Citations
This panel presents information regarding the papers that have cited the interatomic potential (IP) whose page you are on.
The OpenKIM machine learning based Deep Citation framework is used to determine whether the citing article actually used the IP in computations (denoted by "USED") or only provides it as a background citation (denoted by "NOT USED"). For more details on Deep Citation and how to work with this panel, click the documentation link at the top of the panel.
The word cloud to the right is generated from the abstracts of IP principle source(s) (given below in "How to Cite") and the citing articles that were determined to have used the IP in order to provide users with a quick sense of the types of physical phenomena to which this IP is applied.
The bar chart shows the number of articles that cited the IP per year. Each bar is divided into green (articles that USED the IP) and blue (articles that did NOT USE the IP).
Users are encouraged to correct Deep Citation errors in determination by clicking the speech icon next to a citing article and providing updated information. This will be integrated into the next Deep Citation learning cycle, which occurs on a regular basis.
OpenKIM acknowledges the support of the Allen Institute for AI through the Semantic Scholar project for providing citation information and full text of articles when available, which are used to train the Deep Citation ML algorithm.
|
This panel provides information on past usage of this interatomic potential (IP) powered by the OpenKIM Deep Citation framework. The word cloud indicates typical applications of the potential. The bar chart shows citations per year of this IP (bars are divided into articles that used the IP (green) and those that did not (blue)). The complete list of articles that cited this IP is provided below along with the Deep Citation determination on usage. See the Deep Citation documentation for more information.
329 Citations (3 used)
Help us to determine which of the papers that cite this potential actually used it to perform calculations. If you know, click the .
USED (definite) M. Yousefi, “Large-Scale Multiscale Modeling of Phase Transformation in Nanocrystalline Materials: Atomistic and Phase-Field Methods,” arXiv: Materials Science. 2019. link Times cited: 0 Abstract: In this research, atomistic molecular dynamics simulations a… read moreAbstract: In this research, atomistic molecular dynamics simulations are combined with mesoscopic phase-field computational methods in order to investigate phase-transformation in polycrystalline Aluminum microstructure. In fact, microstructural computational modeling of engineering materials could help to optimize their mechanical properties for industrial applications (e.g. directional solidification for turbine blades). As a result, a multiscale modeling approach is developed to find a relation between manufacturing variables (e.g. temperature) and microstructural properties of crystalline materials (e.g. grain size), which could be used to develop an advanced manufacturing process for sensitive applications. The results show that atomistic modeling of grain growth could be used as a first-principle approach in order to study phase transformation's kinetics, which could capture morphology of polycrystalline materials more accurately. On the other hand, phase-field mesoscopic approach needs less computational efforts, but still it relies on semi-empirical data to capture accurate phase transformation regimes, which makes this approach suitable for rapid examining of new manufacturing conditions as well as its effects on microstructural properties of polycrystalline materials. read less USED (high confidence) H. Gao, J. Wang, and J. Sun, “Improve the performance of machine-learning potentials by optimizing descriptors.,” The Journal of chemical physics. 2019. link Times cited: 13 Abstract: Machine-learning (ML) potentials are promising in atomic sim… read moreAbstract: Machine-learning (ML) potentials are promising in atomic simulations due to their comparable accuracy to density functional theory but much lower computational cost. The descriptors to represent atomic environments are of high importance to the performance of ML potentials. Here, we implemented the descriptor in a differentiable way and found that ML potentials with optimized descriptors have some advantages compared with the ones without descriptor optimization, especially when the training dataset is small. Taking aluminum as an example, the trained potentials with proper descriptors can not only predict energies and forces with high accuracy of the first-principles calculations but also reproduce the statistical results of dynamical simulations. These predictions validate the efficiency of our method, which can be applied to improving the performance of machine learning interatomic potentials and will also strongly expand its applications. read less USED (low confidence) F. Z. Dai, B. Wen, H. Xiang, and Y. Zhou, “Grain boundary strengthening in ZrB2 by segregation of W: Atomistic simulations with deep learning potential,” Journal of The European Ceramic Society. 2020. link Times cited: 17 NOT USED (low confidence) Z. Tian, S. Zhang, and G. Chern, “Machine learning for structure-property mapping of Ising models: Scalability and limitations,” Physical Review E. 2023. link Times cited: 0 NOT USED (low confidence) V. Eyert, J. Wormald, W. A. Curtin, and E. Wimmer, “Machine-learned interatomic potentials: Recent developments and prospective applications,” Journal of Materials Research. 2023. link Times cited: 0 NOT USED (low confidence) M. Shi, S. Zhang, and G. Chern, “Machine learning force-field models for metallic spin glass,” ArXiv. 2023. link Times cited: 0 Abstract: Metallic spin glass systems, such as dilute magnetic alloys,… read moreAbstract: Metallic spin glass systems, such as dilute magnetic alloys, are characterized by randomly distributed local moments coupled to each other through a long-range electron-mediated effective interaction. We present a scalable machine learning (ML) framework for dynamical simulations of metallic spin glasses. A Behler-Parrinello type neural-network model, based on the principle of locality, is developed to accurately and efficiently predict electron-induced local magnetic fields that drive the spin dynamics. A crucial component of the ML model is a proper symmetry-invariant representation of local magnetic environment which is direct input to the neural net. We develop such a magnetic descriptor by incorporating the spin degrees of freedom into the atom-centered symmetry function methods which are widely used in ML force-field models for quantum molecular dynamics. We apply our approach to study the relaxation dynamics of an amorphous generalization of the s-d model. Our work highlights the promising potential of ML models for large-scale dynamical modeling of itinerant magnets with quenched disorder. read less NOT USED (low confidence) G. Giunta, G. Campos-Villalobos, and M. Dijkstra, “Coarse-Grained Many-Body Potentials of Ligand-Stabilized Nanoparticles from Machine-Learned Mean Forces,” ACS Nano. 2023. link Times cited: 0 Abstract: Colloidal nanoparticles self-assemble into a variety of supe… read moreAbstract: Colloidal nanoparticles self-assemble into a variety of superstructures with distinctive optical, structural, and electronic properties. These nanoparticles are usually stabilized by a capping layer of organic ligands to prevent aggregation in the solvent. When the ligands are sufficiently long compared to the dimensions of the nanocrystal cores, the effective coarse-grained forces between pairs of nanoparticles are largely affected by the presence of neighboring particles. In order to efficiently investigate the self-assembly behavior of these complex colloidal systems, we propose a machine-learning approach to construct effective coarse-grained many-body interaction potentials. The multiscale methodology presented in this work constitutes a general bottom-up coarse-graining strategy where the coarse-grained forces acting on coarse-grained sites are extracted from measuring the vectorial mean forces on these sites in reference fine-grained simulations. These effective coarse-grained forces, i.e., gradients of the potential of mean force or of the free-energy surface, are represented by a simple linear model in terms of gradients of structural descriptors, which are scalar functions that are rotationally invariant. In this way, we also directly obtain the free-energy surface of the coarse-grained model as a function of all coarse-grained coordinates. We expect that this simple yet accurate coarse-graining framework for the many-body potential of mean force will enable the characterization, understanding, and prediction of the structure and phase behavior of relevant soft-matter systems by direct simulations. The key advantage of this method is its generality, which allows it to be applicable to a broad range of systems. To demonstrate the generality of our method, we also apply it to a colloid–polymer model system, where coarse-grained many-body interactions are pronounced. read less NOT USED (low confidence) M. Ghosh, S. Chowdhury, A. Majumdar, and D. Jana, “Stone–Wales Decorated Phagraphene: A Potential Candidate for Supercapacitor Electrodes and Thermal Transport,” ACS Applied Electronic Materials. 2023. link Times cited: 0 NOT USED (low confidence) Y. Yang, B. Xu, and H. Zong, “Physics infused machine learning force fields for 2D materials monolayers,” Journal of Materials Informatics. 2023. link Times cited: 0 Abstract: Large-scale atomistic simulations of two-dimensional (2D) ma… read moreAbstract: Large-scale atomistic simulations of two-dimensional (2D) materials rely on highly accurate and efficient force fields. Here, we present a physics-infused machine learning framework that enables the efficient development and interpretability of interatomic interaction models for 2D materials. By considering the characteristics of chemical bonds and structural topology, we have devised a set of efficient descriptors. This enables accurate force field training using a small dataset. The machine learning force fields show great success in describing the phase transformation and domain switching behaviors of monolayer Group IV monochalcogenides, e.g., GeSe and PbTe. Notably, this type of force field can be readily extended to other non-transition 2D systems, such as hexagonal boron nitride (h BN), leveraging their structural similarity. Our work provides a straightforward but accurate extension of simulation time and length scales for 2D materials. read less NOT USED (low confidence) L. Ward, B. Blaiszik, C.-W. Lee, T. Martin, I. T. Foster, and A. Schleife, “Accelerating Electronic Stopping Power Predictions by 10 Million Times with a Combination of Time-Dependent Density Functional Theory and Machine Learning,” ArXiv. 2023. link Times cited: 0 Abstract: Knowing the rate at which particle radiation releases energy… read moreAbstract: Knowing the rate at which particle radiation releases energy in a material, the stopping power, is key to designing nuclear reactors, medical treatments, semiconductor and quantum materials, and many other technologies. While the nuclear contribution to stopping power, i.e., elastic scattering between atoms, is well understood in the literature, the route for gathering data on the electronic contribution has for decades remained costly and reliant on many simplifying assumptions, including that materials are isotropic. We establish a method that combines time-dependent density functional theory (TDDFT) and machine learning to reduce the time to assess new materials to mere hours on a supercomputer and provides valuable data on how atomic details influence electronic stopping. Our approach uses TDDFT to compute the electronic stopping contributions to stopping power from first principles in several directions and then machine learning to interpolate to other directions at rates 10 million times higher. We demonstrate the combined approach in a study of proton irradiation in aluminum and employ it to predict how the depth of maximum energy deposition, the"Bragg Peak,"varies depending on incident angle -- a quantity otherwise inaccessible to modelers. The lack of any experimental information requirement makes our method applicable to most materials, and its speed makes it a prime candidate for enabling quantum-to-continuum models of radiation damage. The prospect of reusing valuable TDDFT data for training the model make our approach appealing for applications in the age of materials data science. read less NOT USED (low confidence) C. T. Nguyen, D. T. Ho, and S. Y. Kim, “An Enhanced Sampling Approach for Computing the Free Energy of Solid Surface and Solid–Liquid Interface,” Advanced Theory and Simulations. 2023. link Times cited: 0 Abstract: Free energies of a solid surface and a solid–liquid interfac… read moreAbstract: Free energies of a solid surface and a solid–liquid interface play significant roles in thermodynamics. Due to the limited availability of experimental data, computational methods offer effective alternatives for calculating these properties. This study adopts advanced frameworks of the logarithmic mean force dynamics method to present an enhanced sampling approach for the calculation of the free energy at different temperatures. To achieve this, the free energy profile is constructed along with a pre‐established collective variable within the melting transition and cleavage processes. The values of the solid surface and solid–liquid interface free energies are then extrapolated from the excess free energy related to the formation and persistence of the solid surface or the solid–liquid interface. Furthermore, this methodology is employed to calculate the temperature dependence of the free energy measurements for the (100) and (110) surfaces and interfaces of Cu. It is shown that this methodology is robust and readily applicable in contemporary models of atomic interactions and various systems. read less NOT USED (low confidence) C. Hong et al., “Applications and training sets of machine learning potentials,” Science and Technology of Advanced Materials: Methods. 2023. link Times cited: 0 Abstract: ABSTRACT Recently, machine learning potentials (MLPs) have b… read moreAbstract: ABSTRACT Recently, machine learning potentials (MLPs) have been attracting interest as an alternative to the computationally expensive density-functional theory (DFT) calculations. The data-driven approach in MLPs requires carefully curated training datasets, which define the valid domain of simulations. Therefore, acquiring training datasets that comprehensively span the domain of the desired simulations is important. In this review, we attempt to set guidelines for the systematic construction of training datasets according to target simulations. To this end, we extensively analyze the training sets in previous literature according to four application types: thermal properties, diffusion properties, structure prediction, and chemical reactions. In each application, we summarize characteristic reference structures and discuss specific parameters for DFT calculations such as MD conditions. We hope this review serves as a comprehensive guide for researchers and practitioners aiming to harness the capabilities of MLPs in material simulations. IMPACT STATEMENT This review reports on the selection of training sets for machine learning potentials tailored to their specific applications, which is currently not standardized in the rapidly evolving field. read less NOT USED (low confidence) B. K. Isamura and P. Popelier, “Metaheuristic optimisation of Gaussian process regression model hyperparameters: insights from FEREBUS,” Artificial Intelligence Chemistry. 2023. link Times cited: 0 NOT USED (low confidence) S. Wu et al., “Applications and Advances in Machine Learning Force Fields,” Journal of chemical information and modeling. 2023. link Times cited: 0 Abstract: Force fields (FFs) form the basis of molecular simulations a… read moreAbstract: Force fields (FFs) form the basis of molecular simulations and have significant implications in diverse fields such as materials science, chemistry, physics, and biology. A suitable FF is required to accurately describe system properties. However, an off-the-shelf FF may not be suitable for certain specialized systems, and researchers often need to tailor the FF that fits specific requirements. Before applying machine learning (ML) techniques to construct FFs, the mainstream FFs were primarily based on first-principles force fields (FPFF) and empirical FFs. However, the drawbacks of FPFF and empirical FFs are high cost and low accuracy, respectively, so there is a growing interest in using ML as an effective and precise tool for reconciling this trade-off in developing FFs. In this review, we introduce the fundamental principles of ML and FFs in the context of machine learning force fields (MLFF). We also discuss the advantages and applications of MLFF compared to traditional FFs, as well as the MLFF toolkits widely employed in numerous applications. read less NOT USED (low confidence) N. Fedik et al., “Synergy of semiempirical models and machine learning in computational chemistry.,” The Journal of chemical physics. 2023. link Times cited: 1 Abstract: Catalyzed by enormous success in the industrial sector, many… read moreAbstract: Catalyzed by enormous success in the industrial sector, many research programs have been exploring data-driven, machine learning approaches. Performance can be poor when the model is extrapolated to new regions of chemical space, e.g., new bonding types, new many-body interactions. Another important limitation is the spatial locality assumption in model architecture, and this limitation cannot be overcome with larger or more diverse datasets. The outlined challenges are primarily associated with the lack of electronic structure information in surrogate models such as interatomic potentials. Given the fast development of machine learning and computational chemistry methods, we expect some limitations of surrogate models to be addressed in the near future; nevertheless spatial locality assumption will likely remain a limiting factor for their transferability. Here, we suggest focusing on an equally important effort-design of physics-informed models that leverage the domain knowledge and employ machine learning only as a corrective tool. In the context of material science, we will focus on semi-empirical quantum mechanics, using machine learning to predict corrections to the reduced-order Hamiltonian model parameters. The resulting models are broadly applicable, retain the speed of semiempirical chemistry, and frequently achieve accuracy on par with much more expensive ab initio calculations. These early results indicate that future work, in which machine learning and quantum chemistry methods are developed jointly, may provide the best of all worlds for chemistry applications that demand both high accuracy and high numerical efficiency. read less NOT USED (low confidence) Y. Li, H. Zeng, and H. Zhang, “Atomistic simulations of nucleation and growth of CaCO3 with the influence of inhibitors: A review,” Materials Genome Engineering Advances. 2023. link Times cited: 0 Abstract: Calcium carbonate (CaCO3) is a crucial mineral with great sc… read moreAbstract: Calcium carbonate (CaCO3) is a crucial mineral with great scientific relevance in biomineralization and geoscience. However, excessive precipitation of CaCO3 is posing a threat to industrial production and the aquatic environment. The utilization of chemical inhibitors is typically considered an economical and successful route for addressing the scaling issues, while the underlying mechanism is still debated and needs to be further investigated. In this context, a deep understanding of the crystallization process of CaCO3 and how the inhibitors interact with CaCO3 nuclei and crystals are of great significance in evaluating the performance of scale inhibitors. In recent years, with the rapid development of computing facilities, computer simulations have provided an atomic‐level perspective on the kinetics and thermodynamics of possible association events in CaCO3 solutions as well as the predictions of nucleation pathway and growth mechanism of CaCO3 crystals as a complement to experiment. This review surveys several computational methods and their achievements in this field with a focus on analyzing the functional mechanisms of different types of inhibitors. A general discussion of the current challenges and future directions in applying atomistic simulations to the discovery, design, and development of more effective water‐scale inhibitors is also discussed. read less NOT USED (low confidence) R. Feng et al., “PolyGET: Accelerating Polymer Simulations by Accurate and Generalizable Forcefield with Equivariant Transformer,” ArXiv. 2023. link Times cited: 0 Abstract: Polymer simulation with both accuracy and efficiency is a ch… read moreAbstract: Polymer simulation with both accuracy and efficiency is a challenging task. Machine learning (ML) forcefields have been developed to achieve both the accuracy of ab initio methods and the efficiency of empirical force fields. However, existing ML force fields are usually limited to single-molecule settings, and their simulations are not robust enough. In this paper, we present PolyGET, a new framework for Polymer Forcefields with Generalizable Equivariant Transformers. PolyGET is designed to capture complex quantum interactions between atoms and generalize across various polymer families, using a deep learning model called Equivariant Transformers. We propose a new training paradigm that focuses exclusively on optimizing forces, which is different from existing methods that jointly optimize forces and energy. This simple force-centric objective function avoids competing objectives between energy and forces, thereby allowing for learning a unified forcefield ML model over different polymer families. We evaluated PolyGET on a large-scale dataset of 24 distinct polymer types and demonstrated state-of-the-art performance in force accuracy and robust MD simulations. Furthermore, PolyGET can simulate large polymers with high fidelity to the reference ab initio DFT method while being able to generalize to unseen polymers. read less NOT USED (low confidence) R. Feng et al., “May the Force be with You: Unified Force-Centric Pre-Training for 3D Molecular Conformations,” ArXiv. 2023. link Times cited: 1 Abstract: Recent works have shown the promise of learning pre-trained … read moreAbstract: Recent works have shown the promise of learning pre-trained models for 3D molecular representation. However, existing pre-training models focus predominantly on equilibrium data and largely overlook off-equilibrium conformations. It is challenging to extend these methods to off-equilibrium data because their training objective relies on assumptions of conformations being the local energy minima. We address this gap by proposing a force-centric pretraining model for 3D molecular conformations covering both equilibrium and off-equilibrium data. For off-equilibrium data, our model learns directly from their atomic forces. For equilibrium data, we introduce zero-force regularization and forced-based denoising techniques to approximate near-equilibrium forces. We obtain a unified pre-trained model for 3D molecular representation with over 15 million diverse conformations. Experiments show that, with our pre-training objective, we increase forces accuracy by around 3 times compared to the un-pre-trained Equivariant Transformer model. By incorporating regularizations on equilibrium data, we solved the problem of unstable MD simulations in vanilla Equivariant Transformers, achieving state-of-the-art simulation performance with 2.45 times faster inference time than NequIP. As a powerful molecular encoder, our pre-trained model achieves on-par performance with state-of-the-art property prediction tasks. read less NOT USED (low confidence) S. Röcken and J. Zavadlav, “Accurate machine learning force fields via experimental and simulation data fusion,” ArXiv. 2023. link Times cited: 0 Abstract: Machine Learning (ML)-based force fields are attracting ever… read moreAbstract: Machine Learning (ML)-based force fields are attracting ever-increasing interest due to their capacity to span spatiotemporal scales of classical interatomic potentials at quantum-level accuracy. They can be trained based on high-fidelity simulations or experiments, the former being the common case. However, both approaches are impaired by scarce and erroneous data resulting in models that either do not agree with well-known experimental observations or are under-constrained and only reproduce some properties. Here we leverage both Density Functional Theory (DFT) calculations and experimentally measured mechanical properties and lattice parameters to train an ML potential of titanium. We demonstrate that the fused data learning strategy can concurrently satisfy all target objectives, thus resulting in a molecular model of higher accuracy compared to the models trained with a single data source. The inaccuracies of DFT functionals at target experimental properties were corrected, while the investigated off-target properties remained largely unperturbed. Our approach is applicable to any material and can serve as a general strategy to obtain highly accurate ML potentials. read less NOT USED (low confidence) J. Hinz, V. Karasiev, S. X. Hu, and D. I. Mihaylov, “Development of a machine-learning-based ionic-force correction model for quantum molecular dynamic simulations of warm dense matter,” Physical Review Materials. 2023. link Times cited: 0 NOT USED (low confidence) P. Zhang and W. Yang, “Toward a general neural network force field for protein simulations: Refining the intramolecular interaction in protein.,” The Journal of chemical physics. 2023. link Times cited: 1 Abstract: Molecular dynamics (MD) is an extremely powerful, highly eff… read moreAbstract: Molecular dynamics (MD) is an extremely powerful, highly effective, and widely used approach to understanding the nature of chemical processes in atomic details for proteins. The accuracy of results from MD simulations is highly dependent on force fields. Currently, molecular mechanical (MM) force fields are mainly utilized in MD simulations because of their low computational cost. Quantum mechanical (QM) calculation has high accuracy, but it is exceedingly time consuming for protein simulations. Machine learning (ML) provides the capability for generating accurate potential at the QM level without increasing much computational effort for specific systems that can be studied at the QM level. However, the construction of general machine learned force fields, needed for broad applications and large and complex systems, is still challenging. Here, general and transferable neural network (NN) force fields based on CHARMM force fields, named CHARMM-NN, are constructed for proteins by training NN models on 27 fragments partitioned from the residue-based systematic molecular fragmentation (rSMF) method. The NN for each fragment is based on atom types and uses new input features that are similar to MM inputs, including bonds, angles, dihedrals, and non-bonded terms, which enhance the compatibility of CHARMM-NN to MM MD and enable the implementation of CHARMM-NN force fields in different MD programs. While the main part of the energy of the protein is based on rSMF and NN, the nonbonded interactions between the fragments and with water are taken from the CHARMM force field through mechanical embedding. The validations of the method for dipeptides on geometric data, relative potential energies, and structural reorganization energies demonstrate that the CHARMM-NN local minima on the potential energy surface are very accurate approximations to QM, showing the success of CHARMM-NN for bonded interactions. However, the MD simulations on peptides and proteins indicate that more accurate methods to represent protein-water interactions in fragments and non-bonded interactions between fragments should be considered in the future improvement of CHARMM-NN, which can increase the accuracy of approximation beyond the current mechanical embedding QM/MM level. read less NOT USED (low confidence) F. Khorobrykh et al., “Cluster structure of ultrahard fullerite revealed by Raman spectroscopy,” Carbon. 2023. link Times cited: 0 NOT USED (low confidence) B. Dou et al., “Machine Learning Methods for Small Data Challenges in Molecular Science.,” Chemical reviews. 2023. link Times cited: 11 Abstract: Small data are often used in scientific and engineering rese… read moreAbstract: Small data are often used in scientific and engineering research due to the presence of various constraints, such as time, cost, ethics, privacy, security, and technical limitations in data acquisition. However, big data have been the focus for the past decade, small data and their challenges have received little attention, even though they are technically more severe in machine learning (ML) and deep learning (DL) studies. Overall, the small data challenge is often compounded by issues, such as data diversity, imputation, noise, imbalance, and high-dimensionality. Fortunately, the current big data era is characterized by technological breakthroughs in ML, DL, and artificial intelligence (AI), which enable data-driven scientific discovery, and many advanced ML and DL technologies developed for big data have inadvertently provided solutions for small data problems. As a result, significant progress has been made in ML and DL for small data challenges in the past decade. In this review, we summarize and analyze several emerging potential solutions to small data challenges in molecular science, including chemical and biological sciences. We review both basic machine learning algorithms, such as linear regression, logistic regression (LR), k-nearest neighbor (KNN), support vector machine (SVM), kernel learning (KL), random forest (RF), and gradient boosting trees (GBT), and more advanced techniques, including artificial neural network (ANN), convolutional neural network (CNN), U-Net, graph neural network (GNN), Generative Adversarial Network (GAN), long short-term memory (LSTM), autoencoder, transformer, transfer learning, active learning, graph-based semi-supervised learning, combining deep learning with traditional machine learning, and physical model-based data augmentation. We also briefly discuss the latest advances in these methods. Finally, we conclude the survey with a discussion of promising trends in small data challenges in molecular science. read less NOT USED (low confidence) J. Lin, R. Tamura, Y. Futamura, T. Sakurai, and T. Miyazaki, “Determination of hyper-parameters in the atomic descriptors for efficient and robust molecular dynamics simulations with machine learning forces.,” Physical chemistry chemical physics : PCCP. 2023. link Times cited: 1 Abstract: The atomic descriptors used in machine learning to predict f… read moreAbstract: The atomic descriptors used in machine learning to predict forces are often high dimensional. In general, by retrieving a significant amount of structural information from these descriptors, accurate force predictions can be achieved. On the other hand, to acquire higher robustness for transferability without overfitting, sufficient reduction of descriptors should be necessary. In this study, we propose a method to automatically determine hyperparameters in the atomic descriptors, aiming to obtain accurate machine learning forces while using a small number of descriptors. Our method focuses on identifying an appropriate threshold cut-off for the variance value of the descriptor components. To demonstrate the effectiveness of our method, we apply it to crystalline, liquid, and amorphous structures in SiO2, SiGe, and Si systems. By using both conventional two-body descriptors and our introduced split-type three-body descriptors, we demonstrate that our method can provide machine learning forces that enable efficient and robust molecular dynamics simulations. read less NOT USED (low confidence) X. Cheng, S. Zhang, P. C. H. Nguyen, S. Azarfar, G. Chern, and S. Baek, “Convolutional neural networks for large-scale dynamical modeling of itinerant magnets,” Physical Review Research. 2023. link Times cited: 1 Abstract: Complex spin textures in itinerant electron magnets hold pro… read moreAbstract: Complex spin textures in itinerant electron magnets hold promises for next-generation memory and information technology. The long-ranged and often frustrated electron-mediated spin interactions in these materials give rise to intriguing localized spin structures such as skyrmions. Yet, simulations of magnetization dynamics for such itinerant magnets are computationally difficult due to the need for repeated solutions to the electronic structure problems. We present a convolutional neural network (CNN) model to accurately and efficiently predict the electron-induced magnetic torques acting on local spins. Importantly, as the convolutional operations with a fixed kernel (receptive field) size naturally take advantage of the locality principle for many-electron systems, CNN offers a scalable machine learning approach to spin dynamics. We apply our approach to enable large-scale dynamical simulations of skyrmion phases in itinerant spin systems. By incorporating the CNN model into Landau-Lifshitz-Gilbert dynamics, our simulations successfully reproduce the relaxation process of the skyrmion phase and stabilize a skyrmion lattice in larger systems. The CNN model also allows us to compute the effective receptive fields, thus providing a systematic and unbiased method for determining the locality of the original electron models. read less NOT USED (low confidence) C. Li, B. Gilbert, S. Farrell, and P. Zarzycki, “Rapid Prediction of a Liquid Structure from a Single Molecular Configuration Using Deep Learning,” Journal of Chemical Information and Modeling. 2023. link Times cited: 1 Abstract: Molecular dynamics simulation is an indispensable tool for u… read moreAbstract: Molecular dynamics simulation is an indispensable tool for understanding the collective behavior of atoms and molecules and the phases they form. Statistical mechanics provides accurate routes for predicting macroscopic properties as time-averages over visited molecular configurations - microstates. However, to obtain convergence, we need a sufficiently long record of visited microstates, which translates to the high-computational cost of the molecular simulations. In this work, we show how to use a point cloud-based deep learning strategy to rapidly predict the structural properties of liquids from a single molecular configuration. We tested our approach using three homogeneous liquids with progressively more complex entities and interactions: Ar, NO, and H2O under varying pressure and temperature conditions within the liquid state domain. Our deep neural network architecture allows rapid insight into the liquid structure, here probed by the radial distribution function, and can be used with molecular/atomistic configurations generated by either simulation, first-principle, or experimental methods. read less NOT USED (low confidence) C. Zeng, S. Sahoo, A. Medford, and A. Peterson, “Phase Stability of Large-Size Nanoparticle Alloy Catalysts at Ab Initio Quality Using a Nearsighted Force-Training Approach,” The Journal of Physical Chemistry C. 2023. link Times cited: 1 Abstract: CoPt nanoparticle catalysts are integral to commercial fuel … read moreAbstract: CoPt nanoparticle catalysts are integral to commercial fuel cells. Such systems are prohibitive to fully characterize with electronic structure calculations. Machine-learned potentials offer a scalable solution; however, such potentials are only reliable if representative training data can be employed, which typically requires large electronic structure calculations. Here, we use the nearsighted-force training approach to make high-fidelity machine-learned predictions on large nanoparticles with $>$5,000 atoms using only systematically generated small structures ranging from 38-168 atoms. The resulting ensemble model shows good accuracy and transferability in describing relative energetics for CoPt nanoparticles with various shapes, sizes and Co compositions. It is found that the fcc(100) surface is more likely to form a L1$_0$ ordered structure than the fcc(111) surface. The energy convex hull of the icosahedron shows the most stable particles have Pt-rich skins and Co-rich underlayers. Although the truncated octahedron is the most stable shape across all sizes of Pt nanoparticles, a crossover to icosahedron exists due to a large downshift of surface energy for CoPt nanoparticle alloys. The downshift can be attributed to strain release on the icosahedron surface due to Co alloying. We introduced a simple empirical model to describe the role of Co alloying in the crossover for CoPt nanoparticles. With Monte-Carlo simulations we additionally searched for the most stable atomic arrangement for a truncated octahedron with equal Pt and Co compositions, and also we studied its order-disorder phase transition. We validated the most stable configurations with a new highly scalable density functional theory code called SPARC. Lastly, the order-disorder phase transition for a CoPt nanoparticle exhibits a lower transition temperature and a smoother transition, compared to the bulk CoPt alloy. read less NOT USED (low confidence) T. Pons et al., “Machine Learning Electron Density Prediction Using Weighted Smooth Overlap of Atomic Positions,” Nanomaterials. 2023. link Times cited: 1 Abstract: Having access to accurate electron densities in chemical sys… read moreAbstract: Having access to accurate electron densities in chemical systems, especially for dynamical systems involving chemical reactions, ion transport, and other charge transfer processes, is crucial for numerous applications in materials chemistry. Traditional methods for computationally predicting electron density data for such systems include quantum mechanical (QM) techniques, such as density functional theory. However, poor scaling of these QM methods restricts their use to relatively small system sizes and short dynamic time scales. To overcome this limitation, we have developed a deep neural network machine learning formalism, which we call deep charge density prediction (DeepCDP), for predicting charge densities by only using atomic positions for molecules and condensed phase (periodic) systems. Our method uses the weighted smooth overlap of atomic positions to fingerprint environments on a grid-point basis and map it to electron density data generated from QM simulations. We trained models for bulk systems of copper, LiF, and silicon; for a molecular system, water; and for two-dimensional charged and uncharged systems, hydroxyl-functionalized graphane, with and without an added proton. We showed that DeepCDP achieves prediction R2 values greater than 0.99 and mean squared error values on the order of 10−5e2 Å−6 for most systems. DeepCDP scales linearly with system size, is highly parallelizable, and is capable of accurately predicting the excess charge in protonated hydroxyl-functionalized graphane. We demonstrate how DeepCDP can be used to accurately track the location of charges (protons) by computing electron densities at a few selected grid points in the materials, thus significantly reducing the computational cost. We also show that our models can be transferable, allowing prediction of electron densities for systems on which it has not been trained but that contain a subset of atomic species on which it has been trained. Our approach can be used to develop models that span different chemical systems and train them for the study of large-scale charge transport and chemical reactions. read less NOT USED (low confidence) Y. Liu et al., “Deep learning inter-atomic potential for irradiation damage in 3C-SiC,” Computational Materials Science. 2023. link Times cited: 0 NOT USED (low confidence) Y. Luo, J. A. Meziere, G. Samolyuk, G. Hart, M. Daymond, and L. K. B’eland, “A Set of Moment Tensor Potentials for Zirconium with Increasing Complexity.,” Journal of chemical theory and computation. 2023. link Times cited: 0 Abstract: Machine learning force fields (MLFFs) are an increasingly po… read moreAbstract: Machine learning force fields (MLFFs) are an increasingly popular choice for atomistic simulations due to their high fidelity and improvable nature. Here we propose a hybrid small-cell approach that combines attributes of both offline and active learning to systematically expand a quantum-mechanical (QM) database while constructing MLFFs with increasing model complexity. Our MLFFs employ the moment tensor potential formalism. During this process, we quantitatively assessed the structural properties, elastic properties, dimer potential energies, melting temperatures, phase stability, point defect formation energies, point defect migration energies, free surface energies, and generalized stacking fault (GSF) energies of Zr as predicted by our MLFFs. Unsurprisingly, the model complexity has a positive correlation with prediction accuracy. We also find that the MLFFs were able to predict the properties of out-of-sample configurations without directly including these specific configurations in the training dataset. Additionally, we generated 100 MLFFs of high complexity (1513 parameters each) that reached different local optima during training. Their predictions cluster around the benchmark DFT values, but subtle physical features such as the location of local minima on the GSF energy surface are washed out by statistical noise. read less NOT USED (low confidence) R. Millán, E. Bello-Jurado, M. Moliner, M. Boronat, and R. Gómez-Bombarelli, “Effect of Framework Composition and NH3 on the Diffusion of Cu+ in Cu-CHA Catalysts Predicted by Machine-Learning Accelerated Molecular Dynamics,” ACS Central Science. 2023. link Times cited: 2 Abstract: Cu-exchanged zeolites rely on mobile solvated Cu+ cations fo… read moreAbstract: Cu-exchanged zeolites rely on mobile solvated Cu+ cations for their catalytic activity, but the role of the framework composition in transport is not fully understood. Ab initio molecular dynamics simulations can provide quantitative atomistic insight but are too computationally expensive to explore large length and time scales or diverse compositions. We report a machine-learning interatomic potential that accurately reproduces ab initio results and effectively generalizes to allow multinanosecond simulations of large supercells and diverse chemical compositions. Biased and unbiased simulations of [Cu(NH3)2]+ mobility show that aluminum pairing in eight-membered rings accelerates local hopping and demonstrate that increased NH3 concentration enhances long-range diffusion. The probability of finding two [Cu(NH3)2]+ complexes in the same cage, which is key for SCR-NOx reaction, increases with Cu content and Al content but does not correlate with the long-range mobility of Cu+. Supporting experimental evidence was obtained from reactivity tests of Cu-CHA catalysts with a controlled chemical composition. read less NOT USED (low confidence) L. Ruano, M. Mandado, and J. Nogueira, “Automatic characterization of drug/amino acid interactions by energy decomposition analysis,” Theoretical Chemistry Accounts. 2023. link Times cited: 0 NOT USED (low confidence) A. K. A. Kandy, K. Rossi, A. Raulin-Foissac, G. Laurens, and J. Lam, “Comparing transferability in neural network approaches and linear models for machine-learning interaction potentials,” Physical Review B. 2023. link Times cited: 4 Abstract: Atomic simulations using machine learning interatomic potent… read moreAbstract: Atomic simulations using machine learning interatomic potential (MLIP) have gained a lot of popularity owing to their accuracy in comparison to conventional empirical potentials. However, the transferability of MLIP to systems outside the training-set poses a significant challenge. Here, we compare the transferability of three MLIP approaches: i) Neural Network Potentials (NNP), ii) Physical LassoLars Interactions Potential (PLIP) and iii) Linear Potentials with Belher-Parrinello descriptors, trained over a small but diverse configuration of zinc oxide polymorphs. We compared the obtained models with density functional theory reference results for physical properties including bulk lattice parameters, surface energies, and vibrational density of states and showed the superiority of both NNP and PLIP models. However, the NNP model performed poorly when compared to the other two linear models for the structural optimization of nanoparticles and molecular dynamics simulation of liquid phases, which are systems outside the training-set. While providing less accurate prediction for solid Zinc Oxides phases, both linear models appear more transferable than NNP when testing for nanoscale systems and liquid phases. Our results are finally rationalized by a combination of different statistical analysis including spread in force evaluation, information imbalance, convex hull calculation and density in descriptor space. read less NOT USED (low confidence) A. R. Tan, S. Urata, S. Goldman, J. C. B. Dietschreit, and R. G’omez-Bombarelli, “Single-model uncertainty quantification in neural network potentials does not consistently outperform model ensembles,” npj Computational Materials. 2023. link Times cited: 6 Abstract: Neural networks (NNs) often assign high confidence to their … read moreAbstract: Neural networks (NNs) often assign high confidence to their predictions, even for points far out of distribution, making uncertainty quantification (UQ) a challenge. When they are employed to model interatomic potentials in materials systems, this problem leads to unphysical structures that disrupt simulations, or to biased statistics and dynamics that do not reflect the true physics. Differentiable UQ techniques can find new informative data and drive active learning loops for robust potentials. However, a variety of UQ techniques, including newly developed ones, exist for atomistic simulations and there are no clear guidelines for which are most effective or suitable for a given case. In this work, we examine multiple UQ schemes for improving the robustness of NN interatomic potentials (NNIPs) through active learning. In particular, we compare incumbent ensemble-based methods against strategies that use single, deterministic NNs: mean-variance estimation (MVE), deep evidential regression, and Gaussian mixture models (GMM). We explore three datasets ranging from in-domain interpolative learning to more extrapolative out-of-domain generalization challenges: rMD17, ammonia inversion, and bulk silica glass. Performance is measured across multiple metrics relating model error to uncertainty. Our experiments show that none of the methods consistently outperformed each other across the various metrics. Ensembling remained better at generalization and for NNIP robustness; MVE only proved effective for in-domain interpolation, while GMM was better out-of-domain; and evidential regression, despite its promise, was not the preferable alternative in any of the cases. More broadly, cost-effective, single deterministic models cannot yet consistently match or outperform ensembling for uncertainty quantification in NNIPs. read less NOT USED (low confidence) A. M. Lewis, P. Lazzaroni, and M. Rossi, “Predicting the electronic density response of condensed-phase systems to electric field perturbations.,” The Journal of chemical physics. 2023. link Times cited: 0 Abstract: We present a local and transferable machine-learning approac… read moreAbstract: We present a local and transferable machine-learning approach capable of predicting the real-space density response of both molecules and periodic systems to homogeneous electric fields. The new method, Symmetry-Adapted Learning of Three-dimensional Electron Responses (SALTER), builds on the symmetry-adapted Gaussian process regression symmetry-adapted learning of three-dimensional electron densities framework. SALTER requires only a small, but necessary, modification to the descriptors used to represent the atomic environments. We present the performance of the method on isolated water molecules, bulk water, and a naphthalene crystal. Root mean square errors of the predicted density response lie at or below 10% with barely more than 100 training structures. Derived polarizability tensors and even Raman spectra further derived from these tensors show good agreement with those calculated directly from quantum mechanical methods. Therefore, SALTER shows excellent performance when predicting derived quantities, while retaining all of the information contained in the full electronic response. Thus, this method is capable of predicting vector fields in a chemical context and serves as a landmark for further developments. read less NOT USED (low confidence) Y. Yang, M. Eldred, J. Zádor, and H. Najm, “Multifidelity Neural Network Formulations for Prediction of Reactive Molecular Potential Energy Surfaces,” Journal of chemical information and modeling. 2023. link Times cited: 0 Abstract: This paper focuses on the development of multifidelity model… read moreAbstract: This paper focuses on the development of multifidelity modeling approaches using neural network surrogates, where training data arising from multiple model forms and resolutions are integrated to predict high-fidelity response quantities of interest at lower cost. We focus on the context of quantum chemistry and the integration of information from multiple levels of theory. Important foundations include the use of symmetry function-based atomic energy vector constructions as feature vectors for representing structures across families of molecules and single-fidelity neural network training capabilities that learn the relationships needed to map feature vectors to potential energy predictions. These foundations are embedded within several multifidelity topologies that decompose the high-fidelity mapping into model-based components, including sequential formulations that admit a general nonlinear mapping across fidelities and discrepancy-based formulations that presume an additive decomposition. Methodologies are first explored and demonstrated on a pair of simple analytical test problems and then deployed for potential energy prediction for C5H5 using B2PLYP-D3/6-311++G(d,p) for high-fidelity simulation data and Hartree-Fock 6-31G for low-fidelity data. For the common case of limited access to high-fidelity data, our computational results demonstrate that multifidelity neural network potential energy surface constructions achieve roughly an order of magnitude improvement, either in terms of test error reduction for equivalent total simulation cost or reduction in total cost for equivalent error. read less NOT USED (low confidence) Z. Tian, S. Zhang, and G. Chern, “Machine learning for structure-property relationships: Scalability and limitations,” ArXiv. 2023. link Times cited: 0 Abstract: We present a scalable machine learning (ML) framework for pr… read moreAbstract: We present a scalable machine learning (ML) framework for predicting intensive properties and particularly classifying phases of many-body systems. Scalability and transferability are central to the unprecedented computational efficiency of ML methods. In general, linear-scaling computation can be achieved through the divide and conquer approach, and the locality of physical properties is key to partitioning the system into sub-domains that can be solved separately. Based on the locality assumption, ML model is developed for the prediction of intensive properties of a finite-size block. Predictions of large-scale systems can then be obtained by averaging results of the ML model from randomly sampled blocks of the system. We show that the applicability of this approach depends on whether the block-size of the ML model is greater than the characteristic length scale of the system. In particular, in the case of phase identification across a critical point, the accuracy of the ML prediction is limited by the diverging correlation length. The two-dimensional Ising model is used to demonstrate the proposed framework. We obtain an intriguing scaling relation between the prediction accuracy and the ratio of ML block size over the spin-spin correlation length. Implications for practical applications are also discussed. read less NOT USED (low confidence) D. M. Anstine and O. Isayev, “Machine Learning Interatomic Potentials and Long-Range Physics,” The Journal of Physical Chemistry. a. 2023. link Times cited: 12 Abstract: Advances in machine learned interatomic potentials (MLIPs), … read moreAbstract: Advances in machine learned interatomic potentials (MLIPs), such as those using neural networks, have resulted in short-range models that can infer interaction energies with near ab initio accuracy and orders of magnitude reduced computational cost. For many atom systems, including macromolecules, biomolecules, and condensed matter, model accuracy can become reliant on the description of short- and long-range physical interactions. The latter terms can be difficult to incorporate into an MLIP framework. Recent research has produced numerous models with considerations for nonlocal electrostatic and dispersion interactions, leading to a large range of applications that can be addressed using MLIPs. In light of this, we present a Perspective focused on key methodologies and models being used where the presence of nonlocal physics and chemistry are crucial for describing system properties. The strategies covered include MLIPs augmented with dispersion corrections, electrostatics calculated with charges predicted from atomic environment descriptors, the use of self-consistency and message passing iterations to propagated nonlocal system information, and charges obtained via equilibration schemes. We aim to provide a pointed discussion to support the development of machine learning-based interatomic potentials for systems where contributions from only nearsighted terms are deficient. read less NOT USED (low confidence) J. López-Zorrilla, X. Aretxabaleta, I. W. Yeu, I. Etxebarria, H. Manzano, and N. Artrith, “ænet-PyTorch: A GPU-supported implementation for machine learning atomic potentials training.,” The Journal of chemical physics. 2023. link Times cited: 4 Abstract: In this work, we present ænet-PyTorch, a PyTorch-based imple… read moreAbstract: In this work, we present ænet-PyTorch, a PyTorch-based implementation for training artificial neural network-based machine learning interatomic potentials. Developed as an extension of the atomic energy network (ænet), ænet-PyTorch provides access to all the tools included in ænet for the application and usage of the potentials. The package has been designed as an alternative to the internal training capabilities of ænet, leveraging the power of graphic processing units to facilitate direct training on forces in addition to energies. This leads to a substantial reduction of the training time by one to two orders of magnitude compared to the central processing unit implementation, enabling direct training on forces for systems beyond small molecules. Here, we demonstrate the main features of ænet-PyTorch and show its performance on open databases. Our results show that training on all the force information within a dataset is not necessary, and including between 10% and 20% of the force information is sufficient to achieve optimally accurate interatomic potentials with the least computational resources. read less NOT USED (low confidence) M. Popov, F. Khorobrykh, S. Klimin, V. Churkin, D. Ovsyannikov, and A. Kvashnin, “Surface Tamm States of 2–5 nm Nanodiamond via Raman Spectroscopy,” Nanomaterials. 2023. link Times cited: 2 Abstract: We observed resonance effects in the Raman scattering of nan… read moreAbstract: We observed resonance effects in the Raman scattering of nanodiamonds with an average size of 2–5 nm excited at a wavelength of 1064 nm (1.16 eV). The resonant Raman spectrum of the 2–5 nm nanodiamonds consists of bands at wavelengths of 1325 and 1600 cm−1, a band at 1100–1250 cm−1, and a plateau in the range from 1420 to 1630 cm−1. When excited away from the resonance (at a wavelength of 405 nm, 3.1 eV), the Raman spectrum consists of only three bands at 1325, 1500, and 1600 cm−1. It is important to note that the additional lines (1500 and 1600 cm−1) belong to the sp3-hybridized carbon bonds. The phonon density of states for the nanodiamonds (~1 nm) was calculated using moment tensor potentials (MTP), a class of machine-learning interatomic potentials. The presence of these modes in agreement with the lattice dynamics indicates the existence of bonds with force constants higher than in single-crystal diamonds. The observed resonant phenomena of the Raman scattering and the increase in the bulk modulus are explained by the presence of Tamm states with an energy of electronic transitions of approximately 1 eV, previously observed on the surface of single-crystal diamonds. read less NOT USED (low confidence) L. Schaaf, E. Fako, S. De, A. Schafer, and G. Csányi, “Accurate energy barriers for catalytic reaction pathways: an automatic training protocol for machine learning force fields,” npj Computational Materials. 2023. link Times cited: 2 NOT USED (low confidence) J. Jiang, L.-C. Xu, F. Li, and J. Shao, “Machine Learning Potential Model Based on Ensemble Bispectrum Feature Selection and Its Applicability Analysis,” Metals. 2023. link Times cited: 2 Abstract: With the continuous improvement of machine learning methods,… read moreAbstract: With the continuous improvement of machine learning methods, building the interatomic machine learning potential (MLP) based on the datasets from quantum mechanics calculations has become an effective technical approach to improving the accuracy of classical molecular dynamics simulation. The Spectral Neighbor Analysis Potential (SNAP) is one of the most commonly used machine learning potentials. It uses the bispectrum to encode the local environment of each atom in the lattice. The hyperparameter jmax controls the mapping complexity and precision between the local environment and the bispectrum descriptor. As the hyperparameter jmax increases, the description will become more accurate, but the number of parameters in the bispectrum descriptor will increase dramatically, increasing the computational complexity. In order to reduce the computational complexity without losing the computational accuracy, this paper proposes a two-level ensemble feature selection method (EFS) for a bispectrum descriptor, combining the perturbation method and the feature selector ensemble strategy. Based on the proposed method, the feature subset is selected from the original dataset of the bispectrum descriptor for building the dimension-reduced MLP. As a method application and validation, the data of Fe, Ni, Cu, Li, Mo, Si, and Ge metal elements are used to train the linear regression model based on SNAP for predicting these metals’ atomic energies and forces them to evaluate the performance of the feature subsets. The experimental results show that, compared to the features of SNAP and qSNAP, the training complexity improvement of our EFS method on the qSNAP feature is more effective than SNAP. Compared with the existing methods, when the feature subset size is 0.7 times that of the original features, the proposed EFS method based on the SSWRP ensemble strategy can achieve the best performance in terms of stability, achieving an average stability of 0.94 across all datasets. The training complexity of the linear regression model is reduced by about half, and the prediction complexity is reduced by about 30%. read less NOT USED (low confidence) K. Ghorbani, P. Mirchi, S. Arabha, A. Rajabpour, and S. Volz, “Lattice Thermal Conductivity and Elastic Modulus of Xn4 (X=Be, Mg and Pt),” SSRN Electronic Journal. 2022. link Times cited: 2 Abstract: The newly synthesized BeN4 monolayer has introduced a novel … read moreAbstract: The newly synthesized BeN4 monolayer has introduced a novel group of 2D materials called nitrogen-rich 2D materials. In the present study, the anisotropic mechanical and thermal properties of three members of this group, BeN4, MgN4, and PtN4, are investigated. To this end, a machine learning-based interatomic potential (MLIP) is developed on the basis of the moment tensor potential (MTP) method and utilized in classical molecular dynamics (MD) simulation. Mechanical properties are calculated by extracting the stress-strain curve and thermal properties by non-equilibrium molecular dynamics (NEMD) method. Acquired results show the anisotropic elastic modulus and lattice thermal conductivity of these materials. Generally, elastic modulus and thermal conductivity in the armchair direction are higher than in the zigzag direction. Also, the elastic anisotropy is almost constant at every temperature for BeN4 and MgN4, while for PtN4, this parameter is decreased by increasing the temperature. The findings of this research are not only evidence of the application of machine learning in MD simulations, but also provide information on the basic anisotropic mechanical and thermal properties of these newly discovered 2D nanomaterials. read less NOT USED (low confidence) P. Gao, Z. Liu, J. Zhang, J.-ao Wang, and G. Henkelman, “A Fast, Low-Cost and Simple Method for Predicting Atomic/Inter-Atomic Properties by Combining a Low Dimensional Deep Learning Model with a Fragment Based Graph Convolutional Network,” Crystals. 2022. link Times cited: 2 Abstract: Calculations with high accuracy for atomic and inter-atomic … read moreAbstract: Calculations with high accuracy for atomic and inter-atomic properties, such as nuclear magnetic resonance (NMR) spectroscopy and bond dissociation energies (BDEs) are valuable for pharmaceutical molecule structural analysis, drug exploration, and screening. It is important that these calculations should include relativistic effects, which are computationally expensive to treat. Non-relativistic calculations are less expensive but their results are less accurate. In this study, we present a computational framework for predicting atomic and inter-atomic properties by using machine-learning in a non-relativistic but accurate and computationally inexpensive framework. The accurate atomic and inter-atomic properties are obtained with a low dimensional deep neural network (DNN) embedded in a fragment-based graph convolutional neural network (F-GCN). The F-GCN acts as an atomic fingerprint generator that converts the atomistic local environments into data for the DNN, which improves the learning ability, resulting in accurate results as compared to experiments. Using this framework, the 13C/1H NMR chemical shifts of Nevirapine and phenol O–H BDEs are predicted to be in good agreement with experimental measurement. read less NOT USED (low confidence) S. Ghosal, N. S. Mondal, S. Chowdhury, and D. Jana, “Two novel phases of germa-graphene: Prediction, electronic and transport applications,” Applied Surface Science. 2022. link Times cited: 2 NOT USED (low confidence) L. Shao et al., “Hierarchical Materials from High Information Content Macromolecular Building Blocks: Construction, Dynamic Interventions, and Prediction.,” Chemical reviews. 2022. link Times cited: 7 Abstract: Hierarchical materials that exhibit order over multiple leng… read moreAbstract: Hierarchical materials that exhibit order over multiple length scales are ubiquitous in nature. Because hierarchy gives rise to unique properties and functions, many have sought inspiration from nature when designing and fabricating hierarchical matter. More and more, however, nature's own high-information content building blocks, proteins, peptides, and peptidomimetics, are being coopted to build hierarchy because the information that determines structure, function, and interfacial interactions can be readily encoded in these versatile macromolecules. Here, we take stock of recent progress in the rational design and characterization of hierarchical materials produced from high-information content blocks with a focus on stimuli-responsive and "smart" architectures. We also review advances in the use of computational simulations and data-driven predictions to shed light on how the side chain chemistry and conformational flexibility of macromolecular blocks drive the emergence of order and the acquisition of hierarchy and also on how ionic, solvent, and surface effects influence the outcomes of assembly. Continued progress in the above areas will ultimately usher in an era where an understanding of designed interactions, surface effects, and solution conditions can be harnessed to achieve predictive materials synthesis across scale and drive emergent phenomena in the self-assembly and reconfiguration of high-information content building blocks. read less NOT USED (low confidence) N. Patsalidis, G. Papamokos, G. Floudas, and V. Harmandaris, “Understanding the Interaction between Polybutadiene and Alumina via Density Functional Theory Calculations and Machine-Learned Atomistic Simulations,” The Journal of Physical Chemistry C. 2022. link Times cited: 2 NOT USED (low confidence) S. Chowdhury, V. Demin, L. Chernozatonskii, and A. Kvashnin, “Ultra-Low Thermal Conductivity of Moiré Diamanes,” Membranes. 2022. link Times cited: 4 Abstract: Ultra-thin diamond membranes, diamanes, are one of the most … read moreAbstract: Ultra-thin diamond membranes, diamanes, are one of the most intriguing quasi-2D films, combining unique mechanical, electronic and optical properties. At present, diamanes have been obtained from bi- or few-layer graphene in AA- and AB-stacking by full hydrogenation or fluorination. Here, we study the thermal conductivity of diamanes obtained from bi-layer graphene with twist angle θ between layers forming a Moiré pattern. The combination of DFT calculations and machine learning interatomic potentials makes it possible to perform calculations of the lattice thermal conductivity of such diamanes with twist angles θ of 13.2∘, 21.8∘ and 27.8∘ using the solution of the phonon Boltzmann transport equation. Obtained results show that Moiré diamanes exhibit a wide variety of thermal properties depending on the twist angle, namely a sharp decrease in thermal conductivity from high for “untwisted” diamanes to ultra-low values when the twist angle tends to 30∘, especially for hydrogenated Moiré diamanes. This effect is associated with high anharmonicity and scattering of phonons related to a strong symmetry breaking of the atomic structure of Moiré diamanes compared with untwisted ones. read less NOT USED (low confidence) T. Braeckevelt et al., “Accurately Determining the Phase Transition Temperature of CsPbI3 via Random-Phase Approximation Calculations and Phase-Transferable Machine Learning Potentials,” Chemistry of Materials. 2022. link Times cited: 6 NOT USED (low confidence) S. M. Moosavi et al., “A Data-Science Approach to Predict the Heat Capacity of Nanoporous Materials,” Nature materials. 2022. link Times cited: 27 NOT USED (low confidence) N. Fedik et al., “Extending machine learning beyond interatomic potentials for predicting molecular properties,” Nature Reviews Chemistry. 2022. link Times cited: 30 NOT USED (low confidence) Y. Hu, J. Musielewicz, Z. W. Ulissi, and A. Medford, “Robust and scalable uncertainty estimation with conformal prediction for machine-learned interatomic potentials,” Machine Learning: Science and Technology. 2022. link Times cited: 14 Abstract: Uncertainty quantification (UQ) is important to machine lear… read moreAbstract: Uncertainty quantification (UQ) is important to machine learning (ML) force fields to assess the level of confidence during prediction, as ML models are not inherently physical and can therefore yield catastrophically incorrect predictions. Established a-posteriori UQ methods, including ensemble methods, the dropout method, the delta method, and various heuristic distance metrics, have limitations such as being computationally challenging for large models due to model re-training. In addition, the uncertainty estimates are often not rigorously calibrated. In this work, we propose combining the distribution-free UQ method, known as conformal prediction (CP), with the distances in the neural network’s latent space to estimate the uncertainty of energies predicted by neural network force fields. We evaluate this method (CP+latent) along with other UQ methods on two essential aspects, calibration, and sharpness, and find this method to be both calibrated and sharp under the assumption of independent and identically-distributed (i.i.d.) data. We show that the method is relatively insensitive to hyperparameters selected, and test the limitations of the method when the i.i.d. assumption is violated. Finally, we demonstrate that this method can be readily applied to trained neural network force fields with traditional and graph neural network architectures to obtain estimates of uncertainty with low computational costs on a training dataset of 1 million images to showcase its scalability and portability. Incorporating the CP method with latent distances offers a calibrated, sharp and efficient strategy to estimate the uncertainty of neural network force fields. In addition, the CP approach can also function as a promising strategy for calibrating uncertainty estimated by other approaches. read less NOT USED (low confidence) D. Bishara, Y. Xie, W. K. Liu, and S. Li, “A State-of-the-Art Review on Machine Learning-Based Multiscale Modeling, Simulation, Homogenization and Design of Materials,” Archives of Computational Methods in Engineering. 2022. link Times cited: 24 NOT USED (low confidence) Y.-L. Lee et al., “Data-Driven Enhancement of ZT in SnSe-Based Thermoelectric Systems.,” Journal of the American Chemical Society. 2022. link Times cited: 10 Abstract: Doping and alloying are fundamental strategies to improve th… read moreAbstract: Doping and alloying are fundamental strategies to improve the thermoelectric performance of bare materials. However, identifying outstanding elements and compositions for the development of high-performance thermoelectric materials is challenging. In this study, we present a data-driven approach to improve the thermoelectric performance of SnSe compounds with various doping. Based on the newly generated experimental and computational dataset, we built highly accurate predictive models of thermoelectric properties of doped SnSe compounds. A well-designed feature vector consisting of the chemical properties of a single atom and the electronic structures of a solid plays a key role in achieving accurate predictions for unknown doping elements. Using the machine learning predictive models and calculated map of the solubility limit for each dopant, we rapidly screened high-dimensional material spaces of doped SnSe and evaluated their thermoelectric properties. This data-driven search provided overall strategies to optimize and improve the thermoelectric properties of doped SnSe compounds. In particular, we identified five dopant candidate elements (Ge, Pb, Y, Cd, and As) that provided a high ZT exceeding 2.0 and proposed a design principle for improving the ZT by Sn vacancies depending on the doping elements. Based on the search, we proposed yttrium as a new high-ZT dopant for SnSe with experimental confirmations. Our research is expected to lead to novel high-ZT thermoelectric material candidates and provide cutting-edge research strategies for materials design and extraction of design principles through data-driven research. read less NOT USED (low confidence) B. Xi et al., “Machine-Learning-Assisted Acceleration on High-Symmetry Materials Search: Space Group Predictions from Band Structures,” The Journal of Physical Chemistry C. 2022. link Times cited: 1 NOT USED (low confidence) J. Fox, B. Zhao, B. G. del Rio, S. Rajamanickam, R. Ramprasad, and L. Song, “Concentric Spherical Neural Network for 3D Representation Learning,” 2022 International Joint Conference on Neural Networks (IJCNN). 2022. link Times cited: 1 Abstract: Learning 3D representations of point clouds that generalize … read moreAbstract: Learning 3D representations of point clouds that generalize well to arbitrary orientations is a challenge of practical importance in domains ranging from computer vision to molecular modeling. The proposed approach uses a concentric spherical spatial representation, formed by nesting spheres discretized the icosahedral grid, as the basis for structured learning over point clouds. We propose rotationally equivariant convolutions for learning over the concentric spherical grid, which are incorporated into a novel architecture for representation learning that is robust to general rotations in 3D. We demonstrate the effectiveness and extensibility of our approach to problems in different domains, such as 3D shape recognition and predicting fundamental properties of molecular systems. read less NOT USED (low confidence) S. Pathak et al., “Accurate hellmann-feynman forces with optimized atom-centered gaussian basis sets .,” Proposed for presentation at the Computational Materials Science and Engineering: Gordon Research Conference: Comparing Theories, Algorithms and Computation Protocols in Materials Sci held July 31-August 5, 2022 in Newry, ME. 2022. link Times cited: 1 Abstract: The Hellmann–Feynman (HF) theorem provides a way to compute … read moreAbstract: The Hellmann–Feynman (HF) theorem provides a way to compute forces directly from the electron density, affording an approach to calculating forces of large systems with machine learning (ML) models that predict electron density. The primary issue holding back the general acceptance of the HF approach for atom-centered basis sets is the well-known Pulay force which, if naively discarded, typically constitutes an error upwards of 10 eV/Ang in forces. In this work, we construct specialized atom-centered Gaussian basis sets to reduce the Pulay force, and demonstrate the basis sets’ effectiveness in computing accurate HF forces. We find that HF forces computed using the σ NZHF (N = S ingle, D ouble, T riple) basis sets developed in this work yield comparable accuracy to forces computed with the Pulay term using size matched cc-pVNZ [1] and pcseg-N [2] basis sets for water clusters and pcseg-N and aug-pcseg-N basis sets for DNA fragments. Our results illustrate that the σ NZHF basis sets yield HF forces with state-of-the-art accuracy, paving a clear path for-wards for accurate and efficient calculations of forces for large systems using the HF theorem and ML densities. read less NOT USED (low confidence) S. Chowdhury, S. Ghosal, D. Mondal, and D. Jana, “First-principles and machine-learning study of electronic and phonon transport in carbon-based AA-stacked bilayer biphenylene nanosheets,” Journal of Physics and Chemistry of Solids. 2022. link Times cited: 5 NOT USED (low confidence) E. Minamitani, I. Obayashi, K. Shimizu, and S. Watanabe, “Persistent homology-based descriptor for machine-learning potential of amorphous structures.,” The Journal of chemical physics. 2022. link Times cited: 0 Abstract: High-accuracy prediction of the physical properties of amorp… read moreAbstract: High-accuracy prediction of the physical properties of amorphous materials is challenging in condensed-matter physics. A promising method to achieve this is machine-learning potentials, which is an alternative to computationally demanding ab initio calculations. When applying machine-learning potentials, the construction of descriptors to represent atomic configurations is crucial. These descriptors should be invariant to symmetry operations. Handcrafted representations using a smooth overlap of atomic positions and graph neural networks (GNN) are examples of methods used for constructing symmetry-invariant descriptors. In this study, we propose a novel descriptor based on a persistence diagram (PD), a two-dimensional representation of persistent homology (PH). First, we demonstrated that the normalized two-dimensional histogram obtained from PD could predict the average energy per atom of amorphous carbon at various densities, even when using a simple model. Second, an analysis of the dimensional reduction results of the descriptor spaces revealed that PH can be used to construct descriptors with characteristics similar to those of a latent space in a GNN. These results indicate that PH is a promising method for constructing descriptors suitable for machine-learning potentials without hyperparameter tuning and deep-learning techniques. read less NOT USED (low confidence) C. M. de Armas-Morejón, L. Montero-Cabrera, A. Rubio, and J. Jornet-Somoza, “Electronic Descriptors for Supervised Spectroscopic Predictions,” Journal of Chemical Theory and Computation. 2022. link Times cited: 1 Abstract: Spectroscopic properties of molecules hold great importance … read moreAbstract: Spectroscopic properties of molecules hold great importance for the description of the molecular response under the effect of UV/vis electromagnetic radiation. Computationally expensive ab initio (e.g., MultiConfigurational SCF, Coupled Cluster) or TDDFT methods are commonly used by the quantum chemistry community to compute these properties. In this work, we propose a (supervised) Machine Learning approach to model the absorption spectra of organic molecules. Several supervised ML methods have been tested such as Kernel Ridge Regression (KRR), Multiperceptron Neural Networs (MLP), and Convolutional Neural Networks. [Ramakrishnan et al. J. Chem. Phys.2015, 143, 084111.26328822Ghosh et al. Adv. Sci.2019, 6, 1801367.] The use of only geometrical-atomic number descriptors (e.g., Coulomb Matrix) proved to be insufficient for an accurate training. [Ramakrishnan et al. J. Chem. Phys.2015, 143, 084111.26328822] Inspired by the TDDFT theory, we propose to use a set of electronic descriptors obtained from low-cost DFT methods: orbital energy differences (Δϵia = ϵa – ϵi), transition dipole moment between occupied and unoccupied Kohn–Sham orbitals (⟨ϕi|r|ϕa⟩), and when relevant, charge-transfer character of monoexcitations (Ria). We demonstrate that with these electronic descriptors and the use of Neural Networks we can predict not only a density of excited states but also get a very good estimation of the absorption spectrum and charge-transfer character of the electronic excited states, reaching results close to chemical accuracy (∼2 kcal/mol or ∼0.1 eV). read less NOT USED (low confidence) J. Zhang et al., “Machine Learning Prediction of Superconducting Critical Temperature through the Structural Descriptor,” The Journal of Physical Chemistry C. 2022. link Times cited: 11 NOT USED (low confidence) K. Fujioka and R. Sun, “Interpolating Moving Ridge Regression (IMRR): A machine learning algorithm to predict energy gradients for ab initio molecular dynamics simulations,” Chemical Physics. 2022. link Times cited: 3 NOT USED (low confidence) P. Gao et al., “Graph Convolutional Network-Based Screening Strategy for Rapid Identification of SARS-CoV-2 Cell-Entry Inhibitors,” Journal of Chemical Information and Modeling. 2022. link Times cited: 1 Abstract: The cell entry of SARS-CoV-2 has emerged as an attractive dr… read moreAbstract: The cell entry of SARS-CoV-2 has emerged as an attractive drug development target. We previously reported that the entry of SARS-CoV-2 depends on the cell surface heparan sulfate proteoglycan (HSPG) and the cortex actin, which can be targeted by therapeutic agents identified by conventional drug repurposing screens. However, this drug identification strategy requires laborious library screening, which is time consuming, and often limited number of compounds can be screened. As an alternative approach, we developed and trained a graph convolutional network (GCN)-based classification model using information extracted from experimentally identified HSPG and actin inhibitors. This method allowed us to virtually screen 170,000 compounds, resulting in ∼2000 potential hits. A hit confirmation assay with the uptake of a fluorescently labeled HSPG cargo further shortlisted 256 active compounds. Among them, 16 compounds had modest to strong inhibitory activities against the entry of SARS-CoV-2 pseudotyped particles into Vero E6 cells. These results establish a GCN-based virtual screen workflow for rapid identification of new small molecule inhibitors against validated drug targets. read less NOT USED (low confidence) E. A. Bamidele et al., “Discovery and prediction capabilities in metal-based nanomaterials: An overview of the application of machine learning techniques and some recent advances,” Adv. Eng. Informatics. 2022. link Times cited: 10 NOT USED (low confidence) V. A. Aristizabal-Ferreira, J. M. Guevara‐Vela, A. S.-de la Vega, Á. Pendás, G. Fuentes-Pineda, and T. Rocha‐Rinza, “Computation of photovoltaic and stability properties of hybrid organic–inorganic perovskites via convolutional neural networks,” Theoretical Chemistry Accounts. 2022. link Times cited: 2 NOT USED (low confidence) C. H. Chan, M. Sun, and B. Huang, “Application of machine learning for advanced material prediction and design,” EcoMat. 2022. link Times cited: 23 NOT USED (low confidence) A. Mannodi-Kanakkithodi and M. Chan, “Accelerated screening of functional atomic impurities in halide perovskites using high-throughput computations and machine learning,” Journal of Materials Science. 2022. link Times cited: 7 NOT USED (low confidence) M. Cherukara and A. Mannodi-Kanakkithodi, “Deep learning the properties of inorganic perovskites,” Modelling and Simulation in Materials Science and Engineering. 2022. link Times cited: 1 Abstract: The ability to accurately and quickly predict the stability … read moreAbstract: The ability to accurately and quickly predict the stability of materials and their structural and electronic properties remains a grand challenge in materials science. Density functional theory is widely used as a means of predicting these material properties, but is known to be computationally expensive and scales as the cube of the number of electrons in the material’s unit cell. In this article, for a previously published dataset of inorganic perovskites, we show that a single neural network model using only the elemental properties of the compounds’ constituents can predict lattice constants to within 0.1 Å, heat of formation to within 0.2 eV, and band gaps to within 0.7 eV RMSE. We also compare the performance of the trained network to two widely used regression techniques, namely random forest and Kernel ridge regression, and find that the neural network’s predictions are more accurate for each of the properties. The simultaneous accurate prediction of multiple key properties of technologically relevant materials is promising for rational design and optimization in known and novel chemical spaces. read less NOT USED (low confidence) Z. Wei, C. Zhang, Y. Kan, Y. Zhang, and Y. Chen, “Developing machine learning potential for classical molecular dynamics simulation with superior phonon properties,” Computational Materials Science. 2022. link Times cited: 1 NOT USED (low confidence) Y. Yu, R. Wang, and R. D. Teo, “Machine Learning Approaches for Metalloproteins,” Molecules. 2022. link Times cited: 4 Abstract: Metalloproteins are a family of proteins characterized by me… read moreAbstract: Metalloproteins are a family of proteins characterized by metal ion binding, whereby the presence of these ions confers key catalytic and ligand-binding properties. Due to their ubiquity among biological systems, researchers have made immense efforts to predict the structural and functional roles of metalloproteins. Ultimately, having a comprehensive understanding of metalloproteins will lead to tangible applications, such as designing potent inhibitors in drug discovery. Recently, there has been an acceleration in the number of studies applying machine learning to predict metalloprotein properties, primarily driven by the advent of more sophisticated machine learning algorithms. This review covers how machine learning tools have consolidated and expanded our comprehension of various aspects of metalloproteins (structure, function, stability, ligand-binding interactions, and inhibitors). Future avenues of exploration are also discussed. read less NOT USED (low confidence) Y. Wang et al., “Accelerated Strategy for Fast Ion Conductor Materials Screening and Optimal Doping Scheme Exploration,” Journal of Materiomics. 2022. link Times cited: 0 NOT USED (low confidence) H. Doi, K. Z. Takahashi, and T. Aoyagi, “Screening toward the Development of Fingerprints of Atomic Environments Using Bond-Orientational Order Parameters,” ACS Omega. 2022. link Times cited: 1 Abstract: A combination of atomic numbers and bond-orientational order… read moreAbstract: A combination of atomic numbers and bond-orientational order parameters is considered a candidate for a simple representation that involves information on both the atomic species and their positional relation. The 504 candidates are applied as the fingerprint of the molecules stored in QM9, a data set of computed geometric, energetic, electronic, and thermodynamic properties for 133 885 stable small organic molecules made up of carbon, hydrogen, oxygen, nitrogen, and fluorine atoms. To screen the fingerprints, a regression analysis of the atomic charges given by Open Babel was performed by supervised machine learning. The regression results indicate that the 60 fingerprints successfully estimate Open Babel charges. The results of the dipole moments, an example of a property expressed by charge and position, also had a high accuracy in comparison with the values computed from Open Babel charges. Therefore, the screened 60 fingerprints have the potential to precisely describe the chemical and structural information on the atomic environment of molecules. read less NOT USED (low confidence) A. Pandey, J. Gigax, and R. Pokharel, “Machine Learning Interatomic Potential for High-Throughput Screening of High-Entropy Alloys,” JOM. 2022. link Times cited: 1 NOT USED (low confidence) Q. Feng et al., “Recent Progress and Future Prospects on All-Organic Polymer Dielectrics for Energy Storage Capacitors.,” Chemical reviews. 2021. link Times cited: 151 Abstract: With the development of advanced electronic devices and elec… read moreAbstract: With the development of advanced electronic devices and electric power systems, polymer-based dielectric film capacitors with high energy storage capability have become particularly important. Compared with polymer nanocomposites with widespread attention, all-organic polymers are fundamental and have been proven to be more effective choices in the process of scalable, continuous, and large-scale industrial production, leading to many dielectric and energy storage applications. In the past decade, efforts have intensified in this field with great progress in newly discovered dielectric polymers, fundamental production technologies, and extension toward emerging computational strategies. This review summarizes the recent progress in the field of energy storage based on conventional as well as heat-resistant all-organic polymer materials with the focus on strategies to enhance the dielectric properties and energy storage performances. The key parameters of all-organic polymers, such as dielectric constant, dielectric loss, breakdown strength, energy density, and charge-discharge efficiency, have been thoroughly studied. In addition, the applications of computer-aided calculation including density functional theory, machine learning, and materials genome in rational design and performance prediction of polymer dielectrics are reviewed in detail. Based on a comprehensive understanding of recent developments, guidelines and prospects for the future development of all-organic polymer materials with dielectric and energy storage applications are proposed. read less NOT USED (low confidence) W. B. How, B. Wang, W. Chu, A. Tkatchenko, and O. Prezhdo, “Significance of the Chemical Environment of an Element in Nonadiabatic Molecular Dynamics: Feature Selection and Dimensionality Reduction with Machine Learning.,” The journal of physical chemistry letters. 2021. link Times cited: 8 Abstract: Using supervised and unsupervised machine learning (ML) on f… read moreAbstract: Using supervised and unsupervised machine learning (ML) on features generated from nonadiabatic (NA) molecular dynamics (MD) trajectories under the classical path approximation, we demonstrate that mutual information with the NA Hamiltonian can be used for feature selection and model simplification. Focusing on CsPbI3, a popular metal halide perovskite, we observe that the chemical environment of a single element is sufficient for predicting the NA Hamiltonian. The conclusion applies even to Cs, although Cs does not contribute to the relevant wave functions. Interatomic distances between Cs and I or Pb and the octahedral tilt angle are the most important features. We reduce a typical 360-parameter ML force-field model to just a 12-parameter NA Hamiltonian model, while maintaining a high NA-MD simulation quality. Because NA-MD is a valuable tool for studying excited state processes, overcoming its high computational cost through simple ML models will streamline NA-MD simulations and expand the ranges of accessible system size and simulation time. read less NOT USED (low confidence) N. S. Mondal, S. Nath, S. Chowdhury, and D. Jana, “Electric field-induced electronic-thermoelectric-optical properties of typical isoelectronic HNC6 monolayers: a theoretical study,” Applied Surface Science. 2021. link Times cited: 2 NOT USED (low confidence) R. Magar et al., “AugLiChem: data augmentation library of chemical structures for machine learning,” Machine Learning: Science and Technology. 2021. link Times cited: 11 Abstract: Machine learning (ML) has demonstrated the promise for accur… read moreAbstract: Machine learning (ML) has demonstrated the promise for accurate and efficient property prediction of molecules and crystalline materials. To develop highly accurate ML models for chemical structure property prediction, datasets with sufficient samples are required. However, obtaining clean and sufficient data of chemical properties can be expensive and time-consuming, which greatly limits the performance of ML models. Inspired by the success of data augmentations in computer vision and natural language processing, we developed AugLiChem: the data augmentation library for chemical structures. Augmentation methods for both crystalline systems and molecules are introduced, which can be utilized for fingerprint-based ML models and graph neural networks (GNNs). We show that using our augmentation strategies significantly improves the performance of ML models, especially when using GNNs. In addition, the augmentations that we developed can be used as a direct plug-in module during training and have demonstrated the effectiveness when implemented with different GNN models through the AugliChem library. The Python-based package for our implementation of Auglichem: Data augmentation library for chemical structures, is publicly available at: https://github.com/BaratiLab/AugLiChem. read less NOT USED (low confidence) S. Ghosal, S. Chowdhury, and D. Jana, “Impressive Thermoelectric Figure of Merit in Two-Dimensional Tetragonal Pnictogens: a Combined First-Principles and Machine-Learning Approach.,” ACS applied materials & interfaces. 2021. link Times cited: 21 Abstract: Over the past decade, two-dimensional materials have gained … read moreAbstract: Over the past decade, two-dimensional materials have gained a lot of interest due to their fascinating applications in the field of thermoelectricity. In this study, tetragonal monolayers of group-V elements (T-P, T-As, T-Sb, and T-Bi) are systematically analyzed in the framework of density functional theory in combination with the machine-learning approach. The phonon spectra, as well as the strain profile, dictate that these tetragonal structures are geometrically stable as well as they are potential candidates for experimental synthesis. Electronic analysis suggests that tetragonal pnictogens offer a band gap in the semiconducting regime. Thermal transport characteristics are investigated by solving the semiclassical Boltzmann transport equation. Exceptionally low lattice thermal conductivity has been observed as the atomic number increases in the group. The high Seebeck coefficient and electrical conductivity as well as the low thermal conductivity of T-As, T-Sb, and T-Bi lead to the generation of a very high thermoelectric figure of merit as compared to standard thermoelectric materials. Furthermore, the thermoelectric conversion efficiency of these materials has been observed to be much higher, which ensures their implications in thermoelectric device engineering. read less NOT USED (low confidence) D. Roy, S. Mandal, and B. Pathak, “Machine Learning-Driven High-Throughput Screening of Alloy-Based Catalysts for Selective CO2 Hydrogenation to Methanol.,” ACS applied materials & interfaces. 2021. link Times cited: 35 Abstract: The revolutionary development of machine learning and data s… read moreAbstract: The revolutionary development of machine learning and data science and exploration of its application in material science are huge achievements of the scientific community in the past decade. In this work, we have reported an efficient approach of machine learning-aided high-throughput screening for finding selective earth-abundant high-entropy alloy-based catalysts for CO2 to methanol formation using a machine learning algorithm and microstructure model. For this, we have chosen earth-abundant Cu, Co, Ni, Zn, and Mg metals to form various alloy-based compositions (bimetallic, trimetallic, tetrametallic, and high-entropy alloys) for selective CO2 reduction reaction toward CH3OH. Since there are several possible surface microstructures for different alloys, we have used machine learning along with DFT calculations for high-throughput screening of the catalysts. In this study, the stability of various 8-atom fcc periodic (111) surface unit cells has been calculated using the atomic-size difference factor (δ) as well as the ratio taken from Gibbs free energy of mixing (Ω). Thinking about the simplicity and accuracy, microstructure models by considering the neighboring atoms of the adsorption sites and others as Cu atoms have been considered for different adsorption sites (on-top, bridge, and hollow-hcp). Moreover, the adsorption energies of the *H, *O, *CO, *HCO, *H2CO, and *H3CO intermediates have been predicted using the best fitted algorithm of the training set. The predicted adsorption energies have been screened based on the pure Cu adsorption energy. Furthermore, the screened catalysts have been correlated among different adsorption site microstructures. At the end, we were able to find seven active catalysts, among which two catalysts are CuCoNiZn-based tetrametallic, three catalysts are CuNiZn-based trimetallic, and two catalysts are CuCoZn-based trimetallic alloys. Hence, this work demonstrates not an ultimate but an efficient approach for finding new product-selective catalysts, and we expect that it can be convenient for other similar types of reactions in forthcoming days. read less NOT USED (low confidence) B. Javvaji, B. Mortazavi, X. Zhuang, and T. Rabczuk, “Exploring tensile piezoelectricity and bending flexoelectricity of diamane monolayers by machine learning,” Carbon. 2021. link Times cited: 10 NOT USED (low confidence) B. Kim, Y. Shao, and J. Pu, “Doubly Polarized QM/MM with Machine Learning Chaperone Polarizability.,” Journal of chemical theory and computation. 2021. link Times cited: 4 Abstract: A major shortcoming of semiempirical (SE) molecular orbital … read moreAbstract: A major shortcoming of semiempirical (SE) molecular orbital methods is their severe underestimation of molecular polarizability compared with experimental and ab initio (AI) benchmark data. In a combined quantum mechanical and molecular mechanical (QM/MM) treatment of solution-phase reactions, solute described by SE methods therefore tends to generate inadequate electronic polarization response to solvent electric fields, which often leads to large errors in free energy profiles. To address this problem, here we present a hybrid framework that improves the response property of SE/MM methods through high-level molecular-polarizability fitting. Specifically, we place on QM atoms a set of corrective polarizabilities (referred to as chaperone polarizabilities), whose magnitudes are determined from machine learning (ML) to reproduce the condensed-phase AI molecular polarizability along the minimum free energy path. These chaperone polarizabilities are then used in a machinery similar to a polarizable force field calculation to compensate for the missing polarization energy in the conventional SE/MM simulations. Because QM atoms in this treatment host SE wave functions as well as classical polarizabilities, both polarized by MM electric fields, we name this method doubly polarized QM/MM (dp-QM/MM). We demonstrate the new method on the free energy simulations of the Menshutkin reaction in water. Using AM1/MM as a base method, we show that ML chaperones greatly reduce the error in the solute molecular polarizability from 6.78 to 0.03 Å3 with respect to the density functional theory benchmark. The chaperone correction leads to ∼10 kcal/mol of additional polarization energy in the product region, bringing the simulated free energy profiles to closer agreement with the experimental results. Furthermore, the solute-solvent radial distribution functions show that the chaperone polarizabilities modify the free energy profiles through enhanced solvation corrections when the system evolves from the charge-neutral reactant state to the charge-separated transition and product states. These results suggest that the dp-QM/MM method, enabled by ML chaperone polarizabilities, provides a very physical remedy for the underpolarization problem in SE/MM-based free energy simulations. read less NOT USED (low confidence) J. Peng, J. Guo, R. Ma, and Y. Jiang, “Water-solid interfaces probed by high-resolution atomic force microscopy,” Surface Science Reports. 2021. link Times cited: 14 NOT USED (low confidence) L. Ugwu, Y. Morgan, and H. Ibrahim, “Application of density functional theory and machine learning in heterogenous-based catalytic reactions for hydrogen production,” International Journal of Hydrogen Energy. 2021. link Times cited: 12 NOT USED (low confidence) H. Yao, J. Liu, M. Xu, J. Ji, Q. Dai, and Z. You, “Discussion on molecular dynamics (MD) simulations of the asphalt materials.,” Advances in colloid and interface science. 2021. link Times cited: 57 NOT USED (low confidence) C. M. Wolf, L. Guio, S. Scheiwiller, V. Pakhnyuk, C. Luscombe, and L. Pozzo, “Strategies for the Development of Conjugated Polymer Molecular Dynamics Force Fields Validated with Neutron and X-ray Scattering,” ACS Polymers Au. 2021. link Times cited: 9 Abstract: Conjugated polymers (CPs) enable a wide range of lightweight… read moreAbstract: Conjugated polymers (CPs) enable a wide range of lightweight, lower cost, and flexible organic electronic devices, but a thorough understanding of relationships between molecular structure and dynamics and electronic performance is critical for improved device efficiencies and for new technologies. Molecular dynamics (MD) simulations offer in silico insight into this relationship, but their accuracy relies on the approach used to develop the model’s parameters or force field (FF). In this Perspective, we first review current FFs for CPs and find that most of the models implement an arduous reparameterization of inter-ring torsion potentials and partial charges of classical FFs. However, there are few FFs outside of simple CP molecules, e.g., polythiophenes, that have been developed over the last two decades. There is also limited reparameterization of other parameters, such as nonbonded Lennard-Jones interactions, which we find to be directly influenced by conjugation in these materials. We further provide a discussion on experimental validation of MD FFs, with emphasis on neutron and X-ray scattering. We define multiple ways in which various scattering methods can be directly compared to results of MD simulations, providing a powerful experimental validation metric of local structure and dynamics at relevant length and time scales to charge transport mechanisms in CPs. Finally, we offer a perspective on the use of neutron scattering with machine learning to enable high-throughput parametrization of accurate and experimentally validated CP FFs enabled not only by the ongoing advancements in computational chemistry, data science, and high-performance computing but also using oligomers as proxies for longer polymer chains during FF development. read less NOT USED (low confidence) S. Satsangi, A. Mishra, and A. K. Singh, “Feature Blending: An Approach toward Generalized Machine Learning Models for Property Prediction,” ACS Physical Chemistry Au. 2021. link Times cited: 4 Abstract: From studying the atomic structure and chemical behavior to … read moreAbstract: From studying the atomic structure and chemical behavior to the discovery of new materials and investigating properties of existing materials, machine learning (ML) has been employed in realms that are arduous to probe experimentally. While numerous highly accurate models, specifically for property prediction, have been reported in the literature, there has been a lack of a generalized framework. Herein we propose a novel feature selection approach that enables the development of a unified ML model for property prediction for several classes of materials. It involves an ingenious blending of selected features from various classes of data such that the resultant feature set equips the model with global data descriptors capturing both class-specific as well as global traits. We took accurate band gaps of three distinct classes of 2D materials as our target property to develop the proposed feature blending approach. Using Gaussian process regression (GPR) with the blended features, the ML model developed here resulted in an average root-mean-squared error of 0.12 eV for unseen data belonging to any of the participating classes. The feature blending approach proposed here can be extended to additional classes of materials and also to predict other properties. read less NOT USED (low confidence) P. Liu, C. Verdi, F. Karsai, and G. Kresse, “Phase transitions of zirconia: Machine-learned force fields beyond density functional theory,” Physical Review B. 2021. link Times cited: 12 Abstract: We present an approach to generate machine-learned force fie… read moreAbstract: We present an approach to generate machine-learned force fields (MLFF) with beyond density functional theory (DFT) accuracy. Our approach combines on-the-fly active learning and $\Delta$-machine learning in order to generate an MLFF for zirconia based on the random phase approximation (RPA). Specifically, an MLFF trained on-the-fly during DFT based molecular dynamics simulations is corrected by another MLFF that is trained on the differences between RPA and DFT calculated energies, forces and stress tensors. Thanks to the relatively smooth nature of the differences, the expensive RPA calculations are performed only on a small number of representative structures of small unit cells. These structures are determined by a singular value decomposition rank compression of the kernel matrix with low spatial resolution. This dramatically reduces the computational cost and allows us to generate an MLFF fully capable of reproducing high-level quantum-mechanical calculations beyond DFT. We carefully validate our approach and demonstrate its success in studying the phase transitions of zirconia. read less NOT USED (low confidence) L. Pereira, “Investigating mechanical properties and thermal conductivity of 2D carbon-based materials by computational experiments,” Computational Materials Science. 2021. link Times cited: 9 NOT USED (low confidence) I. Poltavsky and A. Tkatchenko, “Machine Learning Force Fields: Recent Advances and Remaining Challenges.,” The journal of physical chemistry letters. 2021. link Times cited: 42 Abstract: In chemistry and physics, machine learning (ML) methods prom… read moreAbstract: In chemistry and physics, machine learning (ML) methods promise transformative impacts by advancing modeling and improving our understanding of complex molecules and materials. Each ML method comprises a mathematically well-defined procedure, and an increasingly larger number of easy-to-use ML packages for modeling atomistic systems are becoming available. In this Perspective, we discuss the general aspects of ML techniques in the context of creating ML force fields. We describe common features of ML modeling and quantum-mechanical approximations, so-called global and local ML models, and the physical differences behind these two classes of approaches. Finally, we describe the recent developments and emerging directions in the field of ML-driven molecular modeling. This Perspective aims to inspire interdisciplinary collaborations crossing the borders between physical chemistry, chemical physics, computer science, and data science. read less NOT USED (low confidence) S. Ghosal, S. Chowdhury, and D. Jana, “Electronic and thermal transport in novel carbon-based bilayer with tetragonal rings: a combined study using first-principles and machine learning approach.,” Physical chemistry chemical physics : PCCP. 2021. link Times cited: 13 Abstract: In this article, the structural, electronic and thermal tran… read moreAbstract: In this article, the structural, electronic and thermal transport characteristics of bilayer tetragonal graphene (TG) are systematically explored with a combination of first-principles calculations and machine-learning interatomic potential approaches. Optimized ground state geometry of the bilayer TG structure is predicted and examined by employing various stability criteria. Electronic bandstructure analysis confirmed that bilayer TG exhibits a metallic band structure similar to the monolayer T-graphene structure. Thermal transport characteristics of the bilayer TG structure are explored by analysing thermal conductivity, the Seebeck coefficient, and electrical conductivity. The electronic part of the thermal conductivity shows linearly increasing behaviour with temperature, however the lattice part exhibits the opposite character. The lattice thermal conductivity part is investigated in terms of the three phonon scattering rates and weighted phase space. On the other hand, the Seebeck coefficient goes through a transition from negative to positive values with increasing temperature. The Wiedemann-Franz law regarding electrical transport of the bilayer TG is verified and confirms the universal Lorentz number. Specific heat of the bilayer TG structure follows the Debye model at low temperature and constant behaviour at high temperature. Moreover, the Debye temperature of the bilayer TG structure is verified by ab initio calculations as well as fitting the specific heat data using the Debye model. read less NOT USED (low confidence) M. Gilbert et al., “Perspectives on multiscale modelling and experiments to accelerate materials development for fusion,” Journal of Nuclear Materials. 2021. link Times cited: 33 NOT USED (low confidence) P. Gao, J. Zhang, H. Qiu, and S. Zhao, “A general QSPR protocol for the prediction of atomic/inter-atomic properties: a fragment based graph convolutional neural network (F-GCN).,” Physical chemistry chemical physics : PCCP. 2021. link Times cited: 6 Abstract: In this study, a general quantitative structure-property rel… read moreAbstract: In this study, a general quantitative structure-property relationship (QSPR) protocol, fragment based graph convolutional neural network (F-GCN), was developed for the prediction of atomic/inter-atomic properties. We applied this novel artificial intelligence (AI) tool in predictions of NMR chemical shifts and bond dissociation energies (BDEs). The obtained results were comparable to experimental measurements, while the computational cost was substantially reduced, with respect to pure density functional theory (DFT) calculations. The two important features of F-GCN can be summarised as: first, it could utilise different levels of molecular fragments for atomic/inter-atomic information extraction; second, the designed architecture is also open to include additional descriptors for a more accurate solution of the local environment at atomic level, making itself more efficient for structural solutions. And during our test, the averaged prediction error of 1H NMR chemical shifts is as small as 0.32 ppm, and the error of C-H BDE estimation is 2.7 kcal mol-1. Moreover, we further demonstrated the applicability of this developed F-GCN model via several challenging structural assignments. The success of the F-GCN in atomic and inter-atomic predictions also indicates an essential improvement of computational chemistry with the assistance of AI tools. read less NOT USED (low confidence) G. Pilania, “Machine learning in materials science: From explainable predictions to autonomous design,” Computational Materials Science. 2021. link Times cited: 85 NOT USED (low confidence) A. Krishnamoorthy et al., “Dielectric Constant of Liquid Water Determined with Neural Network Quantum Molecular Dynamics.,” Physical review letters. 2021. link Times cited: 11 Abstract: The static dielectric constant ϵ_{0} and its temperature dep… read moreAbstract: The static dielectric constant ϵ_{0} and its temperature dependence for liquid water is investigated using neural network quantum molecular dynamics (NNQMD). We compute the exact dielectric constant in canonical ensemble from NNQMD trajectories using fluctuations in macroscopic polarization computed from maximally localized Wannier functions (MLWF). Two deep neural networks are constructed. The first, NNQMD, is trained on QMD configurations for liquid water under a variety of temperature and density conditions to learn potential energy surface and forces and then perform molecular dynamics simulations. The second network, NNMLWF, is trained to predict locations of MLWF of individual molecules using the atomic configurations from NNQMD. Training data for both the neural networks is produced using a highly accurate quantum-mechanical method, DFT-SCAN that yields an excellent description of liquid water. We produce 280×10^{6} configurations of water at 7 temperatures using NNQMD and predict MLWF centers using NNMLWF to compute the polarization fluctuations. The length of trajectories needed for a converged value of the dielectric constant at 0°C is found to be 20 ns (40×10^{6} configurations with 0.5 fs time step). The computed dielectric constants for 0, 15, 30, 45, 60, 75, and 90°C are in good agreement with experiments. Our scalable scheme to compute dielectric constants with quantum accuracy is also applicable to other polar molecular liquids. read less NOT USED (low confidence) J. A. Suárez, E. Remesal, J. Plata, A. Márquez, and J. Sanz, “Computational Modeling of Carbon Dioxide Catalytic Conversion.” 2021. link Times cited: 1 NOT USED (low confidence) G. Guo, X. Yang, J. Carrete, and W. Li, “Revisiting the thermal conductivity of Si, Ge and diamond from first principles: roles of atomic mass and interatomic potential,” Journal of Physics: Condensed Matter. 2021. link Times cited: 2 Abstract: The thermal conductivity (κ) of nonmetals is determined by t… read moreAbstract: The thermal conductivity (κ) of nonmetals is determined by the constituent atoms, the crystal structure and interatomic potentials. Although the group-IV elemental solids Si, Ge and diamond have been studied extensively, a detailed understanding of the connection between the fundamental features of their energy landscapes and their thermal transport properties is still lacking. Here, starting from first principles, we analyze those factors, including the atomic mass (m) and the second- (harmonic) and third-order (anharmonic) interatomic force constants (IFCs). Both the second- and third-order IFCs of Si and Ge are very similar, and thus Si and Ge represent ideal systems to understand how the atomic mass alone affects κ. Although the group velocity (v) decreases with increasing atomic mass ( v−1∝m ), the phonon lifetime (τ) follows the opposite trend ( τ∝m ). Since the contribution to κ from each phonon mode is approximately proportional v 2 τ, κ is lower for the heavier element, namely Ge. Although the extremely high thermal conductivity of diamond is often attributed to weak anharmonic scattering, the anharmonic component of the interatomic potential is not much weaker than those of Si and Ge, which seems to be overlooked in the literature. In fact, the absolute magnitude of the third-order IFCs is much larger in diamond, and the ratios of the third-order IFCs with respect to the second-order ones are comparable to those of Si and Ge. We also explain the experimentally measured κ of high-quality diamonds (Inyushikin et al 2018 Phys. Rev. B 97 144305) by introducing boundary scattering into the picture, and obtain good agreement between calculations and measurements. read less NOT USED (low confidence) S. Dutta and K. Bose, “Remodelling structure-based drug design using machine learning.,” Emerging topics in life sciences. 2021. link Times cited: 2 Abstract: To keep up with the pace of rapid discoveries in biomedicine… read moreAbstract: To keep up with the pace of rapid discoveries in biomedicine, a plethora of research endeavors had been directed toward Rational Drug Development that slowly gave way to Structure-Based Drug Design (SBDD). In the past few decades, SBDD played a stupendous role in identification of novel drug-like molecules that are capable of altering the structures and/or functions of the target macromolecules involved in different disease pathways and networks. Unfortunately, post-delivery drug failures due to adverse drug interactions have constrained the use of SBDD in biomedical applications. However, recent technological advancements, along with parallel surge in clinical research have led to the concomitant establishment of other powerful computational techniques such as Artificial Intelligence (AI) and Machine Learning (ML). These leading-edge tools with the ability to successfully predict side-effects of a wide range of drugs have eventually taken over the field of drug design. ML, a subset of AI, is a robust computational tool that is capable of data analysis and analytical model building with minimal human intervention. It is based on powerful algorithms that use huge sets of 'training data' as inputs to predict new output values, which improve iteratively through experience. In this review, along with a brief discussion on the evolution of the drug discovery process, we have focused on the methodologies pertaining to the technological advancements of machine learning. This review, with specific examples, also emphasises the tremendous contributions of ML in the field of biomedicine, while exploring possibilities for future developments. read less NOT USED (low confidence) K. Fujioka, Y. Luo, and R. Sun, “Active Machine Learning for Chemical Dynamics Simulations. I. Estimating the Energy Gradient,” ChemRxiv. 2021. link Times cited: 0 Abstract: Ab initio molecular dymamics (AIMD) simulation studies are a… read moreAbstract: Ab initio molecular dymamics (AIMD) simulation studies are a direct
way to visualize chemical reactions and help elucidate non-statistical dynamics that does not follow the intrinsic reaction coordinate. However,
due to the enormous amount of the ab initio energy gradient calculations
needed for AIMD, it has been largely restrained to limited sampling and
low level of theory (i.e., density functional theory with small basis sets).
To overcome this issue, a number of machine learning (ML) methods have
been employed to predict the energy gradient of the system of interest.
In this manuscript, we outline the theoretical foundations of a novel ML
method which trains from a varying set of atomic positions and their
energy gradients, called interpolating moving ridge regression (IMRR),
and directly predicts the energy gradient of a new set of atomic positions.
Several key theoretical findings are presented regarding the inputs used to
train IMRR and the predicted energy gradient. A hyperparameter used to
guide IMRR is rigorously examined as well. The method is then applied to
three bimolecular reactions studied with AIMD, including HBr+ + CO2,
H2S + CH, and C4H2 + CH, to demonstrate IMRR’s performance on different chemical systems of different sizes. This manuscript also compares
the computational cost of the energy gradient calculation with IMRR vs.
ab initio, and the results highlight IMRR as a viable option to greatly
increase the efficiency of AIMD. read less NOT USED (low confidence) M. Xu, T. Zhu, and J. Z. H. Zhang, “Automated Construction of Neural Network Potential Energy Surface: The Enhanced Self-Organizing Incremental Neural Network Deep Potential Method,” Journal of chemical information and modeling. 2021. link Times cited: 4 Abstract: In recent years, the use of deep learning (neural network) p… read moreAbstract: In recent years, the use of deep learning (neural network) potential energy surface (NNPES) in molecular dynamics simulation has experienced explosive growth as it can be as accurate as quantum chemistry methods while being as efficient as classical mechanic methods. However, the development of NNPES is highly nontrivial. In particular, it has been troubling to construct a dataset that is as small as possible yet can cover the target chemical space. In this work, an ESOINN-DP method is developed, which has the enhanced self-organizing incremental neural network (ESOINN) and a newly proposed error indicator at its core. With ESOINN-DP, one can construct the NNPES with little human intervention, and this method ensures that the constructed reference dataset covers the target chemical space with minimum redundancy. The performance of the ESOINN-DP method has been well validated by developing neural network potential energy surfaces for water clusters, tripeptides, and by de-redundancy of a sub-dataset of the ANI-1 database. We believe that the ESOINN-DP method provides a novel idea for the construction of NNPES and, especially, the reference datasets, and it can be used for molecular dynamics (MD) simulations of various gas-phase and condensed-phase chemical systems. read less NOT USED (low confidence) Y. Choi et al., “CHARMM-GUI Polymer Builder for Modeling and Simulation of Synthetic Polymers.,” Journal of chemical theory and computation. 2021. link Times cited: 48 Abstract: Molecular modeling and simulations are invaluable tools for … read moreAbstract: Molecular modeling and simulations are invaluable tools for polymer science and engineering, which predict physicochemical properties of polymers and provide molecular-level insight into the underlying mechanisms. However, building realistic polymer systems is challenging and requires considerable experience because of great variations in structures as well as length and time scales. This work describes Polymer Builder in CHARMM-GUI (http://www.charmm-gui.org/input/polymer), a web-based infrastructure that provides a generalized and automated process to build a relaxed polymer system. Polymer Builder not only provides versatile modeling methods to build complex polymer structures, but also generates realistic polymer melt and solution systems through the built-in coarse-grained model and all-atom replacement. The coarse-grained model parametrization is generalized and extensively validated with various experimental data and all-atom simulations. In addition, the capability of Polymer Builder for generating relaxed polymer systems is demonstrated by density calculations of 34 homopolymer melt systems, characteristic ratio calculations of 170 homopolymer melt systems, a morphology diagram of poly(styrene-b-methyl methacrylate) block copolymers, and self-assembly behavior of amphiphilic poly(ethylene oxide-b-ethylethane) block copolymers in water. We hope that Polymer Builder is useful to carry out innovative and novel polymer modeling and simulation research to acquire insight into structures, dynamics, and underlying mechanisms of complex polymer-containing systems. read less NOT USED (low confidence) B. Mortazavi, F. Shojaei, X. Zhuang, and L. Pereira, “First-principles investigation of electronic, optical, mechanical and heat transport properties of pentadiamond: A comparison with diamond,” Carbon Trends. 2021. link Times cited: 13 NOT USED (low confidence) Y. Liu, O. C. Esan, Z. Pan, and L. An, “Machine learning for advanced energy materials.” 2021. link Times cited: 72 NOT USED (low confidence) J. Westermayr, M. Gastegger, K. T. Schütt, and R. Maurer, “Perspective on integrating machine learning into computational chemistry and materials science.,” The Journal of chemical physics. 2021. link Times cited: 72 Abstract: Machine learning (ML) methods are being used in almost every… read moreAbstract: Machine learning (ML) methods are being used in almost every conceivable area of electronic structure theory and molecular simulation. In particular, ML has become firmly established in the construction of high-dimensional interatomic potentials. Not a day goes by without another proof of principle being published on how ML methods can represent and predict quantum mechanical properties-be they observable, such as molecular polarizabilities, or not, such as atomic charges. As ML is becoming pervasive in electronic structure theory and molecular simulation, we provide an overview of how atomistic computational modeling is being transformed by the incorporation of ML approaches. From the perspective of the practitioner in the field, we assess how common workflows to predict structure, dynamics, and spectroscopy are affected by ML. Finally, we discuss how a tighter and lasting integration of ML methods with computational chemistry and materials science can be achieved and what it will mean for research practice, software development, and postgraduate training. read less NOT USED (low confidence) I. Syuhada, N. Hauwali, A. Rosikhin, E. Sustini, F. A. Noor, and T. Winata, “Bond order redefinition needed to reduce inherent noise in molecular dynamics simulations,” Scientific Reports. 2021. link Times cited: 1 NOT USED (low confidence) M. Nottoli, L. Cupellini, F. Lipparini, G. Granucci, and B. Mennucci, “Multiscale Models for Light-Driven Processes.,” Annual review of physical chemistry. 2021. link Times cited: 20 Abstract: Multiscale models combining quantum mechanical and classical… read moreAbstract: Multiscale models combining quantum mechanical and classical descriptions are a very popular strategy to simulate properties and processes of complex systems. Many alternative formulations have been developed, and they are now available in all of the most widely used quantum chemistry packages. Their application to the study of light-driven processes, however, is more recent, and some methodological and numerical problems have yet to be solved. This is especially the case for the polarizable formulation of these models, the recent advances in which we review here. Specifically, we identify and describe the most important specificities that the polarizable formulation introduces into both the simulation of excited-state dynamics and the modeling of excitation energy and electron transfer processes. Expected final online publication date for the Annual Review of Physical Chemistry, Volume 72 is April 2021. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates. read less NOT USED (low confidence) Z. Chen et al., “Machine Learning on Neutron and X-Ray Scattering.” 2021. link Times cited: 35 Abstract: Neutron and x-ray scattering represent two classes of state-… read moreAbstract: Neutron and x-ray scattering represent two classes of state-of-the-art materials characterization techniques that measure materials structural and dynamical properties with high precision. These techniques play critical roles in understanding a wide variety of materials systems from catalysts to polymers, nanomaterials to macromolecules, and energy materials to quantum materials. In recent years, neutron and x-ray scattering have received a significant boost due to the development and increased application of machine learning to materials problems. This article reviews the recent progress in applying machine learning techniques to augment various neutron and x-ray techniques, including neutron scattering, x-ray absorption, x-ray scattering, and photoemission. We highlight the integration of machine learning methods into the typical workflow of scattering experiments, focusing on problems that challenge traditional analysis approaches but are addressable through machine learning, including leveraging the knowledge of simple materials to model more complicated systems, learning with limited data or incomplete labels, identifying meaningful spectra and materials representations, mitigating spectral noise, and others. We present an outlook on a few emerging roles machine learning may play in broad types of scattering and spectroscopic problems in the foreseeable future. read less NOT USED (low confidence) A. Bafekry et al., “Two-dimensional carbon nitride C6N nanosheet with egg-comb-like structure and electronic properties of a semimetal,” Nanotechnology. 2020. link Times cited: 34 Abstract: In this study, the structural, electronic and optical proper… read moreAbstract: In this study, the structural, electronic and optical properties of theoretically predicted C6N monolayer structure are investigated by means of Density Functional Theory-based First-Principles Calculations. Phonon band dispersion calculations and molecular dynamics simulations reveal the dynamical and thermal stability of the C6N single-layer structure. We found out that the C6N monolayer has large negative in-plane Poisson’s ratios along both X and Y direction and the both values are almost four times that of the famous-pentagraphene. The electronic structure shows that C6N monolayer is a semi-metal and has a Dirac-point in the BZ. The optical analysis using the random phase approximation method constructed over HSE06 illustrates that the first peak of absorption coefficient of the C6N monolayer along all polarizations is located in the IR range of spectrum, while the second absorption peak occurs in the visible range, which suggests its potential applications in optical and electronic devices. Interestingly, optically anisotropic character of this system is highly desirable for the design of polarization-sensitive photodetectors. Thermoelectric properties such as Seebeck coefficient, electrical conductivity, electronic thermal conductivity and power factor are investigated as a function of carrier doping at temperatures 300, 400, and 500 K. In general, we predict that the C6N monolayer could be a new platform for study of novel physical properties in two-dimensional semi-metal materials, which may provide new opportunities to realize high-speed low-dissipation devices. read less NOT USED (low confidence) S. Watanabe et al., “High-dimensional neural network atomic potentials for examining energy materials: some recent simulations,” Journal of Physics: Energy. 2020. link Times cited: 16 Abstract: Owing to their simultaneous accuracy and computational effic… read moreAbstract: Owing to their simultaneous accuracy and computational efficiency, interatomic potentials machine-learned using first-principles calculation data are promising for investigating phenomena closely related to atomic motion in various energy materials. We have been working with one type of these potentials, high-dimensional (HD) neural network potentials (NNPs), and their applications, but we realized that our current understanding of HD NNPs, e.g. the meaning of the atomic energy mapping, remained insufficient, and that tuning their prediction performance for different target properties/phenomena often requires much trial and error. In this article, we illustrate the usefulness of NNPs through our studies on ion migration and thermal transport in energy and related materials. We also share our experiences with data sampling and training strategies and discuss the meaning of atomic energy mapping in HD NNPs. read less NOT USED (low confidence) A. H. Motagamwala and J. Dumesic, “Microkinetic Modeling: A Tool for Rational Catalyst Design.,” Chemical reviews. 2020. link Times cited: 135 Abstract: The design of heterogeneous catalysts relies on understandin… read moreAbstract: The design of heterogeneous catalysts relies on understanding the fundamental surface kinetics that controls catalyst performance, and microkinetic modeling is a tool that can help the researcher in streamlining the process of catalyst design. Microkinetic modeling is used to identify critical reaction intermediates and rate-determining elementary reactions, thereby providing vital information for designing an improved catalyst. In this review, we summarize general procedures for developing microkinetic models using reaction kinetics parameters obtained from experimental data, theoretical correlations, and quantum chemical calculations. We examine the methods required to ensure the thermodynamic consistency of the microkinetic model. We describe procedures required for parameter adjustments to account for the heterogeneity of the catalyst and the inherent errors in parameter estimation. We discuss the analysis of microkinetic models to determine the rate-determining reactions using the degree of rate control and reversibility of each elementary reaction. We introduce incorporation of Brønsted-Evans-Polanyi relations and scaling relations in microkinetic models and the effects of these relations on catalytic performance and formation of volcano curves are discussed. We review the analysis of reaction schemes in terms of the maximum rate of elementary reactions, and we outline a procedure to identify kinetically significant transition states and adsorbed intermediates. We explore the application of generalized rate expressions for the prediction of optimal binding energies of important surface intermediates and to estimate the extent of potential rate improvement. We also explore the application of microkinetic modeling in homogeneous catalysis, electro-catalysis, and transient reaction kinetics. We conclude by highlighting the challenges and opportunities in the application of microkinetic modeling for catalyst design. read less NOT USED (low confidence) R. Batra, L. Song, and R. Ramprasad, “Emerging materials intelligence ecosystems propelled by machine learning,” Nature Reviews Materials. 2020. link Times cited: 121 NOT USED (low confidence) P. Gao, J. Zhang, Y. Sun, and J. Yu, “Toward Accurate Predictions of Atomic Properties via Quantum Mechanics Descriptors Augmented Graph Convolutional Neural Network: Application of This Novel Approach in NMR Chemical Shifts Predictions.,” The journal of physical chemistry letters. 2020. link Times cited: 13 Abstract: In this study, an augmented Graph Convolutional Network (GCN… read moreAbstract: In this study, an augmented Graph Convolutional Network (GCN) with quantum mechanics (QM) descriptors was reported for its accurate predictions of NMR chemical shifts with respect to experimental values. The prediction errors of 13C/1H NMR chemical shifts can be as small as 2.14/0.11 ppm. There are two crucial characteristics for this modified GCN: in one aspect, such a novel neural network could efficiently extract the overall molecule structure information; in another aspect, it could accurately solve the chemical environment of the target atom. As there exists an imperfect linear regression between the experimental NMR chemical shifts (δ) and the density functional theory (DFT) calculated isotropic shielding constants (σ), the inclusion of QM descriptors within GCN can largely improve its performance. Moreover, few-shot learning also becomes feasible with these descriptors. The success of this novel GCN in chemical shifts predictions also indicates its potential applicability for other computational studies. read less NOT USED (low confidence) B. G. del Rio, C. Kuenneth, H. Tran, and R. Ramprasad, “An Efficient Deep Learning Scheme To Predict the Electronic Structure of Materials and Molecules: The Example of Graphene-Derived Allotropes.,” The journal of physical chemistry. A. 2020. link Times cited: 12 Abstract: Computations based on density functional theory (DFT) are tr… read moreAbstract: Computations based on density functional theory (DFT) are transforming various aspects of materials research and discovery. However, the effort required to solve the central equation of DFT, namely the Kohn-Sham equation, which remains a major obstacle for studying large systems with hundreds of atoms in a practical amount of time with routine computational resources. Here, we propose a deep learning architecture that systematically learns the input-output behavior of the Kohn-Sham equation and predicts the electronic density of states, a primary output of DFT calculations, with unprecedented speed and chemical accuracy. The algorithm also adapts and progressively improves in predictive power and versatility as it is exposed to new diverse atomic configurations. We demonstrate this capability for a diverse set of carbon allotropes spanning a large configurational and phase space. The electronic density of states, along with the electronic charge density, may be used downstream to predict a variety of materials properties, bypassing the Kohn-Sham equation, leading to an ultrafast and high-fidelity DFT emulator. read less NOT USED (low confidence) O. T. Unke et al., “Machine Learning Force Fields,” Chemical Reviews. 2020. link Times cited: 441 Abstract: In recent years, the use of machine learning (ML) in computa… read moreAbstract: In recent years, the use of machine learning (ML) in computational chemistry has enabled numerous advances previously out of reach due to the computational complexity of traditional electronic-structure methods. One of the most promising applications is the construction of ML-based force fields (FFs), with the aim to narrow the gap between the accuracy of ab initio methods and the efficiency of classical FFs. The key idea is to learn the statistical relation between chemical structure and potential energy without relying on a preconceived notion of fixed chemical bonds or knowledge about the relevant interactions. Such universal ML approximations are in principle only limited by the quality and quantity of the reference data used to train them. This review gives an overview of applications of ML-FFs and the chemical insights that can be obtained from them. The core concepts underlying ML-FFs are described in detail, and a step-by-step guide for constructing and testing them from scratch is given. The text concludes with a discussion of the challenges that remain to be overcome by the next generation of ML-FFs. read less NOT USED (low confidence) X. Chu et al., “The role of nuclear charges in unifying the descriptions of neural networks (NN)-based force fields,” Materials Letters. 2020. link Times cited: 0 NOT USED (low confidence) M. Kilgour, N. Gastellu, D. Y. Hui, Y. Bengio, and L. Simine, “Generating Multiscale Amorphous Molecular Structures Using Deep Learning: A Study in 2D.,” The journal of physical chemistry letters. 2020. link Times cited: 7 Abstract: Amorphous molecular assemblies appear in a vast array of sys… read moreAbstract: Amorphous molecular assemblies appear in a vast array of systems: from living cells to chemical plants and from everyday items to new devices. The absence of long-range order in amorphous materials implies that precise knowledge of their underlying structures throughout is needed to rationalize and control their properties at the mesoscale. Standard computational simulations suffer from exponentially unfavorable scaling of the required compute with system size. We present a method based on deep learning that leverages the finite range of structural correlations for an autoregressive generation of disordered molecular aggregates up to arbitrary size from small-scale computational or experimental samples. We benchmark performance on self-assembled nanoparticle aggregates and proceed to simulate monolayer amorphous carbon with atomistic resolution. This method bridges the gap between the nanoscale and mesoscale simulations of amorphous molecular systems. read less NOT USED (low confidence) A. S. Kelkar, B. C. Dallin, and R. V. V. Lehn, “Predicting Hydrophobicity by Learning Spatiotemporal Features of Interfacial Water Structure: Combining Molecular Dynamics Simulations with Convolutional Neural Networks.,” The journal of physical chemistry. B. 2020. link Times cited: 14 Abstract: The hydrophobicity of functionalized interfaces can be quant… read moreAbstract: The hydrophobicity of functionalized interfaces can be quantified by the structure and dynamics of water molecules using molecular dynamics (MD) simulations, but existing methods to quantify interfacial hydrophobicity are computationally expensive. In this work, we develop a new machine learning approach that leverages convolutional neural networks (CNNs) to predict the hydration free energy (HFE) as a measure of interfacial hydrophobicity based on water positions sampled from MD simulations. We construct a set of idealized self-assembled monolayers (SAMs) with varying surface polarities and calculate their HFEs using indirect umbrella sampling calculations (INDUS). Using the INDUS-calculated HFEs as labels and physically informed representations of interfacial water density from MD simulations as input, we train and evaluate a series of neural networks to predict SAM HFEs. By systematically varying model hyperparameters, we demonstrate that a 3D CNN trained to analyze both spatial and temporal correlations between interfacial water molecule positions leads to HFE predictions that require an order of magnitude less MD simulation time than INDUS. We showcase the power of this model to explore a large design space by predicting HFEs for a set of 71 chemically heterogeneous SAMs with varying patterns and mole fractions. read less NOT USED (low confidence) J. J. Eriksen, “Mean-field density matrix decompositions.,” The Journal of chemical physics. 2020. link Times cited: 11 Abstract: We introduce new and robust decompositions of mean-field Har… read moreAbstract: We introduce new and robust decompositions of mean-field Hartree-Fock and Kohn-Sham density functional theory relying on the use of localized molecular orbitals and physically sound charge population protocols. The new lossless property decompositions, which allow for partitioning one-electron reduced density matrices into either bond-wise or atomic contributions, are compared to alternatives from the literature with regard to both molecular energies and dipole moments. Besides commenting on possible applications as an interpretative tool in the rationalization of certain electronic phenomena, we demonstrate how decomposed mean-field theory makes it possible to expose and amplify compositional features in the context of machine-learned quantum chemistry. This is made possible by improving upon the granularity of the underlying data. On the basis of our preliminary proof-of-concept results, we conjecture that many of the structure-property inferences in existence today may be further refined by efficiently leveraging an increase in dataset complexity and richness. read less NOT USED (low confidence) N. Khatavkar, S. Swetlana, and A. Singh, “Accelerated prediction of Vickers hardness of Co- and Ni-based superalloys from microstructure and composition using advanced image processing techniques and machine learning,” Acta Materialia. 2020. link Times cited: 32 NOT USED (low confidence) W. Beckner, C. M. Ashraf, J. Lee, D. A. C. Beck, and J. Pfaendtner, “Continuous Molecular Representations of Ionic Liquids.,” The journal of physical chemistry. B. 2020. link Times cited: 6 Abstract: Designing new ionic liquids (ILs) is of crucial importance f… read moreAbstract: Designing new ionic liquids (ILs) is of crucial importance for various industrial applications. However, this always leads to a daunting challenge as the number of possible combinations of cation and anion are very high and it is impossible to experimentally propose and screen a wide pool of potential candidates. Yet, recent applications of machine learning (ML) models have greatly improved the overall chemical discovery pipeline. In this study, we compare different generative methods for producing ionic liquids. In this comparison we show that - 1) when training data is scarce, a transfer learning approach can be applied to variational autoencoders (VAEs) to generate molecular structures of the target molecule type; 2) in a VAE-like structure, separate latent spaces for the cationic and anionic moieties can result in meaningful representations for their combinative, macroscopic properties; 3) interpolating between ILs with desired properties can result in a new IL with attributes similar to the two structural endpoints. read less NOT USED (low confidence) V. L. Deringer, “Modelling and understanding battery materials with machine-learning-driven atomistic simulations,” Journal of Physics: Energy. 2020. link Times cited: 49 Abstract: The realistic computer modelling of battery materials is an … read moreAbstract: The realistic computer modelling of battery materials is an important research goal, with open questions ranging from atomic-scale structure and dynamics to macroscopic phenomena. Quantum-mechanical methods offer high accuracy and predictive power in small-scale atomistic simulations, but they quickly reach their limits when complex electrochemical systems are to be studied—for example, when structural disorder or even fully amorphous phases are present, or when reactions take place at the interface between electrodes and electrolytes. In this Perspective, it is argued that emerging machine learning based interatomic potentials are promising tools for studying battery materials on the atomistic and nanometre length scales, affording quantum-mechanical accuracy yet being many orders of magnitude faster, and thereby extending the capabilities of current battery modelling methodology. Initial applications to solid-state electrolyte and anode materials in lithium-ion batteries are highlighted, and future directions and possible synergies with experiments are discussed. read less NOT USED (low confidence) J. Westermayr and P. Marquetand, “Machine Learning for Electronically Excited States of Molecules,” Chemical Reviews. 2020. link Times cited: 166 Abstract: Electronically excited states of molecules are at the heart … read moreAbstract: Electronically excited states of molecules are at the heart of photochemistry, photophysics, as well as photobiology and also play a role in material science. Their theoretical description requires highly accurate quantum chemical calculations, which are computationally expensive. In this review, we focus on not only how machine learning is employed to speed up such excited-state simulations but also how this branch of artificial intelligence can be used to advance this exciting research field in all its aspects. Discussed applications of machine learning for excited states include excited-state dynamics simulations, static calculations of absorption spectra, as well as many others. In order to put these studies into context, we discuss the promises and pitfalls of the involved machine learning techniques. Since the latter are mostly based on quantum chemistry calculations, we also provide a short introduction into excited-state electronic structure methods and approaches for nonadiabatic dynamics simulations and describe tricks and problems when using them in machine learning for excited states of molecules. read less NOT USED (low confidence) R. Han and S. Luber, “Trajectory-based machine learning method and its application to molecular dynamics,” Molecular Physics. 2020. link Times cited: 3 Abstract: Ab initio molecular dynamics (AIMD) has become a popular sim… read moreAbstract: Ab initio molecular dynamics (AIMD) has become a popular simulation technique but long simulation times are often hampered due to its high computational effort. Alternatively, classical molecular dynamics (MD) based on force fields may be used, which, however, has certain shortcomings compared to AIMD. In order to alleviate that situation, a trajectory-based machine learning (TrajML) approach is introduced for the construction of force fields by learning from AIMD trajectories. Only nuclear trajectories are required, which can be obtained by other methods beyond AIMD as well. We developed an easy-to-use MD machine learning package (TrajML MD) for instant modelling of the force field and system-focussed prediction of molecular configurations for MD trajectories. It consumes similar computational resources as classical MD but can simulate complex systems with a higher accuracy due to the targeted learning on the system of interest. GRAPHICAL ABSTRACT read less NOT USED (low confidence) P. Gao, J. Zhang, Q. Peng, J. Zhang, and V. Glezakou, “General Protocol for the Accurate Prediction of Molecular 13C/1H NMR Chemical Shifts via Machine Learning Augmented DFT,” Journal of chemical information and modeling. 2020. link Times cited: 36 Abstract: Accurate prediction of NMR chemical shifts at affordable com… read moreAbstract: Accurate prediction of NMR chemical shifts at affordable computational cost is very important for different types of structural assignments in experimental studies. Density functional theory (DFT) and gauge-including atomic orbital (GIAO) are two of the most popular computational methods for NMR calculation, yet, they often fail to resolve ambiguities in structural assignments. Here, we present a new method that uses machine learning (ML) techniques (DFT+ML that significantly increases the accuracy of 13C/1H NMR chemical shift prediction for a variety of organic molecules. The input of the generalizable DFT+ML model contains two critical parts: one is a vector providing insights into chemical environments, which can be evaluated without knowing the exact geometry of the molecule; the other one is the DFT calculated isotropic shielding constant. The DFT+ML model was trained with a dataset containing 476 13C and 270 1H experimental chemical shifts. For the DFT methods used here, the root mean square deviations (RMSDs) for the errors between predicted and experimental 13C/1H chemical shifts can be as small as 2.10/0.18 ppm, which is much lower than those from simple DFT (5.54/0.25 ppm), or DFT+linear regression (LR) (4.77/0.23 ppm) approaches. It also has a smaller maximum absolute error than two previously proposed NMR-predicting ML models. The robustness of the DFT+ML model is tested on two classes of organic molecules (TIC10 and hyacinthacines), where the correct isomers were unambiguously assigned to the experimental ones. Overall, the DFT+ML model is showing promise for structural assignments in a variety of systems, including stereoisomers, that are often challenging to determine experimentally. read less NOT USED (low confidence) P. O. Dral, A. Owens, A. Dral, and G. Csányi, “Hierarchical machine learning of potential energy surfaces.,” The Journal of chemical physics. 2020. link Times cited: 49 Abstract: We present hierarchical machine learning (hML) of highly acc… read moreAbstract: We present hierarchical machine learning (hML) of highly accurate potential energy surfaces (PESs). Our scheme is based on adding predictions of multiple Δ-machine learning models trained on energies and energy corrections calculated with a hierarchy of quantum chemical methods. Our (semi-)automatic procedure determines the optimal training set size and composition of each constituent machine learning model, simultaneously minimizing the computational effort necessary to achieve the required accuracy of the hML PES. Machine learning models are built using kernel ridge regression, and training points are selected with structure-based sampling. As an illustrative example, hML is applied to a high-level ab initio CH3Cl PES and is shown to significantly reduce the computational cost of generating the PES by a factor of 100 while retaining similar levels of accuracy (errors of ∼1 cm-1). read less NOT USED (low confidence) J. Westermayr and P. Marquetand, “Machine learning and excited-state molecular dynamics,” Machine Learning: Science and Technology. 2020. link Times cited: 40 Abstract: Machine learning is employed at an increasing rate in the re… read moreAbstract: Machine learning is employed at an increasing rate in the research field of quantum chemistry. While the majority of approaches target the investigation of chemical systems in their electronic ground state, the inclusion of light into the processes leads to electronically excited states and gives rise to several new challenges. Here, we survey recent advances for excited-state dynamics based on machine learning. In doing so, we highlight successes, pitfalls, challenges and future avenues for machine learning approaches for light-induced molecular processes. read less NOT USED (low confidence) S. Chhabra, J. Xie, and A. Frank, “RNAPosers: Machine Learning Classifiers for Ribonucleic Acid-Ligand Poses.,” The journal of physical chemistry. B. 2020. link Times cited: 21 Abstract: Determining the three-dimensional (3D) structures of ribonuc… read moreAbstract: Determining the three-dimensional (3D) structures of ribonucleic acid (RNA)-small molecule ligand complexes is critical to understanding molecular recognition in RNA. Computer docking can, in principle, be used to predict the 3D structure of RNA-small molecule complexes. Unfortunately, retrospective analysis has shown that the scoring functions that are typically used for pose prediction tend to misclassify non-native poses as native and vice versa. Here, we use machine learning to train a set of pose classifiers that estimate the relative "nativeness" of a set of RNA-ligand poses. At the heart of our approach is the use of a pose "fingerprint" (FP) that is a composite of a set of atomic FPs, which individually encode the local "RNA environment" around ligand atoms. We found that by ranking poses based on classification scores from our machine learning classifiers, we were able to recover native-like poses better than when we ranked poses based on their docking scores. With a leave-one-out training and testing approach, we found that one of our classifiers could recover poses that were within 2.5 Å of the native poses in ∼80% of the 80 cases we examined, and, on two separate validation sets, we could recover such poses in ∼60% of the cases. Our set of classifiers, which we refer to as RNAPosers, should find utility as a tool to aid in RNA-ligand pose prediction, and so we make RNAPosers open to the academic community via https://github.com/atfrank/RNAPosers. read less NOT USED (low confidence) J. Roel-Touris and A. Bonvin, “Coarse-grained (hybrid) integrative modeling of biomolecular interactions,” Computational and Structural Biotechnology Journal. 2020. link Times cited: 15 NOT USED (low confidence) X. Gao, F. Ramezanghorbani, O. Isayev, J. S. Smith, and A. Roitberg, “TorchANI: A Free and Open Source PyTorch-Based Deep Learning Implementation of the ANI Neural Network Potentials,” Journal of chemical information and modeling. 2020. link Times cited: 127 Abstract: This paper presents TorchANI, a PyTorch based software for t… read moreAbstract: This paper presents TorchANI, a PyTorch based software for training/inference of ANI (ANAKIN-ME) deep learning models to obtain potential energy surfaces and other physical properties of molecular systems. ANI is an accurate neural network potential originally implemented using C++/CUDA in a program called NeuroChem. Compared with NeuroChem, TorchANI has a design emphasis on being light weight, user friendly, cross platform, and easy to read and modify for fast prototyping, while allowing acceptable sacrice on running performance. Because the computation of atomic environmental vectors (AEVs) and atomic neural networks are all implemented using PyTorch operators, TorchANI is able to use PyTorch's autograd engine to automatically compute analytical forces and Hessian matrices, as well as do force training without additional codes required. TorchANI is open-source and freely available on GitHub: https://github.com/aiqm/torchani. read less NOT USED (low confidence) T. Loeffler, T. Patra, H. Chan, and S. Sankaranarayanan, “Active learning a coarse-grained neural network model for bulk water from sparse training data.” 2020. link Times cited: 5 Abstract: Neural network (NN) based potentials represent flexible alte… read moreAbstract: Neural network (NN) based potentials represent flexible alternatives to pre-defined functional forms. Well-trained NN potentials are transferable and provide a high level of accuracy on-par with the reference model used for training. Despite their tremendous potential and interest in them, there are at least two challenges that need to be addressed – (1) NN models are interpolative, and hence trained by generating large quantities (∼104 or greater) of structural data in hopes that the model has adequately sampled the energy landscape both near and far-from-equilibrium. It is desirable to minimize the number of training data, especially if the underlying reference model is expensive. (2) NN atomistic potentials (like any other classical atomistic model) are limited in the time scales they can access. Coarse-grained NN potentials have emerged as a viable alternative. Here, we address these challenges by introducing an active learning scheme that trains a CG model with a minimal amount of training data. Our active learning workflow starts with a sparse training data set (∼1 to 5 data points), which is continually updated via a nested ensemble Monte Carlo scheme that iteratively queries the energy landscape in regions of failure and improves the network performance. We demonstrate that with ∼300 reference data, our AL-NN is able to accurately predict both the energies and the molecular forces of water, within 2 meV per molecule and 40 meV A−1 of the reference (coarse-grained bond-order potential) model. The AL-NN water model provides good prediction of several structural, thermodynamic, and temperature dependent properties of liquid water, with values close to those obtained from the reference model. The AL-NN also captures the well-known density anomaly of liquid water observed in experiments. Although the AL procedure has been demonstrated for training CG models with sparse reference data, it can be easily extended to develop atomistic NN models against a minimal amount of high-fidelity first-principles data. read less NOT USED (low confidence) D. Kamal, A. Chandrasekaran, R. Batra, and R. Ramprasad, “A charge density prediction model for hydrocarbons using deep neural networks,” Machine Learning: Science and Technology. 2020. link Times cited: 17 Abstract: The electronic charge density distribution ρ(r) of a given m… read moreAbstract: The electronic charge density distribution ρ(r) of a given material is among the most fundamental quantities in quantum simulations from which many large scale properties and observables can be calculated. Conventionally, ρ(r) is obtained using Kohn–Sham density functional theory (KS-DFT) based methods. But, the high computational cost of KS-DFT renders it intractable for systems involving thousands/millions of atoms. Thus, recently there has been efforts to bypass expensive KS equations, and directly predict ρ(r) using machine learning (ML) based methods. Here, we build upon one such scheme to create a robust and reliable ρ(r) prediction model for a diverse set of hydrocarbons, involving huge chemical and morphological complexity /(saturated, unsaturated molecules, cyclo-groups and amorphous and semi-crystalline polymers). We utilize a grid-based fingerprint to capture the atomic neighborhood around an arbitrary point in space, and map it to the reference ρ(r) obtained from standard DFT calculations at that point. Owing to the grid-based learning, dataset sizes exceed billions of points, which is trained using deep neural networks in conjunction with a incremental learning based approach. The accuracy and transferability of the ML approach is demonstrated on not only a diverse test set, but also on a completely unseen system of polystyrene under different strains. Finally, we note that the general approach adopted here could be easily extended to other material systems, and can be used for quick and accurate determination of ρ(r) for DFT charge density initialization, computing dipole or quadrupole, and other observables for which reliable density functional are known. read less NOT USED (low confidence) G. Pilania, P. Balachandran, J. Gubernatis, and T. Lookman, “Data-Based Methods for Materials Design and Discovery: Basic Ideas and General Methods.” 2020. link Times cited: 11 Abstract: Machine learning methods are changing the way we design and … read moreAbstract: Machine learning methods are changing the way we design and discover new materials. This book provides an overview of approaches successfully used in addressing materials problems (alloys,... read less NOT USED (low confidence) P. O. Dral, “Quantum Chemistry in the Age of Machine Learning.,” The journal of physical chemistry letters. 2020. link Times cited: 218 Abstract: As quantum chemistry (QC) community embraces machine learnin… read moreAbstract: As quantum chemistry (QC) community embraces machine learning (ML), the surging number of new methods and applications based on combination of QC and ML is emerging. In this Perspective a view on the current state of affairs in this new exciting research field is offered, challenges of using ML in QC applications are described, and potential future developments are outlined. Specifically, examples of how ML is used to improve the accuracy and accelerate QC research are shown. Generalization and classification of existing techniques is provided to ease the navigation in the sea of literature and to guide the researchers entering the field. The emphasis of this Perspective is on the supervised ML. read less NOT USED (low confidence) M. E. Khatib and W. A. Jong, “ML4Chem: A Machine Learning Package for Chemistry and Materials Science,” ArXiv. 2020. link Times cited: 3 Abstract: ML4Chem is an open-source machine learning library for chemi… read moreAbstract: ML4Chem is an open-source machine learning library for chemistry and materials science. It
provides an extendable platform to develop and deploy machine learning models and pipelines and
is targeted to the non-expert and expert users. ML4Chem follows user-experience design and offers
the needed tools to go from data preparation to inference. Here we introduce its atomistic module
for the implementation, deployment, and reproducibility of atom-centered models. This module is
composed of six core building blocks: data, featurization, models, model optimization, inference,
and visualization. We present their functionality and ease of use with demonstrations utilizing
neural networks and kernel ridge regression algorithms. read less NOT USED (low confidence) N. V. Orupattur, S. H. Mushrif, and V. Prasad, “Catalytic materials and chemistry development using a synergistic combination of machine learning and ab initio methods,” Computational Materials Science. 2020. link Times cited: 18 NOT USED (low confidence) J. Chapman, R. Batra, and R. Ramprasad, “Machine learning models for the prediction of energy, forces, and stresses for Platinum,” Computational Materials Science. 2020. link Times cited: 18 NOT USED (low confidence) A. Chen, X. Zhang, and Z. Zhou, “Machine learning: Accelerating materials development for energy storage and conversion,” InfoMat. 2020. link Times cited: 161 Abstract: With the development of modern society, the requirement for … read moreAbstract: With the development of modern society, the requirement for energy has become increasingly important on a global scale. Therefore, the exploration of novel materials for renewable energy technologies is urgently needed. Traditional methods are difficult to meet the requirements for materials science due to long experimental period and high cost. Nowadays, machine learning (ML) is rising as a new research paradigm to revolutionize materials discovery. In this review, we briefly introduce the basic procedure of ML and common algorithms in materials science, and particularly focus on latest progress in applying ML to property prediction and materials development for energy-related fields, including catalysis, batteries, solar cells, and gas capture. More-over, contributions of ML to experiments are involved as well. We highly expect that this review could lead the way forward in the future development of ML in materials science. read less NOT USED (low confidence) R. Jinnouchi, F. Karsai, and G. Kresse, “Making free-energy calculations routine: Combining first principles with machine learning,” Physical Review B. 2020. link Times cited: 15 NOT USED (low confidence) J. Westermayr, M. Gastegger, and P. Marquetand, “Combining SchNet and SHARC: The SchNarc Machine Learning Approach for Excited-State Dynamics,” The Journal of Physical Chemistry Letters. 2020. link Times cited: 99 Abstract: In recent years, deep learning has become a part of our ever… read moreAbstract: In recent years, deep learning has become a part of our everyday life and is revolutionizing quantum chemistry as well. In this work, we show how deep learning can be used to advance the research field of photochemistry by learning all important properties—multiple energies, forces, and different couplings—for photodynamics simulations. We simplify such simulations substantially by (i) a phase-free training skipping costly preprocessing of raw quantum chemistry data; (ii) rotationally covariant nonadiabatic couplings, which can either be trained or (iii) alternatively be approximated from only ML potentials, their gradients, and Hessians; and (iv) incorporating spin–orbit couplings. As the deep-learning method, we employ SchNet with its automatically determined representation of molecular structures and extend it for multiple electronic states. In combination with the molecular dynamics program SHARC, our approach termed SchNarc is tested on two polyatomic molecules and paves the way toward efficient photodynamics simulations of complex systems. read less NOT USED (low confidence) C. Devereux et al., “Extending the applicability of the ANI deep learning molecular potential to Sulfur and Halogens.,” Journal of chemical theory and computation. 2020. link Times cited: 141 Abstract: Machine learning (ML) methods have become powerful, predicti… read moreAbstract: Machine learning (ML) methods have become powerful, predictive tools in a wide range of applications, such as facial recognition and autonomous vehicles. In the sciences, computational chemists and physicists have been using ML for the prediction of physical phenomena, such as atomistic potential energy surfaces and reaction pathways. Transferable ML potentials, such as ANI-1x, have been developed with the goal of accurately simulating organic molecules containing the chemical elements H, C, N, and O. Here we provide an extension of the ANI-1x model. The new model, dubbed ANI-2x, is trained to three additional chemical elements: S, F, and Cl. Additionally, ANI-2x underwent torsional refinement training to better predict molecular torsion profiles. These new features open a wide range of new applications within organic chemistry and drug development. These seven elements (H, C, N, O, F, Cl, S) make up ~90% of drug like molecules. To show that these additions do not sacrifice accuracy, we have tested this model across a range of organic molecules and applications, including the COMP6 benchmark, dihedral rotations, conformer scoring, and non-bonded interactions. ANI-2x is shown to accurately predict molecular energies compared to DFT with a ~106 factor speedup and a negligible slowdown compared to ANI-1x. The resulting model is a valuable tool for drug development that can potentially replace both quantum calculations and classical force fields for myriad applications. read less NOT USED (low confidence) F. Guo et al., “Intelligent-ReaxFF: Evaluating the reactive force field parameters with machine learning,” Computational Materials Science. 2020. link Times cited: 29 NOT USED (low confidence) Q. Hu, M. Weng, X. Chen, S. Li, F. Pan, and L.-wang Wang, “Neural Network Force Fields For Metal Growth Based On Energy Decompositions.,” The journal of physical chemistry letters. 2020. link Times cited: 6 Abstract: A method using machine learning (ML) is proposed to describe… read moreAbstract: A method using machine learning (ML) is proposed to describe metal growth for simulations, which retains the accuracy of ab initio density functional theory (DFT) and reduces the computational time by thousands of folds. This method is based on atomic energy decomposition from DFT calculations. Compared with other ML methods, our energy decomposition approach can yield much more information with the same DFT calculations. This approach is employed for the amorphous sodium system, where only 1000 DFT molecular dynamics images are enough for training an accurate model. The DFT and NNP (neural network potential) are compared for the dynamics to show that similar structural properties are generated. Finally, metal growths from liquid to solid in a small and larger system are carried out to demonstrate the ability of using NNP to simulate the real growth process. read less NOT USED (low confidence) N. Jackson, A. Bowen, and J. D. de Pablo, “Efficient Multiscale Optoelectronic Prediction for Conjugated Polymers,” Macromolecules. 2020. link Times cited: 17 Abstract: Conjugated polymers represent a high-potential material clas… read moreAbstract: Conjugated polymers represent a high-potential material class due to the tunability of their optoelectronic properties via straightforward processing protocols. However, to in silico tailor these optoelectronic properties, one must compute the conformationally dependent electronic structure over mesoscopic length and time scales. This task represents a challenging multiscale computational problem in which quantum chemistry and atomistic or coarse-grained molecular dynamics must be integrated. Recently, we introduced a methodology, called electronic coarse graining (ECG), that utilizes artificial neural networks to compute ab initio quality electronic structures using only the reduced degrees of freedom of coarse-grained models. Here, we adapt ECG to variable-molecular weight conjugated oligomers by casting the problem in the framework of sequence detection. A machine-learning (ML) architecture employing one-dimensional convolutional neural networks and long short-term memory networks is utilized to comput... read less NOT USED (low confidence) J. Wang, S. Shin, and S. Lee, “Interatomic Potential Model Development: Finite‐Temperature Dynamics Machine Learning,” Advanced Theory and Simulations. 2019. link Times cited: 2 Abstract: Developing an accurate interatomic potential model is a prer… read moreAbstract: Developing an accurate interatomic potential model is a prerequisite for achieving reliable results from classical molecular dynamics (CMD) simulations; however, most of the potentials are biased as specific simulation purposes or conditions are considered in the parameterization. For developing an unbiased potential, a finite‐temperature dynamics machine learning (FTD‐ML) approach is proposed, and its processes and feasibility are demonstrated using the Buckingham potential model and aluminum (Al) as an example. Compared with conventional machine learning approaches, FTD‐ML exhibits three distinguished features: 1) FTD‐ML intrinsically incorporates more extensive configurational and conditional space for enhancing the transferability of developed potentials; 2) FTD‐ML employs various properties calculated directly from CMD, for ML model training and prediction validation against experimental data instead of first‐principles data; 3) FTD‐ML is much more computationally cost effective than first‐principles simulations, especially when the system size increases over 103 atoms as employed in this research for ensuring reliable training data. The Al Buckingham potential developed by the FTD‐ML approach exhibits good performance for general simulation purposes. Thus, the FTD‐ML approach is expected to contribute to a fast development of interatomic potential model suitable for various simulation purposes and conditions, without limitation of model type, while maintaining experimental‐level accuracy. read less NOT USED (low confidence) J. J. Varghese, “Computational design of catalysts for bio-waste upgrading,” Current opinion in chemical engineering. 2019. link Times cited: 5 NOT USED (low confidence) M. Jafary-Zadeh, K. Khoo, R. Laskowski, P. S. Branicio, and A. Shapeev, “Applying a machine learning interatomic potential to unravel the effects of local lattice distortion on the elastic properties of multi-principal element alloys,” Journal of Alloys and Compounds. 2019. link Times cited: 39 NOT USED (low confidence) D. Dubbeldam, K. S. Walton, T. Vlugt, and S. Calero, “Design, Parameterization, and Implementation of Atomic Force Fields for Adsorption in Nanoporous Materials,” Advanced Theory and Simulations. 2019. link Times cited: 42 Abstract: Molecular simulations are an excellent tool to study adsorpt… read moreAbstract: Molecular simulations are an excellent tool to study adsorption and diffusion in nanoporous materials. Examples of nanoporous materials are zeolites, carbon nanotubes, clays, metal‐organic frameworks (MOFs), covalent organic frameworks (COFs) and zeolitic imidazolate frameworks (ZIFs). The molecular confinement these materials offer has been exploited in adsorption and catalysis for almost 50 years. Molecular simulations have provided understanding of the underlying shape selectivity, and adsorption and diffusion effects. Much of the reliability of the modeling predictions depends on the accuracy and transferability of the force field. However, flexibility and the chemical and structural diversity of MOFs add significant challenges for engineering force fields that are able to reproduce experimentally observed structural and dynamic properties. Recent developments in design, parameterization, and implementation of force fields for MOFs and zeolites are reviewed. read less NOT USED (low confidence) V. L. Deringer, M. A. Caro, and G. Csányi, “Machine Learning Interatomic Potentials as Emerging Tools for Materials Science,” Advanced Materials. 2019. link Times cited: 245 Abstract: Atomic‐scale modeling and understanding of materials have ma… read moreAbstract: Atomic‐scale modeling and understanding of materials have made remarkable progress, but they are still fundamentally limited by the large computational cost of explicit electronic‐structure methods such as density‐functional theory. This Progress Report shows how machine learning (ML) is currently enabling a new degree of realism in materials modeling: by “learning” electronic‐structure data, ML‐based interatomic potentials give access to atomistic simulations that reach similar accuracy levels but are orders of magnitude faster. A brief introduction to the new tools is given, and then, applications to some select problems in materials science are highlighted: phase‐change materials for memory devices; nanoparticle catalysts; and carbon‐based electrodes for chemical sensing, supercapacitors, and batteries. It is hoped that the present work will inspire the development and wider use of ML‐based interatomic potentials in diverse areas of materials research. read less NOT USED (low confidence) G. Sun and P. Sautet, “Towards fast and reliable potential energy surfaces for metallic Pt clusters by hierarchical delta neural networks.,” Journal of chemical theory and computation. 2019. link Times cited: 28 Abstract: Data-driven machine learning force fields (MLF) are more and… read moreAbstract: Data-driven machine learning force fields (MLF) are more and more popular in atomistic simulations, and exploit machine learning methods to predict energies and forces for unknown structures based on the knowledge learned from an existing reference database. The latter usually comes from density functional theory calculations. One main drawback of MLFs is that physical laws are not incorporated in the machine learning models and instead, MLFs are designed to be very flexible to simulate complex quantum chemistry potential energy surface (PES). In general, MLFs have poor transferability, and hence a very large trainset is required to span all the target feature space to get a reliable MLF. This procedure becomes more troublesome when the PES is complicated, with a large number of degrees of freedom, in which building a large database is inevitable and very expensive, especially when accurate but costly exchange-correlation functionals have to be used. In this manuscript, we exploit a high dimensional neural network potential (HDNNP) on Pt clusters of size 6 to 20 as one example. Our standard level of energy calculation is DFT GGA (PBE) using a plane wave basis set. We introduce an approximate but fast level with the PBE functional and a minimal atomic orbital basis set, then a more accurate but expensive level, using a hybrid functional or non-local vdw functional and a plane wave basis set, is reliably predicted by learning the difference with HDNNP. The results show that such a differential approach (named ΔHDNNP) can deliver very accurate predictions (error < 10 meV/atom) in reference to converged basis set energies as well as more accurate but expensive xc functional. The overall speedup can be as large as 900 for 20 atom Pt cluster. More importantly, ΔHDNNP shows much better transferability due to the intrinsic smoothness of delta potential energy surface, and accordingly one can use much smaller trainset data to obtain better accuracy than the conventional HDNNP. A multi-layer ΔHDNNP is thus proposed to obtain very accurate predictions versus expensive non-local vdW functional calculations in which the required trainset is further reduced. The approach can be easily generalized to any other machine learning methods and opens a path to study the structure and dynamics of Pt clusters and nanoparticles. read less NOT USED (low confidence) T. D. Huan, R. Batra, J. Chapman, C. Kim, A. Chandrasekaran, and R. Ramprasad, “Iterative-Learning Strategy for the Development of Application-Specific Atomistic Force Fields,” The Journal of Physical Chemistry C. 2019. link Times cited: 18 Abstract: Emerging data-driven approaches in materials science have tr… read moreAbstract: Emerging data-driven approaches in materials science have triggered the development of numerous machine-learning force fields. In practice, they are constructed by training a statistical model on a reference database to predict potential energy and/or atomic forces. Although most of the force fields can accurately recover the properties of the training set, some of them are becoming useful for actual molecular dynamics simulations. In this work, we employ a simple iterative-learning strategy for the development of machine-learning force fields targeted at specific simulations (applications). The strategy involves (1) preparing and fingerprinting a diverse reference database of atomic configurations and forces, (2) generating a pool of machine-learning force fields by learning the reference data, (3) validating the force fields against a series of targeted applications, and (4) selectively and recursively improving the force fields that are unsuitable for a given application while keeping their performance... read less NOT USED (low confidence) G. S. Dhaliwal, P. Nair, and C. V. Singh, “Uncertainty and sensitivity analysis of mechanical and thermal properties computed through Embedded Atom Method potential,” Computational Materials Science. 2019. link Times cited: 9 NOT USED (low confidence) A. Goryaeva, J. Maillet, and M. Marinica, “Towards better efficiency of interatomic linear machine learning potentials,” Computational Materials Science. 2019. link Times cited: 34 NOT USED (low confidence) M. K. Bisbo and B. Hammer, “Efficient Global Structure Optimization with a Machine-Learned Surrogate Model.,” Physical review letters. 2019. link Times cited: 55 Abstract: We propose a scheme for global optimization with first-princ… read moreAbstract: We propose a scheme for global optimization with first-principles energy expressions of atomistic structure. While unfolding its search, the method actively learns a surrogate model of the potential energy landscape on which it performs a number of local relaxations (exploitation) and further structural searches (exploration). Assuming Gaussian processes, deploying two separate kernel widths to better capture rough features of the energy landscape while retaining a good resolution of local minima, an acquisition function is used to decide on which of the resulting structures is the more promising and should be treated at the first-principles level. The method is demonstrated to outperform by 2 orders of magnitude a well established first-principles based evolutionary algorithm in finding surface reconstructions. Finally, global optimization with first-principles energy expressions is utilized to identify initial stages of the edge oxidation and oxygen intercalation of graphene sheets on the Ir(111) surface. read less NOT USED (low confidence) L. Himanen, A. Geurts, A. Foster, and P. Rinke, “Data‐Driven Materials Science: Status, Challenges, and Perspectives,” Advanced Science. 2019. link Times cited: 362 Abstract: Data‐driven science is heralded as a new paradigm in materia… read moreAbstract: Data‐driven science is heralded as a new paradigm in materials science. In this field, data is the new resource, and knowledge is extracted from materials datasets that are too big or complex for traditional human reasoning—typically with the intent to discover new or improved materials or materials phenomena. Multiple factors, including the open science movement, national funding, and progress in information technology, have fueled its development. Such related tools as materials databases, machine learning, and high‐throughput methods are now established as parts of the materials research toolset. However, there are a variety of challenges that impede progress in data‐driven materials science: data veracity, integration of experimental and computational data, data longevity, standardization, and the gap between industrial interests and academic efforts. In this perspective article, the historical development and current state of data‐driven materials science, building from the early evolution of open science to the rapid expansion of materials data infrastructures are discussed. Key successes and challenges so far are also reviewed, providing a perspective on the future development of the field. read less NOT USED (low confidence) E. Flood, C. Boiteux, B. Lev, I. Vorobyov, and T. Allen, “Atomistic Simulations of Membrane Ion Channel Conduction, Gating, and Modulation.,” Chemical reviews. 2019. link Times cited: 70 Abstract: Membrane ion channels are the fundamental electrical compone… read moreAbstract: Membrane ion channels are the fundamental electrical components in the nervous system. Recent developments in X-ray crystallography and cryo-EM microscopy have revealed what these proteins look like in atomic detail but do not tell us how they function. Molecular dynamics simulations have progressed to the point that we can now simulate realistic molecular assemblies to produce quantitative calculations of the thermodynamic and kinetic quantities that control function. In this review, we summarize the state of atomistic simulation methods for ion channels to understand their conduction, activation, and drug modulation mechanisms. We are at a crossroads in atomistic simulation, where long time scale observation can provide unbiased exploration of mechanisms, supplemented by biased free energy methodologies. We illustrate the use of these approaches to describe ion conduction and selectivity in voltage-gated sodium and acid-sensing ion channels. Studies of channel gating present a significant challenge, as activation occurs on longer time scales. Enhanced sampling approaches can ensure convergence on minimum free energy pathways for activation, as illustrated here for pentameric ligand-gated ion channels that are principal to nervous system function and the actions of general anesthetics. We also examine recent studies of local anesthetic and antiepileptic drug binding to a sodium channel, revealing sites and pathways that may offer new targets for drug development. Modern simulations thus offer a range of molecular-level insights into ion channel function and modulation as a learning platform for mechanistic discovery and drug development. read less NOT USED (low confidence) Y. Zuo et al., “A Performance and Cost Assessment of Machine Learning Interatomic Potentials.,” The journal of physical chemistry. A. 2019. link Times cited: 413 Abstract: Machine learning of the quantitative relationship between lo… read moreAbstract: Machine learning of the quantitative relationship between local environment descriptors and the potential energy surface of a system of atoms has emerged as a new frontier in the development of interatomic potentials (IAPs). Here, we present a comprehensive evaluation of ML-IAPs based on four local environment descriptors --- atom-centered symmetry functions (ACSF), smooth overlap of atomic positions (SOAP), the Spectral Neighbor Analysis Potential (SNAP) bispectrum components, and moment tensors --- using a diverse data set generated using high-throughput density functional theory (DFT) calculations. The data set comprising bcc (Li, Mo) and fcc (Cu, Ni) metals and diamond group IV semiconductors (Si, Ge) is chosen to span a range of crystal structures and bonding. All descriptors studied show excellent performance in predicting energies and forces far surpassing that of classical IAPs, as well as predicting properties such as elastic constants and phonon dispersion curves. We observe a general trade-off between accuracy and the degrees of freedom of each model, and consequently computational cost. We will discuss these trade-offs in the context of model selection for molecular dynamics and other applications. read less NOT USED (low confidence) J. Li, K. Song, and J. Behler, “A critical comparison of neural network potentials for molecular reaction dynamics with exact permutation symmetry.,” Physical chemistry chemical physics : PCCP. 2019. link Times cited: 27 Abstract: The availability of accurate full-dimensional potential ener… read moreAbstract: The availability of accurate full-dimensional potential energy surfaces (PESs) is a mandatory condition for efficient computer simulations of molecular systems. Much effort has been devoted to developing reliable PESs with physically sound properties, such as the invariance of the energy with respect to the permutation of chemically identical atoms. In this work, we compare the performance of four neural network (NN)-based approaches with a rigorous permutation symmetry for fitting five typical reaction systems: OH + CO, H + H2S, H + NH3, H + CH4 and OH + CH4. The methods can be grouped into two categories, invariant polynomial based NNs and high-dimensional NN potentials (HD-NNPs). For the invariant polynomial based NNs, three types of polynomials, permutation invariant polynomials (PIPs), non-redundant PIPs (NRPIPs) and fundamental invariants (FIs), are used in the input layer of the NN. In HD-NNPs, the total energy is the sum of atomic contributions, each of which is given by an individual atomic NN with input vectors consisting of sets of atom-centered symmetry functions. Our results show that all methods exhibit a similar level of accuracy for the energies with respect to ab initio data obtained at the explicitly correlated coupled cluster level of theory (CCSD(T)-F12a). The HD-NNP method allows study of systems with larger numbers of atoms, making it more generally applicable than invariant polynomial based approaches, which in turn are computationally more efficient for smaller systems. To illustrate the applicability of the obtained potentials, quasi-classical trajectory calculations have been performed for the OH + CH4 → H2O + CH3 reaction to reveal its complicated mode specificity. read less NOT USED (low confidence) C. Zhan, W. Sun, Y. Xie, D. Jiang, and P. Kent, “Computational Discovery and Design of MXenes for Energy Applications: Status, Successes, and Opportunities.,” ACS applied materials & interfaces. 2019. link Times cited: 84 Abstract: MXenes (M n+1X n, e.g., Ti3C2) are the largest 2D material f… read moreAbstract: MXenes (M n+1X n, e.g., Ti3C2) are the largest 2D material family developed in recent years. They exhibit significant potential in the energy sciences, particularly for energy storage. In this review, we summarize the progress of the computational work regarding the theoretical design of new MXene structures and predictions for energy applications including their fundamental, energy storage, and catalytic properties. We also outline how high-throughput computation, big data, and machine-learning techniques can help broaden the MXene family. Finally, we present some of the major remaining challenges and future research directions needed to mature this novel materials family. read less NOT USED (low confidence) L. Constantin, “Semilocal properties of the Pauli kinetic potential,” Physical Review B. 2019. link Times cited: 11 NOT USED (low confidence) H. Wang, X. Guo, L. Zhang, H. Wang, and J. Xue, “Deep learning inter-atomic potential model for accurate irradiation damage simulations,” Applied Physics Letters. 2019. link Times cited: 32 Abstract: We propose a hybrid scheme that interpolates smoothly the Zi… read moreAbstract: We propose a hybrid scheme that interpolates smoothly the Ziegler-Biersack-Littmark (ZBL) screened nuclear repulsion potential with a newly developed deep learning potential energy model. The resulting DP-ZBL model can not only provide overall good performance on the predictions of near-equilibrium material properties but also capture the right physics when atoms are extremely close to each other, an event that frequently happens in computational simulations of irradiation damage events. We applied this scheme to the simulation of the irradiation damage processes in the face-centered-cubic aluminium system, and found better descriptions in terms of the defect formation energy, evolution of collision cascades, displacement threshold energy, and residual point defects, than the widely-adopted ZBL modified embedded atom method potentials and its variants. Our work provides a reliable and feasible scheme to accurately simulate the irradiation damage processes and opens up new opportunities to solve the predicament of lacking accurate potentials for enormous newly-discovered materials in the irradiation effect field. read less NOT USED (low confidence) A. Mannodi-Kanakkithodi et al., “Comprehensive Computational Study of Partial Lead Substitution in Methylammonium Lead Bromide,” Chemistry of Materials. 2019. link Times cited: 29 Abstract: Impurities in semiconductors, for example, lead-based hybrid… read moreAbstract: Impurities in semiconductors, for example, lead-based hybrid perovskites, have a major influence on their performance as photovoltaic (PV) light absorbers. While impurities could create harmful trap states that lead to nonradiative recombination of charge carriers and adversely affect PV efficiency, they could also potentially increase absorption via midgap energy levels that act as stepping stones for subgap photons or introduce charge carriers via doping. To unearth trends in impurity energy states, we use first principles density functional theory calculations to extensively study partial substitution of Pb in methylammonium lead bromide (MAPbBr3), a representative lead-halide perovskite. Investigation of the density of states and energy levels related to the transition of the substitutional defect from one charge state to another reveals that several elements create midgap energy states in MAPbBr3. We machine learned trends and design rules from the computational data and discovered that a few easily ... read less NOT USED (low confidence) R. Batra, G. Pilania, B. Uberuaga, and R. Ramprasad, “Multifidelity Information Fusion with Machine Learning: A Case Study of Dopant Formation Energies in Hafnia.,” ACS applied materials & interfaces. 2019. link Times cited: 51 Abstract: Cost versus accuracy trade-offs are frequently encountered i… read moreAbstract: Cost versus accuracy trade-offs are frequently encountered in materials science and engineering, where a particular property of interest can be measured/computed at different levels of accuracy or fidelity. Naturally, the most accurate measurement is also the most resource and time intensive, while the inexpensive quicker alternatives tend to be noisy. In such situations, a number of machine learning (ML) based multifidelity information fusion (MFIF) strategies can be employed to fuse information accessible from varying sources of fidelity and make predictions at the highest level of accuracy. In this work, we perform a comparative study on traditionally employed single-fidelity and three MFIF strategies, namely, (1) Δ-learning, (2) low-fidelity as a feature, and (3) multifidelity cokriging (CK) to compare their relative prediction accuracies and efficiencies for accelerated property predictions and high throughput chemical space explorations. We perform our analysis using a dopant formation energy data set for hafnia, which is a well-known high- k material and is being extensively studied for its promising ferroelectric, piezoelectric, and pyroelectric properties. We use a dopant formation energy data set of 42 dopants in hafnia-each studied in six different hafnia phases-computed at two levels of fidelities to find merits and limitations of these ML strategies. The findings of this work indicate that the MFIF based learning schemes outperform the traditional SF machine learning methods, such as Gaussian process regression and CK provides an accurate, inexpensive and flexible alternative to other MFIF strategies. While the results presented here are for the case study of hafnia, they are expected to be general. Therefore, materials discovery problems that involve huge chemical space explorations can be studied efficiently (or even made feasible in some situations) through a combination of a large number of low-fidelity and a few high-fidelity measurements/computations, in conjunction with the CK approach. read less NOT USED (low confidence) L. Chen, S. Venkatram, C. Kim, R. Batra, A. Chandrasekaran, and R. Ramprasad, “Electrochemical Stability Window of Polymeric Electrolytes,” Chemistry of Materials. 2019. link Times cited: 61 Abstract: The electrochemical stability window (ESW) is a fundamental … read moreAbstract: The electrochemical stability window (ESW) is a fundamental consideration for choosing polymers as solid electrolytes in lithium-ion batteries. Morphological and chemical aspects of the polymer matrix and its complex interactions with lithium salts make it difficult to estimate the ESW of the polymer electrolyte, either computationally or experimentally. In this work, we propose a practical computational procedure to estimate the ESW due to just one dominant factor, i.e., the polymer matrix, using first-principles density functional theory computations. Diverse model polymers (10) were investigated, namely, polyethylene, polyketone, poly(ethylene oxide), poly(propylene oxide), poly(vinyl alcohol), polycaprolactone, poly(methyl methacrylate), poly(ethyl acrylate), poly(vinyl chloride), and poly(vinylidene fluoride). For each case, an increasingly complex hierarchy of structural models was considered to elucidate the impact of polymer chemistry and the morphological complexity on the ESW. Favorable agreemen... read less NOT USED (low confidence) N. Jackson, A. Bowen, L. W. Antony, M. A. Webb, V. Vishwanath, and J. D. de Pablo, “Electronic structure at coarse-grained resolutions from supervised machine learning,” Science Advances. 2019. link Times cited: 42 Abstract: Machine learning–enhanced molecular simulation opens a pathw… read moreAbstract: Machine learning–enhanced molecular simulation opens a pathway to multiscale prediction for organic electronics. Computational studies aimed at understanding conformationally dependent electronic structure in soft materials require a combination of classical and quantum-mechanical simulations, for which the sampling of conformational space can be particularly demanding. Coarse-grained (CG) models provide a means of accessing relevant time scales, but CG configurations must be back-mapped into atomistic representations to perform quantum-chemical calculations, which is computationally intensive and inconsistent with the spatial resolution of the CG models. A machine learning approach, denoted as artificial neural network electronic coarse graining (ANN-ECG), is presented here in which the conformationally dependent electronic structure of a molecule is mapped directly to CG pseudo-atom configurations. By averaging over decimated degrees of freedom, ANN-ECG accelerates simulations by eliminating backmapping and repeated quantum-chemical calculations. The approach is accurate, consistent with the CG spatial resolution, and can be used to identify computationally optimal CG resolutions. read less NOT USED (low confidence) S. T. Hutchinson and R. Kobayashi, “Solvent-Specific Featurization for Predicting Free Energies of Solvation through Machine Learning,” Journal of chemical information and modeling. 2019. link Times cited: 30 Abstract: A featurization algorithm based on functional class fingerpr… read moreAbstract: A featurization algorithm based on functional class fingerprints has been implemented within the DeepChem machine learning framework. It is based on descriptors more appropriate for solvation, taking into account intermolecular properties, and has been used in the prediction of free energies of solvation. Tests carried out on solvents with a range of polarity from the FreeSolv and MNSol data sets have shown slightly better accuracy than the commonly used topology-based extended connectivity fingerprint algorithm for hydration free energies. However, improvement was not as significant as hoped and less clear for less polar solvents suggesting that further solvent-specific descriptors may need to be taken into consideration. read less NOT USED (low confidence) J. Chapman, R. Batra, B. Uberuaga, G. Pilania, and R. Ramprasad, “A comprehensive computational study of adatom diffusion on the aluminum (1 0 0) surface,” Computational Materials Science. 2019. link Times cited: 9 NOT USED (low confidence) L. Joss and E. A. Müller, “Machine Learning for Fluid Property Correlations: Classroom Examples with MATLAB,” Journal of Chemical Education. 2019. link Times cited: 39 Abstract: Recent advances in computer hardware and algorithms are spaw… read moreAbstract: Recent advances in computer hardware and algorithms are spawning an explosive growth in the use of computer-based systems aimed at analyzing and ultimately correlating large amounts of experimental and synthetic data. As these machine learning tools become more widespread, it is becoming imperative that scientists and researchers become familiar with them, both in terms of understanding the tools and the current limitations of artificial intelligence, and more importantly being able to critically separate the hype from the real potential. This article presents a classroom exercise aimed at first-year science and engineering college students, where a task is set to produce a correlation to predict the normal boiling point of organic compounds from an unabridged data set of >6000 compounds. The exercise, which is fully documented in terms of the problem statement and the solution, guides the students to initially perform a linear correlation of the boiling point data with a plausible relevant variable (the ... read less NOT USED (low confidence) W. Li and Y. Ando, “Comparison of different machine learning models for the prediction of forces in copper and silicon dioxide.,” Physical chemistry chemical physics : PCCP. 2018. link Times cited: 17 Abstract: Recently, the machine learning (ML) force field has emerged … read moreAbstract: Recently, the machine learning (ML) force field has emerged as a powerful atomic simulation approach because of its high accuracy and low computational cost. However, there have been relatively fewer applications to multicomponent materials. In this study, we construct and compare ML force fields for both an elemental material (Cu) and binary material (SiO2) with varied inputs and regression models. The atomic environments are described by structural fingerprints that take into account the bond angle, and then, different ML techniques, including linear regression, a neural network and a mixture model method, are used to learn the structure-force relationship. We found that using angular structural fingerprints and a mixture model method significantly improves the accuracy of ML force fields. Additionally, we discuss an effective structural fingerprint auto-selection method based on the least absolute shrinkage and selection operator and the genetic algorithm. The atomic simulations conducted for ML force fields are in excellent agreement with ab initio calculations. As a result of the simulation with our ML force field for the structural and vibrational properties of amorphous SiO2, simulated annealing with a slow cooling rate improved the ring statistics in the amorphous structure and the phonon density of states. read less NOT USED (low confidence) D. Das, A. Chandrasekaran, S. Venkatram, and R. Ramprasad, “Effect of Crystallinity on Li Adsorption in Polyethylene Oxide,” Chemistry of Materials. 2018. link Times cited: 19 Abstract: Despite their numerous benefits, the use of solid-polymer el… read moreAbstract: Despite their numerous benefits, the use of solid-polymer electrolytes, such as polyethylene oxide (PEO), in next-generation Li-ion batteries is constrained by their lower ionic conductivity. To overcome this bottleneck and design materials with higher conductivity, it is important to elucidate the underlying atomistic mechanisms of Li-ion adsorption in such materials. Here, we have performed a comprehensive statistical analysis of the interaction of Li and Li+ at numerous locations in crystalline and amorphous PEO. Our in-depth analysis of the Li–O bonding environment using ab initio calculations reveals that Li/Li+ can bind with more number of oxygen atoms in amorphous PEO compared to crystalline case. The maximum value of coordination number, that is, the number of oxygen atoms bonded with Li/Li+ is 3 for crystalline PEO and 5 for amorphous PEO. This can be attributed to the access to more neighboring oxygen atoms in amorphous PEO. Binding energy calculations reveal that the interaction of Li and Li+ s... read less NOT USED (low confidence) S. Sundararaman, L. Huang, S. Ispas, and W. Kob, “New interaction potentials for alkali and alkaline-earth aluminosilicate glasses.,” The Journal of chemical physics. 2018. link Times cited: 36 Abstract: We apply a recently developed optimization scheme to obtain … read moreAbstract: We apply a recently developed optimization scheme to obtain effective potentials for alkali and alkaline-earth aluminosilicate glasses that contain lithium, sodium, potassium, or calcium as modifiers. As input data for the optimization, we used the radial distribution functions of the liquid at high temperature generated by means of ab initio molecular dynamics simulations and density and elastic modulus of glass at room temperature from experiments. The new interaction potentials are able to reproduce reliably the structure and various mechanical and vibrational properties over a wide range of compositions for binary silicates. We have tested these potentials for various ternary systems and find that they are transferable and can be mixed, thus allowing us to reproduce and predict the structure and properties of multicomponent glasses. read less NOT USED (low confidence) A. Jonayat, A. V. van Duin, and M. Janik, “Discovery of Descriptors for Stable Monolayer Oxide Coatings through Machine Learning,” ACS Applied Energy Materials. 2018. link Times cited: 11 Abstract: Monolayer metal oxides (MMOs) can provide tunable chemical p… read moreAbstract: Monolayer metal oxides (MMOs) can provide tunable chemical properties dictated by choice of the support and coating metal oxide. Experimental discovery of (meta)stable coating/support combinations can be accelerated if stability could be predicted based on component physical properties. For such complex systems, machine learning approaches can help to discover underlying principles that dictate system properties. Herein, we use a supervised machine learning (ML) method, regressed against density functional theory-calculated monolayer stabilities, to predict physical properties that are predictive of metal oxide monolayer stability. Monolayer oxide coatings are considered in two classes: (1) “stoichiometric” coatings, in which the monolayer oxide has a stable phase at the same MOx stoichiometry as the substrate, and (2) “nonstoichiometric” coatings. Our ML approach indicates that substrate surface energy, orbital radii, and ionization energies are important for stability of stoichiometric MMOs. The parent ... read less NOT USED (low confidence) L. Chen, R. Batra, R. Ranganathan, G. Sotzing, Y. Cao, and R. Ramprasad, “Electronic Structure of Polymer Dielectrics: The Role of Chemical and Morphological Complexity,” Chemistry of Materials. 2018. link Times cited: 26 Abstract: The electronic structure of polymers contains signatures tha… read moreAbstract: The electronic structure of polymers contains signatures that correlate with their short-term and long-term integrity when subjected to large electric stresses. A detailed picture of the electronic structure of realistic models of polymers has been difficult to obtain, mainly due to the chemical and morphological complexity encountered in polymers. In this work, we have undertaken a comprehensive analysis of the electronic structure of six model polymers displaying chemical and morphological diversity, namely, polyethylene (PE), polypropylene (PP), polystyrene (PS), poly(methyl methacrylate) (PMMA), polyethylene terephthalate (PET), and polybutylene terephthalate (PBT), using first-principles density functional theory computations and classical molecular dynamics simulations. In particular, we have studied the role of monomer chemistry, tacticity, and large-scale morphological disorders in shaping the electronic structure of these polymers. We find that monomer chemistry and morphological disorder coopera... read less NOT USED (low confidence) P. Wang, Y. Shao, H. Wang, and W. Yang, “Accurate interatomic force field for molecular dynamics simulation by hybridizing classical and machine learning potentials,” Extreme Mechanics Letters. 2018. link Times cited: 5 NOT USED (low confidence) K. K. Bejagam, S. Singh, Y. An, and S. A. Deshmukh, “Machine-Learned Coarse-Grained Models.,” The journal of physical chemistry letters. 2018. link Times cited: 51 Abstract: Optimizing force-field (FF) parameters to perform molecular … read moreAbstract: Optimizing force-field (FF) parameters to perform molecular dynamics (MD) simulations is a challenging and time-consuming process. We present a novel FF optimization framework that integrates MD simulations with particle swarm optimization (PSO) algorithm and artificial neural network (ANN). This new ANN-assisted PSO framework was used to develop transferable coarse-grained (CG) models for D2O and DMF as a proof of concept. The PSO algorithm was used to generate the set of input FF parameters for the MD simulations of the CG models of these solvents, which were optimized to reproduce their experimental properties. Herein, for the first time, a reverse approach was employed for on-the-fly training of the ANN model, where results (solvent properties) obtained from the MD simulations and their corresponding FF parameters were used as inputs and outputs, respectively. The ANN model was then required to predict a set of new FF parameters, which were tested for their ability to predict the desired experimental properties. This new framework can be extended to integrate any optimization algorithm with ANN and MD simulations to accelerate the FF development. read less NOT USED (low confidence) C. Kim, A. Chandrasekaran, T. D. Huan, D. Das, and R. Ramprasad, “Polymer Genome: A Data-Powered Polymer Informatics Platform for Property Predictions,” The Journal of Physical Chemistry C. 2018. link Times cited: 254 Abstract: The recent successes of the Materials Genome Initiative have… read moreAbstract: The recent successes of the Materials Genome Initiative have opened up new opportunities for data-centric informatics approaches in several subfields of materials research, including in polymer science and engineering. Polymers, being inexpensive and possessing a broad range of tunable properties, are widespread in many technological applications. The vast chemical and morphological complexity of polymers though gives rise to challenges in the rational discovery of new materials for specific applications. The nascent field of polymer informatics seeks to provide tools and pathways for accelerated property prediction (and materials design) via surrogate machine learning models built on reliable past data. We have carefully accumulated a data set of organic polymers whose properties were obtained either computationally (bandgap, dielectric constant, refractive index, and atomization energy) or experimentally (glass transition temperature, solubility parameter, and density). A fingerprinting scheme that capt... read less NOT USED (low confidence) A. Medford, M. R. Kunz, S. M. Ewing, T. L. Borders, and R. Fushimi, “Extracting Knowledge from Data through Catalysis Informatics,” ACS Catalysis. 2018. link Times cited: 135 Abstract: Catalysis informatics is a distinct subfield that lies at th… read moreAbstract: Catalysis informatics is a distinct subfield that lies at the intersection of cheminformatics and materials informatics but with distinctive challenges arising from the dynamic, surface-sensitive, and multiscale nature of heterogeneous catalysis. The ideas behind catalysis informatics can be traced back decades, but the field is only recently emerging due to advances in data infrastructure, statistics, machine learning, and computational methods. In this work, we review the field from early works on expert systems and knowledge engines to more recent approaches utilizing machine-learning and uncertainty quantification. The data–information–knowledge hierarchy is introduced and used to classify various developments. The chemical master equation and microkinetic models are proposed as a quantitative representation of catalysis knowledge, which can be used to generate explanative and predictive hypotheses for the understanding and discovery of catalytic materials. We discuss future prospects for the field, i... read less NOT USED (low confidence) L. Shen and W. Yang, “Molecular Dynamics Simulations with Quantum Mechanics/Molecular Mechanics and Adaptive Neural Networks.,” Journal of chemical theory and computation. 2018. link Times cited: 92 Abstract: Direct molecular dynamics (MD) simulation with ab initio qua… read moreAbstract: Direct molecular dynamics (MD) simulation with ab initio quantum mechanical and molecular mechanical (QM/MM) methods is very powerful for studying the mechanism of chemical reactions in a complex environment but also very time-consuming. The computational cost of QM/MM calculations during MD simulations can be reduced significantly using semiempirical QM/MM methods with lower accuracy. To achieve higher accuracy at the ab initio QM/MM level, a correction on the existing semiempirical QM/MM model is an attractive idea. Recently, we reported a neural network (NN) method as QM/MM-NN to predict the potential energy difference between semiempirical and ab initio QM/MM approaches. The high-level results can be obtained using neural network based on semiempirical QM/MM MD simulations, but the lack of direct MD samplings at the ab initio QM/MM level is still a deficiency that limits the applications of QM/MM-NN. In the present paper, we developed a dynamic scheme of QM/MM-NN for direct MD simulations on the NN-predicted potential energy surface to approximate ab initio QM/MM MD. Since some configurations excluded from the database for NN training were encountered during simulations, which may cause some difficulties on MD samplings, an adaptive procedure inspired by the selection scheme reported by Behler [ Behler Int. J. Quantum Chem. 2015 , 115 , 1032 ; Behler Angew. Chem., Int. Ed. 2017 , 56 , 12828 ] was employed with some adaptions to update NN and carry out MD iteratively. We further applied the adaptive QM/MM-NN MD method to the free energy calculation and transition path optimization on chemical reactions in water. The results at the ab initio QM/MM level can be well reproduced using this method after 2-4 iteration cycles. The saving in computational cost is about 2 orders of magnitude. It demonstrates that the QM/MM-NN with direct MD simulations has great potentials not only for the calculation of thermodynamic properties but also for the characterization of reaction dynamics, which provides a useful tool to study chemical or biochemical systems in solution or enzymes. read less NOT USED (low confidence) T. Kiyoyuki, “Collaboration between Third and Fourth Paradigms Reveals Complex Atomic Processes.” 2017. link Times cited: 0 Abstract: The machine learning technique (e.g., neural network) enable… read moreAbstract: The machine learning technique (e.g., neural network) enables us to map accurately interatomic potentials obtained by quantum mechanical calculations onto classical potentials. A neural-network potential was successfully applied to the study of Cu atom diffusion in amorphous Ta2O5. read less NOT USED (low confidence) A. Mannodi-Kanakkithodi, T. D. Huan, and R. Ramprasad, “Mining materials design rules from data: The example of polymer dielectrics,” Chemistry of Materials. 2017. link Times cited: 44 Abstract: Mining of currently available and evolving materials databas… read moreAbstract: Mining of currently available and evolving materials databases to discover structure–chemistry–property relationships is critical to developing an accelerated materials design framework. The design of new and advanced polymeric dielectrics for capacitive energy storage has been hampered by the lack of sufficient data encompassing wide enough chemical spaces. Here, data mining and analysis techniques are applied on a recently presented computational data set of around 1100 organic polymers, organometallic polymers, and related molecular crystals, in order to obtain qualitative understanding of the origins of dielectric and electronic properties. By probing the relationships between crucial chemical and structural features of materials and their dielectric constant and band gap, design rules are devised for optimizing either property. Learning from this data set provides guidance to experiments and to future computations, as well as a way of expanding the pool of promising polymer candidates for dielectric ... read less NOT USED (low confidence) T. Bereau, R. Distasio, A. Tkatchenko, and O. A. von Lilienfeld, “Non-covalent interactions across organic and biological subsets of chemical space: Physics-based potentials parametrized from machine learning.,” The Journal of chemical physics. 2017. link Times cited: 128 Abstract: Classical intermolecular potentials typically require an ext… read moreAbstract: Classical intermolecular potentials typically require an extensive parametrization procedure for any new compound considered. To do away with prior parametrization, we propose a combination of physics-based potentials with machine learning (ML), coined IPML, which is transferable across small neutral organic and biologically relevant molecules. ML models provide on-the-fly predictions for environment-dependent local atomic properties: electrostatic multipole coefficients (significant error reduction compared to previously reported), the population and decay rate of valence atomic densities, and polarizabilities across conformations and chemical compositions of H, C, N, and O atoms. These parameters enable accurate calculations of intermolecular contributions-electrostatics, charge penetration, repulsion, induction/polarization, and many-body dispersion. Unlike other potentials, this model is transferable in its ability to handle new molecules and conformations without explicit prior parametrization: All local atomic properties are predicted from ML, leaving only eight global parameters-optimized once and for all across compounds. We validate IPML on various gas-phase dimers at and away from equilibrium separation, where we obtain mean absolute errors between 0.4 and 0.7 kcal/mol for several chemically and conformationally diverse datasets representative of non-covalent interactions in biologically relevant molecules. We further focus on hydrogen-bonded complexes-essential but challenging due to their directional nature-where datasets of DNA base pairs and amino acids yield an extremely encouraging 1.4 kcal/mol error. Finally, and as a first look, we consider IPML for denser systems: water clusters, supramolecular host-guest complexes, and the benzene crystal. read less NOT USED (low confidence) B. K. Carpenter, G. S. Ezra, S. Farantos, Z. C. Kramer, and S. Wiggins, “Empirical Classification of Trajectory Data: An Opportunity for the Use of Machine Learning in Molecular Dynamics.,” The journal of physical chemistry. B. 2017. link Times cited: 18 Abstract: Classical Hamiltonian trajectories are initiated at random p… read moreAbstract: Classical Hamiltonian trajectories are initiated at random points in phase space on a fixed energy shell of a model two degrees of freedom potential, consisting of two interacting minima in an otherwise flat energy plane of infinite extent. Below the energy of the plane, the dynamics are demonstrably chaotic. However, most of the work in this paper involves trajectories at a fixed energy that is 1% above that of the plane, in which regime the dynamics exhibit behavior characteristic of chaotic scattering. The trajectories are analyzed without reference to the potential, as if they had been generated in a typical direct molecular dynamics simulation. The questions addressed are whether one can recover useful information about the structures controlling the dynamics in phase space from the trajectory data alone, and whether, despite the at least partially chaotic nature of the dynamics, one can make statistically meaningful predictions of trajectory outcomes from initial conditions. It is found that key unstable periodic orbits, which can be identified on the analytical potential, appear by simple classification of the trajectories, and that the specific roles of these periodic orbits in controlling the dynamics are also readily discerned from the trajectory data alone. Two different approaches to predicting trajectory outcomes from initial conditions are evaluated, and it is shown that the more successful of them has ∼90% success. The results are compared with those from a simple neural network, which has higher predictive success (97%) but requires the information obtained from the "by-hand" analysis to achieve that level. Finally, the dynamics, which occur partly on the very flat region of the potential, show characteristics of the much-studied phenomenon called "roaming." On this potential, it is found that roaming trajectories are effectively "failed" periodic orbits and that angular momentum can be identified as a key controlling factor, despite the fact that it is not a strictly conserved quantity. It is also noteworthy that roaming on this potential occurs in the absence of a "roaming saddle," which has previously been hypothesized to be a necessary feature for roaming to occur. read less NOT USED (low confidence) J. A. Gomberg, A. Medford, and S. Kalidindi, “Extracting knowledge from molecular mechanics simulations of grain boundaries using machine learning,” Acta Materialia. 2017. link Times cited: 43 NOT USED (low confidence) V. L. Deringer, G. Csányi, and D. Proserpio, “Extracting Crystal Chemistry from Amorphous Carbon Structures,” Chemphyschem. 2017. link Times cited: 71 Abstract: Carbon allotropes have been explored intensively by ab initi… read moreAbstract: Carbon allotropes have been explored intensively by ab initio crystal structure prediction, but such methods are limited by the large computational cost of the underlying density functional theory (DFT). Here we show that a novel class of machine‐learning‐based interatomic potentials can be used for random structure searching and readily predicts several hitherto unknown carbon allotropes. Remarkably, our model draws structural information from liquid and amorphous carbon exclusively, and so does not have any prior knowledge of crystalline phases: it therefore demonstrates true transferability, which is a crucial prerequisite for applications in chemistry. The method is orders of magnitude faster than DFT and can, in principle, be coupled with any algorithm for structure prediction. Machine‐learning models therefore seem promising to enable large‐scale structure searches in the future. read less NOT USED (low confidence) S. Natarajan and J. Behler, “Self-Diffusion of Surface Defects at Copper–Water Interfaces,” Journal of Physical Chemistry C. 2017. link Times cited: 31 Abstract: Solid–liquid interfaces play an important role in many field… read moreAbstract: Solid–liquid interfaces play an important role in many fields like electrochemistry, corrosion, and heterogeneous catalysis. For understanding the related processes, detailed insights into the elementary steps at the atomic level are mandatory. Here we unravel the properties of prototypical surface-defects like adatoms and vacancies at a number of copper–water interfaces including the low-index Cu(111), Cu(100), and Cu(110), as well as the stepped Cu(211) and Cu(311) surfaces. Using a first-principles quality neural network potential constructed from density functional theory reference data in combination with molecular dynamics and metadynamics simulations, we investigate the defect diffusion mechanisms and the associated free energy barriers. Further, the solvent structure and the mobility of the interfacial water molecules close to the defects are analyzed and compared to the defect-free surfaces. We find that, like at the copper–vacuum interface, hopping mechanisms are preferred compared to exchange m... read less NOT USED (low confidence) J. Vermaas, L. Petridis, G. Beckham, and M. Crowley, “Systematic Parameterization of Lignin for the Charmm Force Field,” Biophysical Journal. 2017. link Times cited: 21 NOT USED (low confidence) V. Botu, J. Chapman, and R. Ramprasad, “A study of adatom ripening on an Al (1 1 1) surface with machine learning force fields,” Computational Materials Science. 2016. link Times cited: 31 NOT USED (low confidence) M. Cusentino et al., “Molecular dynamics of high pressure tin phases: Empirical and machine learned interatomic potentials,” SHOCK COMPRESSION OF CONDENSED MATTER - 2022: Proceedings of the Conference of the American Physical Society Topical Group on Shock Compression of Condensed Matter. 2023. link Times cited: 0 NOT USED (low confidence) G. Sivaraman et al., “A Combined Machine Learning and High-Energy X-ray Diffraction Approach to Understanding Liquid and Amorphous Metal Oxides,” Journal of the Physical Society of Japan. 2022. link Times cited: 2 NOT USED (low confidence) X. Liu, Q. Wang, and J. Zhang, “Machine Learning Interatomic Force Fields for Carbon Allotropic Materials.” 2021. link Times cited: 0 NOT USED (low confidence) P. O. Dral, “Quantum chemistry assisted by machine learning.” 2020. link Times cited: 14 NOT USED (low confidence) A. Chandrasekaran, C. Kim, and R. Ramprasad, “Polymer Genome: A Polymer Informatics Platform to Accelerate Polymer Discovery.” 2020. link Times cited: 5 NOT USED (low confidence) G. Ackland and G. Bonny, “Interatomic Potential Development,” Comprehensive Nuclear Materials. 2020. link Times cited: 4 NOT USED (high confidence) B. G. del Rio, B. Phan, and R. Ramprasad, “A deep learning framework to emulate density functional theory,” npj Computational Materials. 2023. link Times cited: 0 NOT USED (high confidence) H. Jin, E. Zhang, and H. Espinosa, “Recent Advances and Applications of Machine Learning in Experimental Solid Mechanics: A Review,” ArXiv. 2023. link Times cited: 6 Abstract:
For many decades, experimental solid mechanics has played … read moreAbstract:
For many decades, experimental solid mechanics has played a crucial role in characterizing and understanding the mechanical properties of natural and novel materials. Recent advances in machine learning (ML) provide new opportunities for the field, including experimental design, data analysis, uncertainty quantification, and inverse problems. As the number of papers published in recent years in this emerging field is exploding, it is timely to conduct a comprehensive and up-to-date review of recent ML applications in experimental solid mechanics. Here, we first provide an overview of common ML algorithms and terminologies that are pertinent to this review, with emphasis placed on physics-informed and physics-based ML methods. Then, we provide thorough coverage of recent ML applications in traditional and emerging areas of experimental mechanics, including fracture mechanics, biomechanics, nano- and micro-mechanics, architected materials, and 2D material. Finally, we highlight some current challenges of applying ML to multi-modality and multi-fidelity experimental datasets and propose several future research directions. This review aims to provide valuable insights into the use of ML methods as well as a variety of examples for researchers in solid mechanics to integrate into their experiments. read less NOT USED (high confidence) C. Lapointe, T. Swinburne, L. Proville, C. Becquart, N. Mousseau, and M. Marinica, “Machine learning surrogate models for strain-dependent vibrational properties and migration rates of point defects,” Physical Review Materials. 2022. link Times cited: 2 NOT USED (high confidence) K. S. Csizi and M. Reiher, “Universal QM/MM approaches for general nanoscale applications,” Wiley Interdisciplinary Reviews: Computational Molecular Science. 2022. link Times cited: 6 Abstract: Quantum mechanics/molecular mechanics (QM/MM) hybrid models … read moreAbstract: Quantum mechanics/molecular mechanics (QM/MM) hybrid models allow one to address chemical phenomena in complex molecular environments. Whereas this modeling approach can cope with a large system size at moderate computational costs, the models are often tedious to construct and require manual preprocessing and expertise. As a result, transferability to new application areas can be limited and the many parameters are not easy to adjust to reference data that are typically scarce. Therefore, it is desirable to devise automated procedures of controllable accuracy, which enables such modeling in a standardized and black‐box‐type manner. Although diverse best‐practice protocols have been set up for the construction of individual components of a QM/MM model (e.g., the MM potential, the type of embedding, the choice of the QM region), automated procedures that reconcile all steps of the QM/MM model construction are still rare. Here, we review the state of the art of QM/MM modeling with a focus on automation. We elaborate on MM model parametrization, on atom‐economical physically‐motivated QM region selection, and on embedding schemes that incorporate mutual polarization as critical components of the QM/MM model. In view of the broad scope of the field, we mostly restrict the discussion to methodologies that build de novo models based on first‐principles data, on uncertainty quantification, and on error mitigation with a high potential for automation. Ultimately, it is desirable to be able to set up reliable QM/MM models in a fast and efficient automated way without being constrained by specific chemical or technical limitations. read less NOT USED (high confidence) A. Pedone, M. Bertani, L. Brugnoli, and A. Pallini, “Interatomic potentials for oxide glasses: Past, present, and future,” Journal of Non-Crystalline Solids: X. 2022. link Times cited: 2 NOT USED (high confidence) S. Pathak et al., “Accurate Hellmann-Feynman forces from density functional calculations with augmented Gaussian basis sets.,” The Journal of chemical physics. 2022. link Times cited: 2 Abstract: The Hellmann-Feynman (HF) theorem provides a way to compute … read moreAbstract: The Hellmann-Feynman (HF) theorem provides a way to compute forces directly from the electron density, enabling efficient force calculations for large systems through machine learning (ML) models for the electron density. The main issue holding back the general acceptance of the HF approach for atom-centered basis sets is the well-known Pulay force which, if naively discarded, typically constitutes an error upward of 10 eV/Å in forces. In this work, we demonstrate that if a suitably augmented Gaussian basis set is used for density functional calculations, the Pulay force can be suppressed, and HF forces can be computed as accurately as analytical forces with state-of-the-art basis sets, allowing geometry optimization and molecular dynamics to be reliably performed with HF forces. Our results pave a clear path forward for the accurate and efficient simulation of large systems using ML densities and the HF theorem. read less NOT USED (high confidence) Y.-H. Liu, S. Zhang, P. Zhang, T.-K. Lee, and G. Chern, “Machine learning predictions for local electronic properties of disordered correlated electron systems,” ArXiv. 2022. link Times cited: 4 Abstract: Yi-Hsuan Liu, 2 Sheng Zhang, Puhan Zhang, Ting-Kuo Lee, 2, 4… read moreAbstract: Yi-Hsuan Liu, 2 Sheng Zhang, Puhan Zhang, Ting-Kuo Lee, 2, 4 and Gia-Wei Chern Department of Physics, National Tsing Hua University, Hsinchu 30013, Taiwan Institute of Physics, Academia Sinica, Nankang 11529, Taiwan Department of Physics, University of Virginia, Charlottesville, Virginia 22904, USA Department of Physics, National Sun Yat-sen University, Kaohsiun 80424, Taiwan (Dated: April 13, 2022) read less NOT USED (high confidence) K. Zongo, Béland, and O. ‐ Plamondon, “First-principles database for fitting a machine-learning silicon interatomic force field,” MRS Advances. 2022. link Times cited: 0 Abstract: Data-driven machine learning has emerged to address the limi… read moreAbstract: Data-driven machine learning has emerged to address the limitations of traditional methods when modeling interatomic interactions in materials, such as electronic density functional theory (DFT) and semi-empirical potentials. These machine-learning frameworks involve mathematical models coupled to quantum mechanical data. In the present article, we focus on the moment tensor potential (MTP) machine-learning framework. More specifically, we provide an account of the development of a preliminary MTP for silicon, including details pertaining to the construction of a DFT database. read less NOT USED (high confidence) G. S. Dhaliwal, P. Nair, and C. V. Singh, “Machine learned interatomic potentials using random features,” npj Computational Materials. 2022. link Times cited: 10 NOT USED (high confidence) L. Chen, X. Zhang, A. Chen, S. Yao, X. Hu, and Z. Zhou, “Targeted design of advanced electrocatalysts by machine learning,” Chinese Journal of Catalysis. 2022. link Times cited: 20 NOT USED (high confidence) P. Zhang and G. Chern, “Machine learning nonequilibrium electron forces for spin dynamics of itinerant magnets,” npj Computational Materials. 2021. link Times cited: 4 NOT USED (high confidence) Z. Li, K. Meidani, P. Yadav, and A. Farimani, “Graph Neural Networks Accelerated Molecular Dynamics,” The Journal of chemical physics. 2021. link Times cited: 28 Abstract: Molecular Dynamics (MD) simulation is a powerful tool for un… read moreAbstract: Molecular Dynamics (MD) simulation is a powerful tool for understanding the dynamics and structure of matter. Since the resolution of MD is atomic-scale, achieving long timescale simulations with femtosecond integration is very expensive. In each MD step, numerous iterative computations are performed to calculate energy based on different types of interaction and their corresponding spatial gradients. These repetitive computations can be learned and surrogated by a deep learning model, such as a Graph Neural Network (GNN). In this work, we developed a GNN Accelerated MD (GAMD) model that directly predicts forces, given the state of the system (atom positions, atom types), bypassing the evaluation of potential energy. By training the GNN on a variety of data sources (simulation data derived from classical MD and density functional theory), we show that GAMD can predict the dynamics of two typical molecular systems, Lennard-Jones system and water system, in the NVT ensemble with velocities regulated by a thermostat. We further show that GAMD's learning and inference are agnostic to the scale, where it can scale to much larger systems at test time. We also perform a comprehensive benchmark test comparing our implementation of GAMD to production-level MD software, showing GAMD's competitive performance on the large-scale simulation. read less NOT USED (high confidence) M. Mansoor et al., “Ab-initio calculation of point defect equilibria during heat treatment: Nitrogen, hydrogen, and silicon doped diamond,” Diamond and Related Materials. 2021. link Times cited: 5 NOT USED (high confidence) A. Goryaeva et al., “Efficient and transferable machine learning potentials for the simulation of crystal defects in bcc Fe and W,” Physical Review Materials. 2021. link Times cited: 20 NOT USED (high confidence) A. Ghosh, M. Ziatdinov, O. Dyck, B. Sumpter, and S. Kalinin, “Bridging microscopy with molecular dynamics and quantum simulations: an atomAI based pipeline,” npj Computational Materials. 2021. link Times cited: 7 NOT USED (high confidence) J. Goniakowski, S. Menon, G. Laurens, and J. Lam, “Nonclassical Nucleation of Zinc Oxide from a Physically Motivated Machine-Learning Approach,” The Journal of Physical Chemistry C. 2021. link Times cited: 3 Abstract: Observing non-classical nucleation pathways remains challeng… read moreAbstract: Observing non-classical nucleation pathways remains challenging in simulations of complex materials with technological interests. This is because it requires very accurate force fields that can capture the whole complexity of their underlying interatomic interactions and an advanced structural analysis. Here, we first report the construction of a machine-learning force field for zinc oxide interactions using the Physical LassoLars Interaction Potentials approach which allows us to be predictive even for untrained structures. Then, we carried out freezing simulations from a liquid and observed the crystal formation with atomistic precision. Our results, which are analyzed using a data-driven approach based on bond order parameters, demonstrate the presence of both prenucleation clusters and two-step nucleation scenarios thus retrieving seminal predictions of non-classical nucleation pathways made on much simpler models. read less NOT USED (high confidence) P. Rajak, A. Krishnamoorthy, A. Mishra, R. Kalia, A. Nakano, and P. Vashishta, “Autonomous reinforcement learning agent for chemical vapor deposition synthesis of quantum materials,” npj Computational Materials. 2021. link Times cited: 15 NOT USED (high confidence) J. Behler and G. Csányi, “Machine learning potentials for extended systems: a perspective,” The European Physical Journal B. 2021. link Times cited: 66 NOT USED (high confidence) V. F. Hernandes, M. S. Marques, and J. R. Bordin, “Phase classification using neural networks: application to supercooled, polymorphic core-softened mixtures,” Journal of Physics: Condensed Matter. 2021. link Times cited: 4 Abstract: Characterization of phases of soft matter systems is a chall… read moreAbstract: Characterization of phases of soft matter systems is a challenge faced in many physical chemical problems. For polymorphic fluids it is an even greater challenge. Specifically, glass forming fluids, as water, can have, besides solid polymorphism, more than one liquid and glassy phases, and even a liquid–liquid critical point. In this sense, we apply a neural network algorithm to analyze the phase behavior of a mixture of core-softened fluids that interact through the continuous-shouldered well (CSW) potential, which have liquid polymorphism and liquid–liquid critical points, similar to water. We also apply the neural network to mixtures of CSW fluids and core-softened alcohols models. We combine and expand methods based on bond-orientational order parameters to study mixtures, applied to mixtures of hardcore fluids and to supercooled water, to include longer range coordination shells. With this, the trained neural network was able to properly predict the crystalline solid phases, the fluid phases and the amorphous phase for the pure CSW and CSW-alcohols mixtures with high efficiency. More than this, information about the phase populations, obtained from the network approach, can help verify if the phase transition is continuous or discontinuous, and also to interpret how the metastable amorphous region spreads along the stable high density fluid phase. These findings help to understand the behavior of supercooled polymorphic fluids and extend the comprehension of how amphiphilic solutes affect the phases behavior. read less NOT USED (high confidence) Y. Zamora, L. T. Ward, G. Sivaraman, I. T. Foster, and H. Hoffmann, “Proxima: accelerating the integration of machine learning in atomistic simulations,” Proceedings of the ACM International Conference on Supercomputing. 2021. link Times cited: 7 Abstract: Atomistic-scale simulations are prominent scientific applica… read moreAbstract: Atomistic-scale simulations are prominent scientific applications that require the repetitive execution of a computationally expensive routine to calculate a system's potential energy. Prior work shows that these expensive routines can be replaced with a machine-learned surrogate approximation to accelerate the simulation at the expense of the overall accuracy. The exact balance of speed and accuracy depends on the specific configuration of the surrogate-modeling workflow and the science itself, and prior work leaves it up to the scientist to find a configuration that delivers the required accuracy for their science problem. Unfortunately, due to the underlying system dynamics, it is rare that a single surrogate configuration presents an optimal accuracy/latency trade-off for the entire simulation. In practice, scientists must choose conservative configurations so that accuracy is always acceptable, forgoing possible acceleration. As an alternative, we propose Proxima, a systematic and automated method for dynamically tuning a surrogate-modeling configuration in response to real-time feedback from the ongoing simulation. Proxima estimates the uncertainty of applying a surrogate approximation in each step of an iterative simulation. Using this information, the specific surrogate configuration can be adjusted dynamically to ensure maximum speedup while sustaining a required accuracy metric. We evaluate Proxima using a Monte Carlo sampling application and find that Proxima respects a wide range of user-defined accuracy goals while achieving speedups of 1.02--5.5X relative to a standard read less NOT USED (high confidence) X. Chen, I. Fonseca, M. Ravnik, V. Slastikov, C. Zannoni, and A. Zarnescu, “Topics in the mathematical design of materials,” Philosophical Transactions of the Royal Society A. 2021. link Times cited: 1 Abstract: We present a perspective on several current research directi… read moreAbstract: We present a perspective on several current research directions relevant to the mathematical design of new materials. We discuss: (i) design problems for phase-transforming and shape-morphing materials, (ii) epitaxy as an approach of central importance in the design of advanced semiconductor materials, (iii) selected design problems in soft matter, (iv) mathematical problems in magnetic materials, (v) some open problems in liquid crystals and soft materials and (vi) mathematical problems on liquid crystal colloids. The presentation combines topics from soft and hard condensed matter, with specific focus on those design themes where mathematical approaches could possibly lead to exciting progress. This article is part of the theme issue ‘Topics in mathematical design of complex materials’. read less NOT USED (high confidence) J. Allcock et al., “The Prospects of Monte Carlo Antibody Loop Modelling on a Fault-Tolerant Quantum Computer,” Frontiers in Drug Discovery. 2021. link Times cited: 3 Abstract: Quantum computing for the biological sciences is an area of … read moreAbstract: Quantum computing for the biological sciences is an area of rapidly growing interest, but specific industrial applications remain elusive. Quantum Markov chain Monte Carlo has been proposed as a method for accelerating a broad class of computational problems, including problems of pharmaceutical interest. Here we investigate the prospects of quantum advantage via this approach, by applying it to the problem of modelling antibody structure, a crucial task in drug development. To minimize the resources required while maintaining pharmaceutical-level accuracy, we propose a specific encoding of molecular dihedral angles into registers of qubits and a method for implementing, in quantum superposition, a Markov chain Monte Carlo update step based on a classical all-atom force field. We give the first detailed analysis of the resources required to solve a problem of industrial size and relevance and find that, though the time and space requirements of using a quantum computer in this way are considerable, continued technological improvements could bring the required resources within reach in the future. read less NOT USED (high confidence) T. Schlick and S. Portillo‐Ledesma, “Biomolecular modeling thrives in the age of technology,” Nature Computational Science. 2021. link Times cited: 41 NOT USED (high confidence) G. Sivaraman et al., “Experimentally Driven Automated Machine-Learned Interatomic Potential for a Refractory Oxide.,” Physical review letters. 2021. link Times cited: 12 Abstract: Understanding the structure and properties of refractory oxi… read moreAbstract: Understanding the structure and properties of refractory oxides is critical for high temperature applications. In this work, a combined experimental and simulation approach uses an automated closed loop via an active learner, which is initialized by x-ray and neutron diffraction measurements, and sequentially improves a machine-learning model until the experimentally predetermined phase space is covered. A multiphase potential is generated for a canonical example of the archetypal refractory oxide, HfO_{2}, by drawing a minimum number of training configurations from room temperature to the liquid state at ∼2900 °C. The method significantly reduces model development time and human effort. read less NOT USED (high confidence) R. Massafra et al., “A Clinical Decision Support System for Predicting Invasive Breast Cancer Recurrence: Preliminary Results,” Frontiers in Oncology. 2021. link Times cited: 15 Abstract: The mortality associated to breast cancer is in many cases r… read moreAbstract: The mortality associated to breast cancer is in many cases related to metastasization and recurrence. Personalized treatment strategies are critical for the outcomes improvement of BC patients and the Clinical Decision Support Systems can have an important role in medical practice. In this paper, we present the preliminary results of a prediction model of the Breast Cancer Recurrence (BCR) within five and ten years after diagnosis. The main breast cancer-related and treatment-related features of 256 patients referred to Istituto Tumori “Giovanni Paolo II” of Bari (Italy) were used to train machine learning algorithms at the-state-of-the-art. Firstly, we implemented several feature importance techniques and then we evaluated the prediction performances of BCR within 5 and 10 years after the first diagnosis by means different classifiers. By using a small number of features, the models reached highly performing results both with reference to the BCR within 5 years and within 10 years with an accuracy of 77.50% and 80.39% and a sensitivity of 92.31% and 95.83% respectively, in the hold-out sample test. Despite validation studies are needed on larger samples, our results are promising for the development of a reliable prognostic supporting tool for clinicians in the definition of personalized treatment plans. read less NOT USED (high confidence) Y. Mishin, “Machine-Learning Interatomic Potentials for Materials Science,” Electrical Engineering eJournal. 2021. link Times cited: 103 NOT USED (high confidence) D. Schwalbe-Koda, A. R. Tan, and R. Gómez-Bombarelli, “Differentiable sampling of molecular geometries with uncertainty-based adversarial attacks,” Nature Communications. 2021. link Times cited: 36 NOT USED (high confidence) Y.-S. Lin, G. P. P. Pun, and Y. Mishin, “Development of a physically-informed neural network interatomic potential for tantalum,” Computational Materials Science. 2021. link Times cited: 9 NOT USED (high confidence) S. T. Lam, Q. J. Li, R. Ballinger, C. Forsberg, and J. Li, “Modeling LiF and FLiBe Molten Salts with Robust Neural Network Interatomic Potential.,” ACS applied materials & interfaces. 2021. link Times cited: 21 Abstract: Lithium-based molten salts have attracted significant attent… read moreAbstract: Lithium-based molten salts have attracted significant attention due to their applications in energy storage, advanced fission reactors, and fusion devices. Lithium fluorides and particularly 66.6%LiF-33.3%BeF2 (Flibe) are of considerable interest in nuclear systems, as they show an excellent combination of favorable heat transfer, neutron moderation, and transmutation characteristics. For nuclear salts, the range of possible local structures, compositions, and thermodynamic conditions presents significant challenges in atomistic modeling. In this work, we demonstrate that atom-centered neural network interatomic potentials (NNIPs) provide a fast method for performing molecular dynamics of molten salts that is as accurate as ab initio molecular dynamics. For LiF, these potentials are able to accurately reproduce ab initio interactions of dimers, crystalline solids under deformation, crystalline LiF near the melting point, and liquid LiF at high temperatures. For Flibe, NNIPs accurately predict the structures and dynamics at normal operating conditions, high-temperature-pressure conditions, and in the crystalline solid phase. Furthermore, we show that NNIP-based molecular dynamics of molten salts are scalable to reach long time scales (e.g., nanosecond) and large system sizes (e.g., 105 atoms) while maintaining ab initio density functional theory accuracy and providing more than 3 orders of magnitude of computational speedup for calculating structure and transport properties. read less NOT USED (high confidence) B. Mortazavi, B. Javvaji, F. Shojaei, T. Rabczuk, A. Shapeev, and X. Zhuang, “Exceptional piezoelectricity, high thermal conductivity and stiffness and promising photocatalysis in two-dimensional MoSi2N4 family confirmed by first-principles,” Nano Energy. 2020. link Times cited: 208 NOT USED (high confidence) D. Yoo, J. Jung, W. Jeong, and S. Han, “Metadynamics sampling in atomic environment space for collecting training data for machine learning potentials,” npj Computational Materials. 2020. link Times cited: 7 NOT USED (high confidence) P. Liu, C. Verdi, F. Karsai, and G. Kresse, “α−β
phase transition of zirconium predicted by on-the-fly machine-learned force field,” Physical Review Materials. 2020. link Times cited: 8 Abstract: The accurate prediction of solid-solid structural phase tran… read moreAbstract: The accurate prediction of solid-solid structural phase transitions at finite temperature is a challenging task, since the dynamics is so slow that direct simulations of the phase transitions by first-principles (FP) methods are typically not possible. Here, we study the $\alpha$-$\beta$ phase transition of Zr at ambient pressure by means of on-the-fly machine-learned force fields. These are automatically generated during FP molecular dynamics (MD) simulations without the need of human intervention, while retaining almost FP accuracy. Our MD simulations successfully reproduce the first-order displacive nature of the phase transition, which is manifested by an abrupt jump of the volume and a cooperative displacement of atoms at the phase transition temperature. The phase transition is further identified by the simulated X-ray powder diffraction, and the predicted phase transition temperature is in reasonable agreement with experiment. Furthermore, we show that using a singular value decomposition and pseudo inversion of the design matrix generally improves the machine-learned force field compared to the usual inversion of the squared matrix in the regularized Bayesian regression. read less NOT USED (high confidence) M. Yang, L. Bonati, D. Polino, and M. Parrinello, “Using metadynamics to build neural network potentials for reactive events: the case of urea decomposition in water,” arXiv: Chemical Physics. 2020. link Times cited: 51 NOT USED (high confidence) G. Laurens, M. Rabary, J. Lam, D. Peláez, and A. Allouche, “Infrared spectra of neutral polycyclic aromatic hydrocarbons based on machine learning potential energy surface and dipole mapping,” Theoretical Chemistry Accounts. 2020. link Times cited: 3 NOT USED (high confidence) A. Kerr, G. Jose, C. Riggert, and K. Mullen, “Automatic learning of topological phase boundaries.,” Physical review. E. 2020. link Times cited: 5 Abstract: Topological phase transitions, which do not adhere to Landau… read moreAbstract: Topological phase transitions, which do not adhere to Landau's phenomenological model (i.e., a spontaneous symmetry breaking process and vanishing local order parameters), have been actively researched in condensed matter physics. Machine learning of topological phase transitions has generally proved difficult due to the global nature of the topological indices. Only recently has the method of diffusion maps been shown to be effective at identifying changes in topological order. However, previous diffusion map results required adjustments of two hyperparameters: a data length scale and the number of phase boundaries. In this article we introduce a heuristic that requires no such tuning. This heuristic allows computer programs to locate appropriate hyperparameters without user input. We demonstrate this method's efficacy by drawing remarkably accurate phase diagrams in three physical models: the Haldane model of graphene, a generalization of the Su-Schreiffer-Haeger model, and a model for a quantum ring with tunnel junctions. These diagrams are drawn, without human intervention, from a supplied range of model parameters. read less NOT USED (high confidence) P. Gao, J. Zhang, Y. Sun, and J. Yu, “Accurate predictions of aqueous solubility of drug molecules via the multilevel graph convolutional network (MGCN) and SchNet architectures.,” Physical chemistry chemical physics : PCCP. 2020. link Times cited: 15 Abstract: Deep learning based methods have been widely applied to pred… read moreAbstract: Deep learning based methods have been widely applied to predict various kinds of molecular properties in the pharmaceutical industry with increasingly more success. In this study, we propose two novel models for aqueous solubility predictions, based on the Multilevel Graph Convolutional Network (MGCN) and SchNet architectures, respectively. The advantage of the MGCN lies in the fact that it could extract the graph features of the target molecules directly from the (3D) structural information; therefore, it doesn't need to rely on a lot of intra-molecular descriptors to learn the features, which are of significance for accurate predictions of the molecular properties. The SchNet performs well in modelling the interatomic interactions inside a molecule, and such a deep learning architecture is also capable of extracting structural information and further predicting the related properties. The actual accuracy of these two novel approaches was systematically benchmarked with four different independent datasets. We found that both the MGCN and SchNet models performed well for aqueous solubility predictions. In the future, we believe such promising predictive models will be applicable to enhancing the efficiency of the screening, crystallization and delivery of drug molecules, essentially as a useful tool to promote the development of molecular pharmaceutics. read less NOT USED (high confidence) J. Chapman and R. Ramprasad, “Multiscale Modeling of Defect Phenomena in Platinum Using Machine Learning of Force Fields,” JOM. 2020. link Times cited: 5 NOT USED (high confidence) S. Lee et al., “Applying Machine Learning Algorithms to Predict Potential Energies and Atomic Forces during C-H Activation,” Journal of the Korean Physical Society. 2020. link Times cited: 2 Abstract: Molecular dynamics (MD) simulations are useful in understand… read moreAbstract: Molecular dynamics (MD) simulations are useful in understanding the interaction between solid materials and molecules. However, performing MD simulations is possible only when interatomic potentials are available and constructing such interatomic potentials usually requires additional computational work. Recently, generating interatomic potentials was shown to be much easier when machine learning (ML) algorithms were used. In addition, ML algorithms require new descriptors for improved performance. Here, we present an ML approach with several categories of atomic descriptors to predict the parameters necessary for MD simulations, such as the potential energies and the atomic forces. We propose several atomic descriptors based on structural information and find that better descriptors can be generated from eXtreme gradient boosting (XGBoost). Moreover, we observe fewer descriptors that perform better in predicting the potential energies and the forces during methane activation processes on a catalytic Pt(111) surface. These results were consistently observed in two different ML algorithms: fully-connected neural network (FNN) and XGBoost. Taking into account the advantages of FNN and XGBoost, we propose an efficient ML model for estimating potential energies. Our findings will be helpful in developing new ML potentials for long-time MD simulations. read less NOT USED (high confidence) T. W. Ko, J. A. Finkler, S. Goedecker, and J. Behler, “A fourth-generation high-dimensional neural network potential with accurate electrostatics including non-local charge transfer,” Nature Communications. 2020. link Times cited: 177 NOT USED (high confidence) Z. Chen et al., “Direct Prediction of Phonon Density of States With Euclidean Neural Networks,” Advanced Science. 2020. link Times cited: 40 Abstract: Machine learning has demonstrated great power in materials d… read moreAbstract: Machine learning has demonstrated great power in materials design, discovery, and property prediction. However, despite the success of machine learning in predicting discrete properties, challenges remain for continuous property prediction. The challenge is aggravated in crystalline solids due to crystallographic symmetry considerations and data scarcity. Here, the direct prediction of phonon density‐of‐states (DOS) is demonstrated using only atomic species and positions as input. Euclidean neural networks are applied, which by construction are equivariant to 3D rotations, translations, and inversion and thereby capture full crystal symmetry, and achieve high‐quality prediction using a small training set of ≈103 examples with over 64 atom types. The predictive model reproduces key features of experimental data and even generalizes to materials with unseen elements, and is naturally suited to efficiently predict alloy systems without additional computational cost. The potential of the network is demonstrated by predicting a broad number of high phononic specific heat capacity materials. The work indicates an efficient approach to explore materials' phonon structure, and can further enable rapid screening for high‐performance thermal storage materials and phonon‐mediated superconductors. read less NOT USED (high confidence) J. Chapman and R. Ramprasad, “Nanoscale Modeling of Surface Phenomena in Aluminum Using Machine Learning Force Fields,” The Journal of Physical Chemistry C. 2020. link Times cited: 7 Abstract: The study of nano-scale surface phenomena is essential in un… read moreAbstract: The study of nano-scale surface phenomena is essential in understanding the physical processes that aid in technologically relevant applications, such as catalysis, material growth, and failure nuc... read less NOT USED (high confidence) B. Mortazavi, E. Podryabinkin, I. Novikov, T. Rabczuk, X. Zhuang, and A. Shapeev, “Accelerating first-principles estimation of thermal conductivity by machine-learning interatomic potentials: A MTP/ShengBTE solution,” Comput. Phys. Commun. 2020. link Times cited: 90 NOT USED (high confidence) M. Paleico and J. Behler, “A bin and hash method for analyzing reference data and descriptors in machine learning potentials,” Machine Learning: Science and Technology. 2020. link Times cited: 1 Abstract: In recent years the development of machine learning potentia… read moreAbstract: In recent years the development of machine learning potentials (MLPs) has become a very active field of research. Numerous approaches have been proposed, which allow one to perform extended simulations of large systems at a small fraction of the computational costs of electronic structure calculations. The key to the success of modern MLPs is the close-to first principles quality description of the atomic interactions. This accuracy is reached by using very flexible functional forms in combination with high-level reference data from electronic structure calculations. These data sets can include up to hundreds of thousands of structures covering millions of atomic environments to ensure that all relevant features of the potential energy surface are well represented. The handling of such large data sets is nowadays becoming one of the main challenges in the construction of MLPs. In this paper we present a method, the bin-and-hash (BAH) algorithm, to overcome this problem by enabling the efficient identification and comparison of large numbers of multidimensional vectors. Such vectors emerge in multiple contexts in the construction of MLPs. Examples are the comparison of local atomic environments to identify and avoid unnecessary redundant information in the reference data sets that is costly in terms of both the electronic structure calculations as well as the training process, the assessment of the quality of the descriptors used as structural fingerprints in many types of MLPs, and the detection of possibly unreliable data points. The BAH algorithm is illustrated for the example of high-dimensional neural network potentials using atom-centered symmetry functions for the geometrical description of the atomic environments, but the method is general and can be combined with any current type of MLP. read less NOT USED (high confidence) B. Parsaeifard et al., “An assessment of the structural resolution of various fingerprints commonly used in machine learning,” Machine Learning: Science and Technology. 2020. link Times cited: 40 Abstract: Atomic environment fingerprints are widely used in computati… read moreAbstract: Atomic environment fingerprints are widely used in computational materials science, from machine learning potentials to the quantification of similarities between atomic configurations. Many approaches to the construction of such fingerprints, also called structural descriptors, have been proposed. In this work, we compare the performance of fingerprints based on the overlap matrix, the smooth overlap of atomic positions, Behler–Parrinello atom-centered symmetry functions, modified Behler–Parrinello symmetry functions used in the ANI-1ccx potential and the Faber–Christensen–Huang–Lilienfeld fingerprint under various aspects. We study their ability to resolve differences in local environments and in particular examine whether there are certain atomic movements that leave the fingerprints exactly or nearly invariant. For this purpose, we introduce a sensitivity matrix whose eigenvalues quantify the effect of atomic displacement modes on the fingerprint. Further, we check whether these displacements correlate with the variation of localized physical quantities such as forces. Finally, we extend our examination to the correlation between molecular fingerprints obtained from the atomic fingerprints and global quantities of entire molecules. read less NOT USED (high confidence) S. Chibani and F.-X. Coudert, “Machine learning approaches for the prediction of materials properties,” APL Materials. 2020. link Times cited: 92 Abstract: We give here a brief overview of the use of machine learning… read moreAbstract: We give here a brief overview of the use of machine learning (ML) in our field, for chemists and materials scientists with no experience with these techniques. We illustrate the workflow of ML for computational studies of materials, with a specific interest in the prediction of materials properties. We present concisely the fundamental ideas of ML, and for each stage of the workflow, we give examples of the possibilities and questions to be considered in implementing ML-based modeling. read less NOT USED (high confidence) M.-P. V. Christiansen, H. L. Mortensen, S. A. Meldgaard, and B. Hammer, “Gaussian representation for image recognition and reinforcement learning of atomistic structure.,” The Journal of chemical physics. 2020. link Times cited: 11 Abstract: The success of applying machine learning to speed up structu… read moreAbstract: The success of applying machine learning to speed up structure search and improve property prediction in computational chemical physics depends critically on the representation chosen for the atomistic structure. In this work, we investigate how different image representations of two planar atomistic structures (ideal graphene and graphene with a grain boundary region) influence the ability of a reinforcement learning algorithm [the Atomistic Structure Learning Algorithm (ASLA)] to identify the structures from no prior knowledge while interacting with an electronic structure program. Compared to a one-hot encoding, we find a radial Gaussian broadening of the atomic position to be beneficial for the reinforcement learning process, which may even identify the Gaussians with the most favorable broadening hyperparameters during the structural search. Providing further image representations with angular information inspired by the smooth overlap of atomic positions method, however, is not found to cause further speedup of ASLA. read less NOT USED (high confidence) F. Thiemann, P. Rowe, E. A. Müller, and A. Michaelides, “Machine Learning Potential for Hexagonal Boron Nitride Applied to Thermally and Mechanically Induced Rippling,” The Journal of Physical Chemistry C. 2020. link Times cited: 18 Abstract: We introduce an interatomic potential for hexagonal boron ni… read moreAbstract: We introduce an interatomic potential for hexagonal boron nitride (hBN) based on the Gaussian approximation potential (GAP) machine learning methodology. The potential is based on a training set of... read less NOT USED (high confidence) A. S. Christensen and A. V. Lilienfeld, “On the role of gradients for machine learning of molecular energies and forces,” Machine Learning: Science and Technology. 2020. link Times cited: 78 Abstract: The accuracy of any machine learning potential can only be a… read moreAbstract: The accuracy of any machine learning potential can only be as good as the data used in the fitting process. The most efficient model therefore selects the training data that will yield the highest accuracy compared to the cost of obtaining the training data. We investigate the convergence of prediction errors of quantum machine learning models for organic molecules trained on energy and force labels, two common data types in molecular simulations. When training models for the potential energy surface of a single molecule, we find that the inclusion of atomic forces in the training data increases the accuracy of the predicted energies and forces 7-fold, compared to models trained on energy only. Surprisingly, for models trained on sets of organic molecules of varying size and composition in non-equilibrium conformations, inclusion of forces in the training does not improve the predicted energies of unseen molecules in new conformations. Predicted forces, however, improve about 7-fold. For the systems studied, we find that force labels and energy labels contribute equally per label to the convergence of the prediction errors. The optimal choice of what type of training data to include depends on several factors: the computational cost of acquiring the force and energy labels for training, the application domain, the property of interest and the complexity of the machine learning model. Based on our observations we describe key considerations for the creation of new datasets for potential energy surfaces of molecules which maximize the efficiency of the resulting machine learning models. read less NOT USED (high confidence) J. Westermayr and P. Marquetand, “Deep learning for UV absorption spectra with SchNarc: First steps toward transferability in chemical compound space.,” The Journal of chemical physics. 2020. link Times cited: 43 Abstract: Machine learning (ML) has shown to advance the research fiel… read moreAbstract: Machine learning (ML) has shown to advance the research field of quantum chemistry in almost any possible direction and has also recently been applied to investigate the multifaceted photochemistry of molecules. In this paper, we pursue two goals: (i) We show how ML can be used to model permanent dipole moments for excited states and transition dipole moments by adapting the charge model of Gastegger et al. [Chem. Sci. 8, 6924-6935 (2017)], which was originally proposed for the permanent dipole moment vector of the electronic ground state. (ii) We investigate the transferability of our excited-state ML models in chemical space, i.e., whether an ML model can predict the properties of molecules that it has never been trained on and whether it can learn the different excited states of two molecules simultaneously. To this aim, we employ and extend our previously reported SchNarc approach for excited-state ML. We calculate UV absorption spectra from excited-state energies and transition dipole moments as well as electrostatic potentials from latent charges inferred by the ML model of the permanent dipole moment vectors. We train our ML models on CH2NH2 + and C2H4, while predictions are carried out for these molecules and additionally for CHNH2, CH2NH, and C2H5 +. The results indicate that transferability is possible for the excited states. read less NOT USED (high confidence) L. Zhang, M. Chen, X. Wu, H. Wang, W. E, and R. Car, “Deep neural network for the dielectric response of insulators,” Physical Review B. 2020. link Times cited: 45 Abstract: We introduce a deep neural network to model in a symmetry pr… read moreAbstract: We introduce a deep neural network to model in a symmetry preserving way the environmental dependence of the centers of the electronic charge. The model learns from ab-initio density functional theory, wherein the electronic centers are uniquely assigned by the maximally localized Wannier functions. When combined with the Deep Potential model of the atomic potential energy surface, the scheme predicts the dielectric response of insulators for trajectories inaccessible to direct ab-initio simulation. The scheme is non-perturbative and can capture the response of a mutating chemical environment. We demonstrate the approach by calculating the infrared spectra of liquid water at standard conditions, and of ice under extreme pressure, when it transforms from a molecular to an ionic crystal. read less NOT USED (high confidence) M. Paleico and J. Behler, “Global optimization of copper clusters at the ZnO(101¯0) surface using a DFT-based neural network potential and genetic algorithms.,” The Journal of chemical physics. 2020. link Times cited: 28 Abstract: The determination of the most stable structures of metal clu… read moreAbstract: The determination of the most stable structures of metal clusters supported at solid surfaces by computer simulations represents a formidable challenge due to the complexity of the potential-energy surface. Here, we combine a high-dimensional neural network potential, which allows us to predict the energies and forces of a large number of structures with first-principles accuracy, with a global optimization scheme employing genetic algorithms. This very efficient setup is used to identify the global minima and low-energy local minima for a series of copper clusters containing between four and ten atoms adsorbed at the ZnO(101¯0) surface. A series of structures with common structural features resembling the Cu(111) and Cu(110) surfaces at the metal-oxide interface has been identified, and the geometries of the emerging clusters are characterized in detail. We demonstrate that the frequently employed approximation of a frozen substrate surface in global optimization can result in missing the most relevant structures. read less NOT USED (high confidence) M. Eckhoff, F. Schönewald, M. Risch, C. Volkert, P. Blöchl, and J. Behler, “Closing the gap between theory and experiment for lithium manganese oxide spinels using a high-dimensional neural network potential,” Physical Review B. 2020. link Times cited: 17 Abstract: Many positive electrode materials in lithium ion batteries i… read moreAbstract: Many positive electrode materials in lithium ion batteries include transition metals, which are difficult to describe by electronic structure methods like density functional theory (DFT) due to the presence of multiple oxidation states. A prominent example is the lithium manganese oxide spinel ${\mathrm{Li}}_{x}{\mathrm{Mn}}_{2}{\mathrm{O}}_{4}$ with $0\ensuremath{\le}x\ensuremath{\le}2$. While DFT, employing the local hybrid functional PBE0r, provides a reliable description, the need for extended computer simulations of large structural models remains a significant challenge. Here, we close this gap by constructing a DFT-based high-dimensional neural network potential (HDNNP) providing accurate energies and forces at a fraction of the computational costs. As different oxidation states and the resulting Jahn-Teller distortions represent a new level of complexity for HDNNPs, the potential is carefully validated by performing x-ray diffraction experiments. We demonstrate that the HDNNP provides atomic level details and is able to predict a series of properties like the lattice parameters and expansion with increasing Li content or temperature, the orthorhombic to cubic transition, the lithium diffusion barrier, and the phonon frequencies. We show that for understanding these properties access to large time and length scales as enabled by the HDNNP is essential to close the gap between theory and experiment. read less NOT USED (high confidence) D. Morgan and R. Jacobs, “Opportunities and Challenges for Machine Learning in Materials Science,” Annual Review of Materials Research. 2020. link Times cited: 167 Abstract: Advances in machine learning have impacted myriad areas of m… read moreAbstract: Advances in machine learning have impacted myriad areas of materials science, such as the discovery of novel materials and the improvement of molecular simulations, with likely many more important developments to come. Given the rapid changes in this field, it is challenging to understand both the breadth of opportunities and the best practices for their use. In this review, we address aspects of both problems by providing an overview of the areas in which machine learning has recently had significant impact in materials science, and then we provide a more detailed discussion on determining the accuracy and domain of applicability of some common types of machine learning models. Finally, we discuss some opportunities and challenges for the materials community to fully utilize the capabilities of machine learning. read less NOT USED (high confidence) R. Jinnouchi, F. Karsai, C. Verdi, R. Asahi, and G. Kresse, “Descriptors representing two- and three-body atomic distributions and their effects on the accuracy of machine-learned inter-atomic potentials.,” The Journal of chemical physics. 2020. link Times cited: 47 Abstract: When determining machine-learning models for inter-atomic po… read moreAbstract: When determining machine-learning models for inter-atomic potentials, the potential energy surface is often described as a non-linear function of descriptors representing two- and three-body atomic distribution functions. It is not obvious how the choice of the descriptors affects the efficiency of the training and the accuracy of the final machine-learned model. In this work, we formulate an efficient method to calculate descriptors that can separately represent two- and three-body atomic distribution functions, and we examine the effects of including only two- or three-body descriptors, as well as including both, in the regression model. Our study indicates that non-linear mixing of two- and three-body descriptors is essential for an efficient training and a high accuracy of the final machine-learned model. The efficiency can be further improved by weighting the two-body descriptors more strongly. We furthermore examine a sparsification of the three-body descriptors. The three-body descriptors usually provide redundant representations of the atomistic structure, and the number of descriptors can be significantly reduced without loss of accuracy by applying an automatic sparsification using a principal component analysis. Visualization of the reduced descriptors using three-body distribution functions in real-space indicates that the sparsification automatically removes the components that are less significant for describing the distribution function. read less NOT USED (high confidence) C. Lapointe et al., “Machine learning surrogate models for prediction of point defect vibrational entropy,” Physical Review Materials. 2020. link Times cited: 5 Abstract: The temperature variation of the defect densities in a cryst… read moreAbstract: The temperature variation of the defect densities in a crystal depends on vibrational entropy. This contribution to the system thermodynamics remains computationally challenging as it requires a diagonalization of the system's Hessian which scales as $O({N}^{3})$ for a crystal made of $N$ atoms. Here, to circumvent such a heavy computational task and make it feasible even for systems containing millions of atoms, the harmonic vibrational entropy of point defects is estimated directly from the relaxed atomic positions through a linear-in-descriptor machine learning approach of order $O(N)$. With a size-independent descriptor dimension and fixed model parameters, an excellent predictive power is demonstrated on a wide range of defect configurations, supercell sizes, and external deformations well outside the training database. In particular, formation entropies in a range of $250{k}_{B}$ are predicted with less than $1.6{k}_{B}$ error from a training database whose formation entropies span only $25{k}_{B}$ (training error less than $1.0{k}_{B}$). This exceptional transferability is found to hold even when the training is limited to a low-energy superbasin in the phase space while the tests are performed for a different liquid-like superbasin at higher energies. read less NOT USED (high confidence) J. Chapman and R. Ramprasad, “Predicting the dynamic behavior of the mechanical properties of platinum with machine learning.,” The Journal of chemical physics. 2020. link Times cited: 2 Abstract: Over the last few decades, computational tools have been ins… read moreAbstract: Over the last few decades, computational tools have been instrumental in understanding the behavior of materials at the nano-meter length scale. Until recently, these tools have been dominated by two levels of theory: quantum mechanics (QM) based methods and semi-empirical/classical methods. The former are time-intensive but accurate and versatile, while the latter methods are fast but are significantly limited in veracity, versatility, and transferability. Recently, machine learning (ML) methods have shown the potential to bridge the gap between these two chasms due to their (i) low cost, (ii) accuracy, (iii) transferability, and (iv) ability to be iteratively improved. In this work, we further extend the scope of ML for atomistic simulations by capturing the temperature dependence of the mechanical and structural properties of bulk platinum through molecular dynamics simulations. We compare our results directly with experiments, showcasing that ML methods can be used to accurately capture large-scale materials phenomena that are out of reach of QM calculations. We also compare our predictions with those of a reliable embedded atom method potential. We conclude this work by discussing how ML methods can be used to push the boundaries of nano-scale materials research by bridging the gap between QM and experimental methods. read less NOT USED (high confidence) C. Zhai, T. Li, H. Shi, and J. Yeo, “Discovery and design of soft polymeric bio-inspired materials with multiscale simulations and artificial intelligence.,” Journal of materials chemistry. B. 2020. link Times cited: 28 Abstract: Materials chemistry is at the forefront of the global "… read moreAbstract: Materials chemistry is at the forefront of the global "Fourth Industrial Revolution", in part by establishing a "Materials 4.0" paradigm. A key aspect of this paradigm is developing methods to effectively integrate hardware, software, and biological systems. Towards this end, we must have intimate knowledge of the virtual space in materials design: materials omics (materiomics), materials informatics, computational modelling and simulations, artificial intelligence (AI), and big data. We focus on the discovery and design of next-generation bio-inspired materials because the design space is so huge as to be almost intractable. With nature providing researchers with specific guiding principles, this material design space may be probed most efficiently through digital, high-throughput methods. Therefore, to enhance awareness and adoption of digital approaches in soft polymeric bio-inspired materials discovery and design, we detail multiscale simulation techniques in soft matter from the molecular level to the macroscale. We also highlight the unique role that artificial intelligence and materials databases will play in molecular simulations as well as soft materials discovery. Finally, we showcase several case studies that concretely apply computational modelling and simulations for integrative soft bio-inspired materials design with experiments. read less NOT USED (high confidence) M. Raeisi, B. Mortazavi, E. Podryabinkin, F. Shojaei, X. Zhuang, and A. Shapeev, “High thermal conductivity in semiconducting Janus and non-Janus diamanes,” Carbon. 2020. link Times cited: 35 NOT USED (high confidence) R. Hoffmann and J. Malrieu, “Simulation vs Understanding A Tension, in Quantum Chemistry and Beyond. PART B The March of Simulation, for Better or Worse.,” Angewandte Chemie. 2020. link Times cited: 15 Abstract: In the second part of this essay, we leave philosophy, simpl… read moreAbstract: In the second part of this essay, we leave philosophy, simply describing Roald's being trashed by simulation. This leads us to a general sketch of artificial intelligence (AI), Searle's Chinese room, and Strevens' account of what a go-playing program knows. Back to our terrain -- we ask "Quantum Chemistry, † ca. 2020?" Then move to examples of Big Data, machine learning and neural networks in action, first in chemistry and then affecting social matters. trivial to scary. We argue that moral decisions are hardly to be left to a computer. And that causes are so much deeper than correlations. read less NOT USED (high confidence) R. V. Lommel, J. Zhao, W. D. D. Borggraeve, F. de Proft, and M. Alonso, “Molecular dynamics based descriptors for predicting supramolecular gelation,” Chemical Science. 2020. link Times cited: 13 Abstract: Four molecular dynamics-based descriptors were derived able … read moreAbstract: Four molecular dynamics-based descriptors were derived able to classify gelator–solvent combinations as a gel, precipitate or clear solution. read less NOT USED (high confidence) S. Srinivasan et al., “Machine learning the metastable phase diagram of covalently bonded carbon,” Nature Communications. 2020. link Times cited: 7 NOT USED (high confidence) R. Batra and S. Sankaranarayanan, “Machine learning for multi-fidelity scale bridging and dynamical simulations of materials,” Journal of Physics: Materials. 2020. link Times cited: 12 Abstract: Molecular dynamics (MD) is a powerful and popular tool for u… read moreAbstract: Molecular dynamics (MD) is a powerful and popular tool for understanding the dynamical evolution of materials at the nano and mesoscopic scales. There are various flavors of MD ranging from the high fidelity albeit computationally expensive ab initio MD to relatively lower fidelity but much more efficient classical MD such as atomistic and coarse-grained models. Each of these different flavors of MD have been independently used by materials scientists to bring about breakthroughs in materials discovery and design. A significant gulf exists between the various MD flavors, each having varying levels of fidelity. The accuracy of DFT or ab initio MD is generally much higher than that of classical atomistic simulations which is higher than that of coarse-grained models. Multi-fidelity scale bridging to combine the accuracy and flexibility of ab initio MD with efficiency classical MD has been a longstanding goal. The advent of big-data analytics has brought to the forefront powerful machine learning methods that can be deployed to achieve this goal. Here, we provide our perspective on the challenges in multi-fidelity scale bridging and trace the developments leading up to the use of machine learning algorithms and data-science towards addressing this grand challenge. read less NOT USED (high confidence) W. Zhao, Q. Li, X.-H. Huang, L. Bie, and J. Gao, “Toward the Prediction of Multi-Spin State Charges of a Heme Model by Random Forest Regression,” Frontiers in Chemistry. 2020. link Times cited: 1 Abstract: The random forest regression (RFR) model was introduced to p… read moreAbstract: The random forest regression (RFR) model was introduced to predict the multiple spin state charges of a heme model, which is important for the molecular dynamic simulation of the spin crossover phenomenon. In this work, a multiple spin state structure data set with 39,368 structures of the simplified heme–oxygen binding model was built from the non-adiabatic dynamic simulation trajectories. The ESP charges of each atom were calculated and used as the real-valued response. The conformational adapted charge model (CAC) of three spin states was constructed by an RFR model using symmetry functions. The results show that our RFR model can effectively predict the on the fly atomic charges with the varying conformations as well as the atomic charge of different spin states in the same conformation, thus achieving the balance of accuracy and efficiency. The average mean absolute error of the predicted charges of each spin state is <0.02 e. The comparison studies on descriptors showed a maximum 0.06 e improvement in prediction of the charge of Fe2+ by using 11 manually selected structural parameters. We hope that this model can not only provide variable parameters for developing the force field of the multi-spin state but also facilitate automation, thus enabling large-scale simulations of atomistic systems. read less NOT USED (high confidence) K. Momeni et al., “Multiscale computational understanding and growth of 2D materials: a review,” npj Computational Materials. 2020. link Times cited: 85 NOT USED (high confidence) M. Wen and E. Tadmor, “Uncertainty quantification in molecular simulations with dropout neural network potentials,” npj Computational Materials. 2020. link Times cited: 46 NOT USED (high confidence) M. Karabin and D. Perez, “An entropy-maximization approach to automated training set generation for interatomic potentials.,” The Journal of chemical physics. 2020. link Times cited: 17 Abstract: Machine learning-based interatomic potentials are currently … read moreAbstract: Machine learning-based interatomic potentials are currently garnering a lot of attention as they strive to achieve the accuracy of electronic structure methods at the computational cost of empirical potentials. Given their generic functional forms, the transferability of these potentials is highly dependent on the quality of the training set, the generation of which can be highly labor-intensive. Good training sets should at once contain a very diverse set of configurations while avoiding redundancies that incur cost without providing benefits. We formalize these requirements in a local entropy-maximization framework and propose an automated sampling scheme to sample from this objective function. We show that this approach generates much more diverse training sets than unbiased sampling and is competitive with hand-crafted training sets. read less NOT USED (high confidence) T. Mueller, A. Hernandez, and C. Wang, “Machine learning for interatomic potential models.,” The Journal of chemical physics. 2020. link Times cited: 189 Abstract: The use of supervised machine learning to develop fast and a… read moreAbstract: The use of supervised machine learning to develop fast and accurate interatomic potential models is transforming molecular and materials research by greatly accelerating atomic-scale simulations with little loss of accuracy. Three years ago, Jörg Behler published a perspective in this journal providing an overview of some of the leading methods in this field. In this perspective, we provide an updated discussion of recent developments, emerging trends, and promising areas for future research in this field. We include in this discussion an overview of three emerging approaches to developing machine-learned interatomic potential models that have not been extensively discussed in existing reviews: moment tensor potentials, message-passing networks, and symbolic regression. read less NOT USED (high confidence) B. Mortazavi et al., “Exploring phononic properties of two-dimensional materials using machine learning interatomic potentials,” arXiv: Materials Science. 2020. link Times cited: 111 NOT USED (high confidence) R. Biswas, R. Rashmi, and U. Lourderaj, “Machine Learning in Chemical Dynamics,” Resonance. 2020. link Times cited: 5 NOT USED (high confidence) R. Biswas, R. Rashmi, and U. Lourderaj, “Machine Learning in Chemical Dynamics,” Resonance. 2020. link Times cited: 0 NOT USED (high confidence) J. Westermayr, F. A. Faber, A. S. Christensen, O. von Lilienfeld, and P. Marquetand, “Neural networks and kernel ridge regression for excited states dynamics of CH2NH 2+ : From single-state to multi-state representations and multi-property machine learning models,” Machine Learning: Science and Technology. 2019. link Times cited: 34 Abstract: Excited-state dynamics simulations are a powerful tool to in… read moreAbstract: Excited-state dynamics simulations are a powerful tool to investigate photo-induced reactions of molecules and materials and provide complementary information to experiments. Since the applicability of these simulation techniques is limited by the costs of the underlying electronic structure calculations, we develop and assess different machine learning models for this task. The machine learning models are trained on ab initio calculations for excited electronic states, using the methylenimmonium cation (CH2NH 2+ ) as a model system. Two distinct strategies for modeling excited state properties are tested in this work. The first strategy is to treat each state separately in a kernel ridge regression model and all states together in a multiclass neural network. The second strategy is to instead encode the state as input into the model, which is tested with both models. Numerical evidence suggests that using the state as input yields the best performance. An important goal for excited-state machine learning models is their use in dynamics simulations, which needs not only state-specific information but also couplings, i.e. properties involving pairs of states. Accordingly, we investigate how well machine learning models can predict the couplings. Furthermore, we explore how combining all properties in a single neural network affects the accuracy. Finally, machine learning predicted energies, forces, and couplings are used to carry out excited-state dynamics simulations. Results demonstrate the scopes and possibilities of machine learning to model excited-state properties. read less NOT USED (high confidence) T. Cova and A. Pais, “Deep Learning for Deep Chemistry: Optimizing the Prediction of Chemical Patterns,” Frontiers in Chemistry. 2019. link Times cited: 108 Abstract: Computational Chemistry is currently a synergistic assembly … read moreAbstract: Computational Chemistry is currently a synergistic assembly between ab initio calculations, simulation, machine learning (ML) and optimization strategies for describing, solving and predicting chemical data and related phenomena. These include accelerated literature searches, analysis and prediction of physical and quantum chemical properties, transition states, chemical structures, chemical reactions, and also new catalysts and drug candidates. The generalization of scalability to larger chemical problems, rather than specialization, is now the main principle for transforming chemical tasks in multiple fronts, for which systematic and cost-effective solutions have benefited from ML approaches, including those based on deep learning (e.g. quantum chemistry, molecular screening, synthetic route design, catalysis, drug discovery). The latter class of ML algorithms is capable of combining raw input into layers of intermediate features, enabling bench-to-bytes designs with the potential to transform several chemical domains. In this review, the most exciting developments concerning the use of ML in a range of different chemical scenarios are described. A range of different chemical problems and respective rationalization, that have hitherto been inaccessible due to the lack of suitable analysis tools, is thus detailed, evidencing the breadth of potential applications of these emerging multidimensional approaches. Focus is given to the models, algorithms and methods proposed to facilitate research on compound design and synthesis, materials design, prediction of binding, molecular activity, and soft matter behavior. The information produced by pairing Chemistry and ML, through data-driven analyses, neural network predictions and monitoring of chemical systems, allows (i) prompting the ability to understand the complexity of chemical data, (ii) streamlining and designing experiments, (ii) discovering new molecular targets and materials, and also (iv) planning or rethinking forthcoming chemical challenges. In fact, optimization engulfs all these tasks directly. read less NOT USED (high confidence) T. Chen and T. Manz, “A collection of forcefield precursors for metal–organic frameworks,” RSC Advances. 2019. link Times cited: 12 Abstract: A host of important performance properties for metal–organic… read moreAbstract: A host of important performance properties for metal–organic frameworks (MOFs) and other complex materials can be calculated by modeling statistical ensembles. The principle challenge is to develop accurate and computationally efficient interaction models for these simulations. Two major approaches are (i) ab initio molecular dynamics in which the interaction model is provided by an exchange–correlation theory (e.g., DFT + dispersion functional) and (ii) molecular mechanics in which the interaction model is a parameterized classical force field. The first approach requires further development to improve computational speed. The second approach requires further development to automate accurate forcefield parameterization. Because of the extreme chemical diversity across thousands of MOF structures, this problem is still mostly unsolved today. For example, here we show structures in the 2014 CoRE MOF database contain more than 8 thousand different atom types based on first and second neighbors. Our results showed that atom types based on both first and second neighbors adequately capture the chemical environment, but atom types based on only first neighbors do not. For 3056 MOFs, we used density functional theory (DFT) followed by DDEC6 atomic population analysis to extract a host of important forcefield precursors: partial atomic charges; atom-in-material (AIM) C6, C8, and C10 dispersion coefficients; AIM dipole and quadrupole moments; various AIM polarizabilities; quantum Drude oscillator parameters; AIM electron cloud parameters; etc. Electrostatic parameters were validated through comparisons to the DFT-computed electrostatic potential. These forcefield precursors should find widespread applications to developing MOF force fields. read less NOT USED (high confidence) C. Lu et al., “Deep Learning for Optoelectronic Properties of Organic Semiconductors,” The Journal of Physical Chemistry C. 2019. link Times cited: 34 Abstract: Atomistic modeling of energetic disorder in organic semicond… read moreAbstract: Atomistic modeling of energetic disorder in organic semiconductors (OSCs) and its effects on the optoelectronic properties of OSCs requires a large number of excited-state electronic-structure calculations, a computationally daunting task for many OSC applications. In this work, we advocate the use of deep learning to address this challenge and demonstrate that state-of-the-art deep neural networks (DNNs) are capable of predicting the electronic properties of OSCs at an accuracy comparable with the quantum chemistry methods used for generating training data. We extensively investigate the performances of four recent DNNs (deep tensor neural network, SchNet, message passing neural network, and multilevel graph convolutional neural network) in predicting various electronic properties of an important class of OSCs, i.e., oligothiophenes (OTs), including their HOMO and LUMO energies, excited-state energies and associated transition dipole moments. We find that SchNet shows the best performance for OTs of different sizes (from bithiophene to sexithiophene), achieving average prediction errors in the range of 20-80meV compared to the results from (time-dependent) density functional theory. We show that SchNet also consistently outperforms shallow feed-forward neural networks, especially in difficult cases with large molecules or limited training data. We further show that SchNet could predict the transition dipole moment accurately, a task previously known to be difficult for feed-forward neural networks, and we ascribe the relatively large errors in transition dipole prediction seen for some OT configurations to the charge-transfer character of their excited states. Finally, we demonstrate the effectiveness of SchNet by modeling the UV-Vis absorption spectra of OTs in dichloromethane and a good agreement is observed between the calculated and experimental spectra. read less NOT USED (high confidence) J. Caceres-Delpiano, L. P. Wang, and J. Essex, “The automated optimisation of a coarse-grained force field using free energy data,” Physical Chemistry Chemical Physics. 2019. link Times cited: 3 Abstract: Atomistic models provide a detailed representation of molecu… read moreAbstract: Atomistic models provide a detailed representation of molecular systems, but are sometimes inadequate for simulations of large systems over long timescales. Coarse-grained models enable accelerated simulations by reducing the number of degrees of freedom, at the cost of reduced accuracy. New optimisation processes to parameterise these models could improve their quality and range of applicability. We present an automated approach for the optimisation of coarse-grained force fields, by reproducing free energy data derived from atomistic molecular simulations. To illustrate the approach, we implemented hydration free energy gradients as a new target for force field optimisation in ForceBalance and applied it successfully to optimise the un-charged side-chains and the protein backbone in the SIRAH protein coarse-grain force field. The optimised parameters closely reproduced hydration free energies of atomistic models and gave improved agreement with experiment. read less NOT USED (high confidence) A. Bochkarev, A. Roekeghem, S. Mossa, and N. Mingo, “Anharmonic thermodynamics of vacancies using a neural network potential,” Physical Review Materials. 2019. link Times cited: 11 Abstract: Lattice anharmonicity is thought to strongly affect vacancy … read moreAbstract: Lattice anharmonicity is thought to strongly affect vacancy concentrations in metals at high temperatures. It is however non-trivial to account for this effect directly using density functional theory (DFT). Here we develop a deep neural network potential for aluminum that overcomes the limitations inherent to DFT, and we use it to obtain accurate anharmonic vacancy formation free energies as a function of temperature. While confirming the important role of anharmonicity at high temperatures, the calculation unveils a markedly nonlinear behavior of the vacancy formation entropy and shows that the vacancy formation free energy only violates Arrhenius law at temperatures above 600 K, in contrast with previous DFT calculations. read less NOT USED (high confidence) A. Bhowmik, I. Castelli, J. M. García‐Lastra, P. B. Jørgensen, O. Winther, and T. Vegge, “A perspective on inverse design of battery interphases using multi-scale modelling, experiments and generative deep learning,” Energy Storage Materials. 2019. link Times cited: 67 NOT USED (high confidence) R. Vasudevan et al., “Materials science in the artificial intelligence age: high-throughput library generation, machine learning, and a pathway from correlations to the underpinning physics,” MRS Communications. 2019. link Times cited: 14 Abstract: The use of statistical/machine learning (ML) approaches to m… read moreAbstract: The use of statistical/machine learning (ML) approaches to materials science is experiencing explosive growth. Here, we review recent work focusing on the generation and application of libraries from both experiment and theoretical tools. The library data enables classical correlative ML and also opens the pathway for exploration of underlying causative physical behaviors. We highlight key advances facilitated by this approach and illustrate how modeling, macroscopic experiments, and imaging can be combined to accelerate the understanding and development of new materials systems. These developments point toward a data-driven future wherein knowledge can be aggregated and synthesized, accelerating the advancement of materials science. read less NOT USED (high confidence) P. Schlexer Lamoureux et al., “Machine Learning for Computational Heterogeneous Catalysis,” ChemCatChem. 2019. link Times cited: 179 Abstract: Big data and artificial intelligence has revolutionized scie… read moreAbstract: Big data and artificial intelligence has revolutionized science in almost every field – from economics to physics. In the area of materials science and computational heterogeneous catalysis, this revolution has led to the development of scientific data repositories, as well as data mining and machine learning tools to investigate the vast materials space. The goal of using these tools is to establish a deeper understanding of the relations between materials properties and activity, selectivity and stability – the important figures of merit in catalysis. Based on these insights, catalyst design principles can be established, which hopefully lead us to discover highly efficient catalysts to solve pressing issues for a sustainable future and the synthesis of highly functional materials, chemicals and pharmaceuticals. The inherent complexity of catalytic reactions quests for machine learning methods to efficiently navigate through the high‐dimensional hyper‐surfaces in structure optimization problems to determine relevant chemical structures and transition states. In this review, we show how cutting edge data infrastructures and machine learning methods are being used to address problems in computational heterogeneous catalysis. read less NOT USED (high confidence) A. K. Pervaje, C. C. Walker, and E. Santiso, “Molecular simulation of polymers with a SAFT-γ Mie approach,” Molecular Simulation. 2019. link Times cited: 8 Abstract: ABSTRACT We review the group contribution Statistical Associ… read moreAbstract: ABSTRACT We review the group contribution Statistical Associating Fluid Theory with Mie interaction potentials (SAFT-γ Mie) approach for building coarse-grained models for molecular simulation of polymeric systems. In this top-down method, force field parameters for coarse-grained polymer models can be derived from thermodynamic information on constituent monomer units using the SAFT-γ Mie equation of state (EoS). This strategy can facilitate high-throughput computational screening of polymeric materials, with a corresponding states correlation expediting the force field fitting. Accurate and transferable non-bonded parameters linked to macroscopic thermodynamic data allow for calculation of properties beyond those obtainable from the EoS alone. To overcome limitations of SAFT-γ Mie regarding polymer chain stiffness and branching, hybrid top-down/bottom-up approaches have combined non-bonded parameters from SAFT-γ Mie with bond-stretching and angle-bending potentials from higher-resolution force fields. Our review critically evaluates the performance of recent SAFT-γ Mie polymer models, highlighting the strengths and weaknesses in the context of other equation of state and coarse-graining methods. read less NOT USED (high confidence) J. C. Thomas, J. S. Bechtel, A. Natarajan, and A. V. der Ven, “Machine learning the density functional theory potential energy surface for the inorganic halide perovskite
CsPbBr3,” Physical Review B. 2019. link Times cited: 7 Abstract: Structural phase transitions as a function of temperature di… read moreAbstract: Structural phase transitions as a function of temperature dictate the structure--functionality relationships in many technologically important materials. Harmonic Hamiltonians have proven successful in predicting the vibrational properties of many materials. However, they are inadequate for modeling structural phase transitions in crystals with potential energy surfaces that are either strongly anharmonic or no\ n-convex with respect to collective atomic displacements or homogeneous strains. In this paper we develop a framework to express highly anharmonic first-principles potential energy surfaces as polynomials of collective cluster deformati\ ons. We further adapt the approach to a nonlinear extension of the cluster expansion formalism through the use of an artificial neural net model. The machine learning models are trained on a large database of first-principles calculations and are shown to reproduce the potential energy surface with l\ ow error. read less NOT USED (high confidence) S. Chhabra, J. Xie, and A. Frank, “RNAPosers: Machine Learning Classifiers For RNA-Ligand Poses,” bioRxiv. 2019. link Times cited: 0 Abstract: Determining the 3-dimensional (3D) structures of ribonucleic… read moreAbstract: Determining the 3-dimensional (3D) structures of ribonucleic acid (RNA)-small molecule complexes is critical to understanding molecular recognition in RNA. Computer docking can, in principle, be used to predict the 3D structure of RNA-small molecule complexes. Unfortunately, retrospective analysis has shown that the scoring functions that are typically used to rank poses tend to misclassify non-native poses as native, and vice versa. This misclassification of non-native poses severely limits the utility of computer docking in the context pose prediction, as well as in virtual screening. Here, we use machine learning to train a set of pose classifiers that estimate the relative “nativeness” of a set of RNA-ligand poses. At the heart of our approach is the use of a pose “fingerprint” that is a composite of a set of atomic fingerprints, which individually encode the local “RNA environment” around ligand atoms. We found that by ranking poses based on the classification scores from our machine learning classifiers, we were able to recover native-like poses better than when we ranked poses based on their docking scores. With a leave-one-out training and testing approach, we found that one of our classifiers could recover poses that were within 2.5 Å of the native poses in ∼80% of the 88 cases we examined, and similarly, on a separate validation set, we could recover such poses in ∼70% of the cases. Our set of classifiers, which we refer to as RNAPosers, should find utility as a tool to aid in RNA-ligand pose prediction and so we make RNAPosers open to the academic community via https://github.com/atfrank/RNAPosers. read less NOT USED (high confidence) G. Schmitz, I. H. Godtliebsen, and O. Christiansen, “Machine learning for potential energy surfaces: An extensive database and assessment of methods.,” The Journal of chemical physics. 2019. link Times cited: 42 Abstract: On the basis of a new extensive database constructed for the… read moreAbstract: On the basis of a new extensive database constructed for the purpose, we assess various Machine Learning (ML) algorithms to predict energies in the framework of potential energy surface (PES) construction and discuss black box character, robustness, and efficiency. The database for training ML algorithms in energy predictions based on the molecular structure contains SCF, RI-MP2, RI-MP2-F12, and CCSD(F12*)(T) data for around 10.5 × 106 configurations of 15 small molecules. The electronic energies as function of molecular structure are computed from both static and iteratively refined grids in the context of automized PES construction for anharmonic vibrational computations within the n-mode expansion. We explore the performance of a range of algorithms including Gaussian Process Regression (GPR), Kernel Ridge Regression, Support Vector Regression, and Neural Networks (NNs). We also explore methods related to GPR such as sparse Gaussian Process Regression, Gaussian process Markov Chains, and Sparse Gaussian Process Markov Chains. For NNs, we report some explorations of architecture, activation functions, and numerical settings. Different delta-learning strategies are considered, and the use of delta learning targeting CCSD(F12*)(T) predictions using, for example, RI-MP2 combined with machine learned CCSD(F12*)(T)-RI-MP2 differences is found to be an attractive option. read less NOT USED (high confidence) T. Manz, T. Chen, D. Cole, N. G. Limas, and B. Fiszbein, “New scaling relations to compute atom-in-material polarizabilities and dispersion coefficients: part 1. Theory and accuracy,” RSC Advances. 2019. link Times cited: 16 Abstract: Polarizabilities and London dispersion forces are important … read moreAbstract: Polarizabilities and London dispersion forces are important to many chemical processes. Force fields for classical atomistic simulations can be constructed using atom-in-material polarizabilities and Cn (n = 6, 8, 9, 10…) dispersion coefficients. This article addresses the key question of how to efficiently assign these parameters to constituent atoms in a material so that properties of the whole material are better reproduced. We develop a new set of scaling laws and computational algorithms (called MCLF) to do this in an accurate and computationally efficient manner across diverse material types. We introduce a conduction limit upper bound and m-scaling to describe the different behaviors of surface and buried atoms. We validate MCLF by comparing results to high-level benchmarks for isolated neutral and charged atoms, diverse diatomic molecules, various polyatomic molecules (e.g., polyacenes, fullerenes, and small organic and inorganic molecules), and dense solids (including metallic, covalent, and ionic). We also present results for the HIV reverse transcriptase enzyme complexed with an inhibitor molecule. MCLF provides the non-directionally screened polarizabilities required to construct force fields, the directionally-screened static polarizability tensor components and eigenvalues, and environmentally screened C6 coefficients. Overall, MCLF has improved accuracy compared to the TS-SCS method. For TS-SCS, we compared charge partitioning methods and show DDEC6 partitioning yields more accurate results than Hirshfeld partitioning. MCLF also gives approximations for C8, C9, and C10 dispersion coefficients and quantum Drude oscillator parameters. This method should find widespread applications to parameterize classical force fields and density functional theory (DFT) + dispersion methods. read less NOT USED (high confidence) R. Batra et al., “General Atomic Neighborhood Fingerprint for Machine Learning-Based Methods,” The Journal of Physical Chemistry C. 2019. link Times cited: 34 Abstract: To facilitate chemical space exploration for material screen… read moreAbstract: To facilitate chemical space exploration for material screening or to accelerate computationally expensive first-principles simulations, inexpensive surrogate models that capture electronic, atomistic, or macroscopic materials properties have become an increasingly popular tool over the last decade. The most fundamental quantity common across all such machine learning (ML)-based methods is the fingerprint used to numerically represent a material or its structure. To increase the learning capability of the ML methods, the common practice is to construct fingerprints that satisfy the same symmetry relations as displayed by the target material property of interest (for which the ML model is being developed). Thus, in this work, we present a general, simple, and elegant fingerprint that can be used to learn different electronic/atomistic/structural properties, irrespective of their scalar, vector, or tensorial nature. This fingerprint is based on the concept of multipole terms and can be systematically increa... read less NOT USED (high confidence) N. E. R. Zimmermann and A. Jain, “Local structure order parameters and site fingerprints for quantification of coordination environment and crystal structure similarity,” RSC Advances. 2019. link Times cited: 49 Abstract: Structure characterization and classification is frequently … read moreAbstract: Structure characterization and classification is frequently based on local environment information of all or selected atomic sites in the crystal structure. Therefore, reliable and robust procedures to find coordinated neighbors and to evaluate the resulting coordination pattern (e.g., tetrahedral, square planar) are critically important for both traditional and machine learning approaches that aim to exploit site or structure information for predicting materials properties. Here, we introduce new local structure order parameters (LoStOPs) that are specifically designed to rapidly detect highly symmetric local coordination environments (e.g., Platonic solids such as a tetrahedron or an octahedron) as well as less symmetric ones (e.g., Johnson solids such as a square pyramid). Furthermore, we introduce a Monte Carlo optimization approach to ensure that the different LoStOPs are comparable with each other. We then apply the new local environment descriptors to define site and structure fingerprints and to measure similarity between 61 known coordination environments and 40 commonly studied crystal structures, respectively. After extensive testing and optimization, we determine the most accurate structure similarity assessment procedure to compute all 2.45 billion structure similarities between each pair of the ≈70 000 materials that are currently present in the Materials Project database. read less NOT USED (high confidence) Y. Okamoto, “Data sampling scheme for reproducing energies along reaction coordinates in high-dimensional neural network potentials.,” The Journal of chemical physics. 2019. link Times cited: 3 Abstract: We propose a data sampling scheme for high-dimensional neura… read moreAbstract: We propose a data sampling scheme for high-dimensional neural network potentials that can predict energies along the reaction pathway calculated using the hybrid density functional theory. We observed that a data sampling scheme that combined partial geometry optimization of intermediate structures with random displacement of atoms successfully predicted the energies along the reaction path with respect to five chemical reactions: Claisen rearrangement, Diels-Alder reaction, [1,5]-sigmatropic hydrogen shift, concerted hydrogen transfer in the water hexamer, and Cornforth rearrangement. read less NOT USED (high confidence) J. Vandermause et al., “On-the-fly active learning of interpretable Bayesian force fields for atomistic rare events,” npj Computational Materials. 2019. link Times cited: 189 NOT USED (high confidence) Z. Aitken, V. Sorkin, and Y.-W. Zhang, “Atomistic modeling of nanoscale plasticity in high-entropy alloys,” Journal of Materials Research. 2019. link Times cited: 32 Abstract: Lattice structures, defect structures, and deformation mecha… read moreAbstract: Lattice structures, defect structures, and deformation mechanisms of high-entropy alloys (HEAs) have been studied using atomistic simulations to explain their remarkable mechanical properties. These atomistic simulation techniques, such as first-principles calculations and molecular dynamics allow atomistic-level resolution of structure, defect configuration, and energetics. Following the structure–property paradigm, such understandings can be useful for guiding the design of high-performance HEAs. Although there have been a number of atomistic studies on HEAs, there is no comprehensive review on the state-of-the-art techniques and results of atomistic simulations of HEAs. This article is intended to fill the gap, providing an overview of the state-of-the-art atomistic simulations on HEAs. In particular, we discuss how atomistic simulations can elucidate the nanoscale mechanisms of plasticity underlying the outstanding properties of HEAs, and further present a list of interesting problems for forthcoming atomistic simulations of HEAs. read less NOT USED (high confidence) M. Wood, M. Cusentino, B. Wirth, and A. Thompson, “Data-driven material models for atomistic simulation,” Physical Review B. 2019. link Times cited: 37 Abstract: The central approximation made in classical molecular dynami… read moreAbstract: The central approximation made in classical molecular dynamics simulation of materials is the interatomic potential used to calculate the forces on the atoms. Great effort and ingenuity is required to construct viable functional forms and find accurate parametrizations for potentials using traditional approaches. Machine learning has emerged as an effective alternative approach to develop accurate and robust interatomic potentials. Starting with a very general model form, the potential is learned directly from a database of electronic structure calculations and therefore can be viewed as a multiscale link between quantum and classical atomistic simulations. Risk of inaccurate extrapolation exists outside the narrow range of time and length scales where the two methods can be directly compared. In this work, we use the spectral neighbor analysis potential (SNAP) and show how a fit can be produced with minimal interpolation errors which is also robust in extrapolating beyond training. To demonstrate the method, we have developed a tungsten-beryllium potential suitable for the full range of binary compositions. Subsequently, large-scale molecular dynamics simulations were performed of high energy Be atom implantation onto the (001) surface of solid tungsten. The machine learned W-Be potential generates a population of implantation structures consistent with quantum calculations of defect formation energies. A very shallow ($l2\phantom{\rule{0.16em}{0ex}}\mathrm{nm}$) average Be implantation depth is predicted which may explain ITER diverter degradation in the presence of beryllium. read less NOT USED (high confidence) A. Jha, A. Chandrasekaran, C. Kim, and R. Ramprasad, “Impact of dataset uncertainties on machine learning model predictions: the example of polymer glass transition temperatures,” Modelling and Simulation in Materials Science and Engineering. 2019. link Times cited: 50 Abstract: Over the past decade, there has been a resurgence in the imp… read moreAbstract: Over the past decade, there has been a resurgence in the importance of data-driven techniques in materials science and engineering. The utilization of state-of-the art algorithms, coupled with the increased availability of experimental and computational data, has led to the development of surrogate models offering the promise of rapid and accurate predictions of materials’ properties based solely on their structure or composition. Such machine learning (ML) models are trained on available past data and are thus susceptible to the intrinsic uncertainties/errors associate with these past measurements. The glass transition temperature (Tg) of polymers, a property of paramount interest in polymer science, is one strong example of a material property that can show widespread variation in the final reported value as a result of a variety of intrinsic and extrinsic factors that occur during the experimental measurement process. In the current work, we curate a large database of Tg measurements from a variety of data sources and proceed to investigate the statistical nature of the inherent uncertainties in the database. Through the partitioning of the dataset using statistically relevant measures, we investigate the effect of variations in the dataset on the performance of the final ML model. We demonstrate that the measure of central tendency, median is a valid approximation when dealing with multiple reported values for Tg when dealing with multiple reported values of Tg for the same polymeric material. Moreover, the Bayesian model noise/uncertainty that emerges from our machine-learning pipeline is able to represent quantitatively the underlying noise/uncertainties in the experimental measurement of Tg. read less NOT USED (high confidence) E. Iype and S. Urolagin, “Machine learning model for non-equilibrium structures and energies of simple molecules.,” The Journal of chemical physics. 2019. link Times cited: 8 Abstract: Predicting molecular properties using a Machine Learning (ML… read moreAbstract: Predicting molecular properties using a Machine Learning (ML) method is gaining interest among research as it offers quantum chemical accuracy at molecular mechanics speed. This prediction is performed by training an ML model using a set of reference data [mostly Density Functional Theory (DFT)] and then using it to predict properties. In this work, kernel based ML models are trained (using Bag of Bonds as well as many body tensor representation) against datasets containing non-equilibrium structures of six molecules (water, methane, ethane, propane, butane, and pentane) to predict their atomization energies and to perform a Metropolis Monte Carlo (MMC) run with simulated annealing to optimize molecular structures. The optimized structures and energies of the molecules are found to be comparable with DFT optimized structures, energies, and forces. Thus, this method offers the possibility to use a trained ML model to perform a classical simulation such as MMC without using any force field, thereby improving the accuracy of the simulation at low computational cost. read less NOT USED (high confidence) J. Westermayr et al., “Machine learning enables long time scale molecular photodynamics simulations† †Electronic supplementary information (ESI) available. See DOI: 10.1039/c9sc01742a,” Chemical Science. 2018. link Times cited: 44 Abstract: Machine learning enables excited-state molecular dynamics si… read moreAbstract: Machine learning enables excited-state molecular dynamics simulations including nonadiabatic couplings on nanosecond time scales. read less NOT USED (high confidence) B. Xu and C. Nan, “Modeling and predicting responses of magnetoelectric materials,” MRS Bulletin. 2018. link Times cited: 2 Abstract: Magnetoelectric (ME) materials exhibit cross-coupling effect… read moreAbstract: Magnetoelectric (ME) materials exhibit cross-coupling effects between magnetization and polarization, by which one can manipulate the magnetization (or polarization) with an electric (or magnetic) field. To better understand the responses of ME materials and the coupling mechanisms involved, various simulation methods at different scales, ranging from electronic and atomic scale to the mesoscale, have been developed in the past decades. In this article, we summarize recent progress in modeling and predicting responses of ME materials, and present our perspectives on key issues that require further study, including multiscale simulation methods and approaches dealing with dynamic processes. The simulation methods have the potential to illuminate the dynamic processes in ME materials and device response to external fields and eventually be used for guidance for the data-driven computational design of new ME materials and devices. read less NOT USED (high confidence) S. Jindal and S. Bulusu, “A transferable artificial neural network model for atomic forces in nanoparticles.,” The Journal of chemical physics. 2018. link Times cited: 10 Abstract: We have designed a new method to fit the energy and atomic f… read moreAbstract: We have designed a new method to fit the energy and atomic forces using a single artificial neural network (SANN) for any number of chemical species present in a molecular system. The traditional approach for fitting the potential energy surface for a multicomponent system using artificial neural network (ANN) is to consider n number of networks for n number of chemical species in the system. This shoots the computational cost and makes it difficult to apply to a system containing more number of species. We present a new strategy of using a SANN to compute energy and forces of a chemical system. Since atomic forces are significant for geometry optimizations and molecular dynamics simulations for any chemical system, their accurate prediction is of utmost importance. So, to predict the atomic forces, we have modified the traditional way of fitting forces from underlying energy expression. We have applied our strategy to study geometry optimizations and dynamics in gold-silver nanoalloys and thiol protected gold nanoclusters. Also, force fitting has made it possible to train smaller sized systems and extrapolate the parameters to make accurate predictions for larger systems. This proposed strategy has definitely made the mapping and fitting of atomic forces easier and can be applied to a wide variety of molecular systems. read less NOT USED (high confidence) J. R. Mohallem, P. F. G. Velloso, and A. Arapiraca, “Probing molecular environments with a fictitious isotopic dipole,” International Journal of Quantum Chemistry. 2018. link Times cited: 1 Abstract: A HD-like isotopic dipole moment is proposed as a sensible p… read moreAbstract: A HD-like isotopic dipole moment is proposed as a sensible probe for molecular environments, in particular for electrostatic fields and polarizable (reactive) sites of molecules. Fictitious nuclear masses are chosen in order to yield a rigid dipole with appropriate magnitude. Upon subtracting the Born-Oppenheimer energy, the interaction is reduced to the Field-dipole-like and the dipole-polarizability-like terms, the last one being particularly informative since connected to potentially reactive sites. The Field strength and orientation are easily obtained by identifying the minimum Field-dipole energy configuration and flipping the dipole from it. In this case the method appears to have a superior accuracy in comparison with ab initio approaches. In tests with hydrogen, water, benzene and chlorobenzene molecules and with a frustrated Lewis pair, the potential of the method is assessed. read less NOT USED (high confidence) C. Zanette et al., “Toward Learned Chemical Perception of Force Field Typing Rules.,” Journal of chemical theory and computation. 2018. link Times cited: 34 Abstract: Molecular mechanics force fields define how the energy and f… read moreAbstract: Molecular mechanics force fields define how the energy and forces in a molecular system are computed from its atomic positions, thus enabling the study of such systems through computational methods like molecular dynamics and Monte Carlo simulations. Despite progress toward automated force field parametrization, considerable human expertise is required to develop or extend force fields. In particular, human input has long been required to define atom types, which encode chemically unique environments that determine which parameters will be assigned. However, relying on humans to establish atom types is suboptimal. Human-created atom types are often developed without statistical justification, leading to over- or under-fitting of data. Human-created types are also difficult to extend in a systematic and consistent manner when new chemistries must be modeled or new data becomes available. Finally, human effort is not scalable when force fields must be generated for new (bio)polymers, compound classes, or materials. To remedy these deficiencies, our long-term goal is to replace human specification of atom types with an automated approach, based on rigorous statistics and driven by experimental and/or quantum chemical reference data. In this work, we describe novel methods that automate the discovery of appropriate chemical perception: SMARTY allows for the creation of atom types, while SMIRKY goes further by automating the creation of fragment (nonbonded, bonds, angles, and torsions) types. These approaches enable the creation of move sets in atom or fragment type space, which are used within a Monte Carlo optimization approach. We demonstrate the power of these new methods by automating the rediscovery of human defined atom types (SMARTY) or fragment types (SMIRKY) in existing small molecule force fields. We assess these approaches using several molecular data sets, including one which covers a diverse subset of the DrugBank database. read less NOT USED (high confidence) T. Patra et al., “Defect Dynamics in 2-D MoS2 Probed by Using Machine Learning, Atomistic Simulations, and High-Resolution Microscopy.,” ACS nano. 2018. link Times cited: 63 Abstract: Structural defects govern various physical, chemical, and op… read moreAbstract: Structural defects govern various physical, chemical, and optoelectronic properties of two-dimensional transition-metal dichalcogenides (TMDs). A fundamental understanding of the spatial distribution and dynamics of defects in these low-dimensional systems is critical for advances in nanotechnology. However, such understanding has remained elusive primarily due to the inaccessibility of (a) necessary time scales via standard atomistic simulations and (b) required spatiotemporal resolution in experiments. Here, we take advantage of supervised machine learning, in situ high-resolution transmission electron microscopy (HRTEM) and molecular dynamics (MD) simulations to overcome these limitations. We combine genetic algorithms (GA) with MD to investigate the extended structure of point defects, their dynamical evolution, and their role in inducing the phase transition between the semiconducting (2H) and metallic (1T) phase in monolayer MoS2. GA-based structural optimization is used to identify the long-range structure of randomly distributed point defects (sulfur vacancies) for various defect densities. Regardless of the density, we find that organization of sulfur vacancies into extended lines is the most energetically favorable. HRTEM validates these findings and suggests a phase transformation from the 2H-to-1T phase that is localized near these extended defects when exposed to high electron beam doses. MD simulations elucidate the molecular mechanism driving the onset of the 2H to 1T transformation and indicate that finite amounts of 1T phase can be retained by increasing the defect concentration and temperature. This work significantly advances the current understanding of defect structure/evolution and structural transitions in 2D TMDs, which is crucial for designing nanoscale devices with desired functionality. read less NOT USED (high confidence) S. A. Meldgaard, E. L. Kolsbjerg, and B. Hammer, “Machine learning enhanced global optimization by clustering local environments to enable bundled atomic energies.,” The Journal of chemical physics. 2018. link Times cited: 31 Abstract: We show how to speed up global optimization of molecular str… read moreAbstract: We show how to speed up global optimization of molecular structures using machine learning methods. To represent the molecular structures, we introduce the auto-bag feature vector that combines (i) a local feature vector for each atom, (ii) an unsupervised clustering of such feature vectors for many atoms across several structures, and (iii) a count for a given structure of how many times each cluster is represented. During subsequent global optimization searches, accumulated structure-energy relations of relaxed structural candidates are used to assign local energies to each atom using supervised learning. Specifically, the local energies follow from assigning energies to each cluster of local feature vectors and demanding the sum of local energies to amount to the structural energies in the least squares sense. The usefulness of the method is demonstrated in basin hopping searches for 19-atom structures described by single- or double-well Lennard-Jones type potentials and for 24-atom carbon structures described by density functional theory. In all cases, utilizing the local energy information derived on-the-fly enhances the rate at which the global minimum energy structure is found. read less NOT USED (high confidence) T. Xie and J. Grossman, “Hierarchical Visualization of Materials Space with Graph Convolutional Neural Networks,” The Journal of chemical physics. 2018. link Times cited: 52 Abstract: The combination of high throughput computation and machine l… read moreAbstract: The combination of high throughput computation and machine learning has led to a new paradigm in materials design by allowing for the direct screening of vast portions of structural, chemical, and property spaces. The use of these powerful techniques leads to the generation of enormous amounts of data, which in turn calls for new techniques to efficiently explore and visualize the materials space to help identify underlying patterns. In this work, we develop a unified framework to hierarchically visualize the compositional and structural similarities between materials in an arbitrary material space with representations learned from different layers of graph convolutional neural networks. We demonstrate the potential for such a visualization approach by showing that patterns emerge automatically that reflect similarities at different scales in three representative classes of materials: perovskites, elemental boron, and general inorganic crystals, covering material spaces of different compositions, structures, and both. For perovskites, elemental similarities are learned that reflects multiple aspects of atom properties. For elemental boron, structural motifs emerge automatically showing characteristic boron local environments. For inorganic crystals, the similarity and stability of local coordination environments are shown combining different center and neighbor atoms. The method could help transition to a data-centered exploration of materials space in automated materials design. read less NOT USED (high confidence) B. Goldsmith, J. Esterhuizen, J.-X. Liu, C. J. Bartel, and C. Sutton, “Machine learning for heterogeneous catalyst design and discovery,” Aiche Journal. 2018. link Times cited: 233 Abstract: A dvances in machine learning (ML) are making a large impact… read moreAbstract: A dvances in machine learning (ML) are making a large impact in many fields, including: artificial intelligence, materials science, and chemical engineering. Generally, ML tools learn from data to find insights or make fast predictions of target properties. Recently, ML is also greatly influencing heterogeneous catalysis research due to the availability of ML (e.g., Python Scikit-learn, TensorFlow) and workflow management tools (e.g., ASE, Atomate), the growing amount of data in materials databases (e.g., Novel Materials Discovery Laboratory, Citrination, Materials Project, CatApp), and algorithmic improvements. New catalysts are needed for sustainable chemical production, alternative energy, and pollution mitigation applications to meet the demands of our world’s rising population. It is a challenging endeavor, however, to make novel heterogeneous catalysts with good performance (i.e., stable, active, selective) because their performance depends on many properties: composition, support, surface termination, particle size, particle morphology, and atomic coordination environment. Additionally, the properties of heterogeneous catalysts can change under reaction conditions through various phenomena such as Ostwald ripening, particle disintegration, surface oxidation, and surface reconstruction. Many heterogeneous catalyst structures are disordered or amorphous in their active state, which further complicates their atomic-level characterization by modeling and experiment. Computational modeling using quantum mechanical (QM) methods such as density functional theory (DFT) can accelerate catalyst screening by enabling rapid prototyping and revealing active sites and structure-activity relations. The high computational cost of QM methods, however, limits the range of catalyst spaces that can be examined. Recent progress in merging ML with QM modeling and experiments promises to drive forward rational catalyst design. Therefore, it is timely to highlight the ability of ML tools to accelerate heterogeneous catalyst research. A key question we aim to address in this perspective is how machine learning can aid heterogeneous catalyst design and discovery. ML has been used in catalysis research since at least the 1990s. Early studies used neural networks to correlate catalyst physicochemical properties and reaction conditions with measured catalytic performance, but these studies were limited in the number of systems considered. Recently, ML has been applied to the high-throughput screening of heterogeneous catalysts and found to be predictive and applicable across a broad space of catalysts. ML algorithms such as decision trees, kernel ridge regression, neural networks, support vector machines, principal component analysis, and compressed sensing can help create predictive models of catalyst target properties, which are typically figures of merit corresponding to stability, activity, selectivity. In this perspective, we discuss various areas where ML is making an impact on heterogeneous catalysis research. ML is also aiding homogeneous catalysis research and shares many similarities (and differences) with ML for heterogeneous catalysis, but this discussion is beyond the perspective’s scope (for interested readers, see Ref. 26–28). Here, we emphasize the ability of ML combined with QM calculations to speed-up the search for optimal catalysts in combinatorial large spaces, such as alloys. ML-derived interatomic potentials for accurate and fast catalyst simulations will also be assessed, as well as the opportunity for ML to help find descriptors of catalyst performance in large datasets. The use of ML to aid transition state search algorithms (to compute reaction mechanisms) will Correspondence concerning this article should be addressed to B. R. Goldsmith at bgoldsm@umich.edu. read less NOT USED (high confidence) S. Bhattacharya, L. Xu, and D. Thompson, “Revisiting the earliest signatures of amyloidogenesis: Roadmaps emerging from computational modeling and experiment,” Wiley Interdisciplinary Reviews: Computational Molecular Science. 2018. link Times cited: 11 Abstract: Neurodegenerative amyloidogenesis begins with the aggregatio… read moreAbstract: Neurodegenerative amyloidogenesis begins with the aggregation of intrinsically disordered proteins (IDPs), which is the first step in a cascade of assembly events that can lead to insoluble fibrous deposits in brain tissue. IDP conformations that promote formation of toxic oligomers remain poorly understood, and are the most fundamental target of putative treatments for neurodegenerative disease. Rapid advances in theory, simulation and experimental methods, hold the promise of reversing protein aggregation by identifying and developing inhibitors of the transient amyloidogenic IDP conformations. To make meaningful progress it is important to appreciate the benefits and limitations of the latest developments in computational methods of conformational and ensemble modeling, and their integration and validation with experiments. Integrated studies are beginning to provide significant conceptual and mechanistic insights, including identification of the properties of amyloidogenic IDPs in their free, unbound form. At the same time, contradicting viewpoints have emerged concerning convergence of IDP ensemble signatures and properties from parallel studies, and there also remains a pressing need to develop physical models that can deliver reliable predictions across different IDP families. Focussing on the four most common amyloidogenic IDPs of Amyloid β, Tau, α‐synuclein and Prions, improvements are proposed for next‐generation models and experiments that can potentially identify drug treatments for neurodegenerative disease via incorporation of the extended cellular environment. read less NOT USED (high confidence) J. Warren, “The Materials Genome Initiative and artificial intelligence,” MRS Bulletin. 2018. link Times cited: 27 Abstract: The Materials Genome Initiative (MGI) seeks to accelerate th… read moreAbstract: The Materials Genome Initiative (MGI) seeks to accelerate the discovery, design, development, and deployment of new materials through the creation of a materials innovation infrastructure. This infrastructure is essentially a system for providing data and tools that encapsulate our existing knowledge about materials, and the means to create new knowledge. Given this approach, MGI is also deeply linked to the ongoing exponential growth in applications of machine learning and artificial intelligence (AI) to materials research. This article explores the connections between MGI, the consequent need for data publication, the implications for data-driven science, and the application of AI to materials design. Examples will demonstrate how materials research is transforming in remarkable ways, and that the MGI vision of accelerated materials discovery is within reach. read less NOT USED (high confidence) A. Bartók, J. Kermode, N. Bernstein, and G. Csányi, “Machine Learning a General-Purpose Interatomic Potential for Silicon,” Physical Review X. 2018. link Times cited: 291 Abstract: The success of first principles electronic structure calcula… read moreAbstract: The success of first principles electronic structure calculation for predictive modeling in chemistry, solid state physics, and materials science is constrained by the limitations on simulated length and time scales due to computational cost and its scaling. Techniques based on machine learning ideas for interpolating the Born-Oppenheimer potential energy surface without explicitly describing electrons have recently shown great promise, but accurately and efficiently fitting the physically relevant space of configurations has remained a challenging goal. Here we present a Gaussian Approximation Potential for silicon that achieves this milestone, accurately reproducing density functional theory reference results for a wide range of observable properties, including crystal, liquid, and amorphous bulk phases, as well as point, line, and plane defects. We demonstrate that this new potential enables calculations that would be extremely expensive with a first principles electronic structure method, such as finite temperature phase boundary lines, self-diffusivity in the liquid, formation of the amorphous by slow quench, and dynamic brittle fracture. We show that the uncertainty quantification inherent to the Gaussian process regression framework gives a qualitative estimate of the potential's accuracy for a given atomic configuration. The success of this model shows that it is indeed possible to create a useful machine-learning-based interatomic potential that comprehensively describes a material, and serves as a template for the development of such models in the future. read less NOT USED (high confidence) D. Setiawan, J. Brender, and Y. Zhang, “Recent advances in automated protein design and its future challenges,” Expert Opinion on Drug Discovery. 2018. link Times cited: 19 Abstract: ABSTRACT Introduction: Protein function is determined by pro… read moreAbstract: ABSTRACT Introduction: Protein function is determined by protein structure which is in turn determined by the corresponding protein sequence. If the rules that cause a protein to adopt a particular structure are understood, it should be possible to refine or even redefine the function of a protein by working backwards from the desired structure to the sequence. Automated protein design attempts to calculate the effects of mutations computationally with the goal of more radical or complex transformations than are accessible by experimental techniques. Areas covered: The authors give a brief overview of the recent methodological advances in computer-aided protein design, showing how methodological choices affect final design and how automated protein design can be used to address problems considered beyond traditional protein engineering, including the creation of novel protein scaffolds for drug development. Also, the authors address specifically the future challenges in the development of automated protein design. Expert opinion: Automated protein design holds potential as a protein engineering technique, particularly in cases where screening by combinatorial mutagenesis is problematic. Considering solubility and immunogenicity issues, automated protein design is initially more likely to make an impact as a research tool for exploring basic biology in drug discovery than in the design of protein biologics. read less NOT USED (high confidence) S. Sundararaman, L. Huang, S. Ispas, and W. Kob, “New optimization scheme to obtain interaction potentials for oxide glasses.,” The Journal of chemical physics. 2018. link Times cited: 53 Abstract: We propose a new scheme to parameterize effective potentials… read moreAbstract: We propose a new scheme to parameterize effective potentials that can be used to simulate atomic systems such as oxide glasses. As input data for the optimization, we use the radial distribution functions of the liquid and the vibrational density of state of the glass, both obtained from ab initio simulations, as well as experimental data on the pressure dependence of the density of the glass. For the case of silica, we find that this new scheme facilitates finding pair potentials that are significantly more accurate than the previous ones even if the functional form is the same, thus demonstrating that even simple two-body potentials can be superior to more complex three-body potentials. We have tested the new potential by calculating the pressure dependence of the elastic moduli and found a good agreement with the corresponding experimental data. read less NOT USED (high confidence) Q. Tong, L. Xue, J. Lv, Y. Wang, and Y. Ma, “Accelerating CALYPSO structure prediction by data-driven learning of a potential energy surface.,” Faraday discussions. 2018. link Times cited: 48 Abstract: Ab initio structure prediction methods have been nowadays wi… read moreAbstract: Ab initio structure prediction methods have been nowadays widely used as powerful tools for structure searches and materials discovery. However, they are generally restricted to small systems owing to the heavy computational cost of the underlying density functional theory (DFT) calculations in structure optimizations. In this work, by combining a state-of-art machine learning (ML) potential with our in-house developed CALYPSO structure prediction method, we developed two acceleration schemes for the structure prediction of large systems, in which a ML potential is pre-constructed to fully replace DFT calculations or trained in an on-the-fly manner from scratch during the structure searches. The developed schemes have been applied to medium- and large-sized boron clusters, both of which are challenging cases for either the construction of ML potentials or extensive structure searches. Experimental structures of B36 and B40 clusters can be readily reproduced, and the putative global minimum structure for the B84 cluster is proposed, where the computational cost is substantially reduced by ∼1-2 orders of magnitude if compared with full DFT-based structure searches. Our results demonstrate a viable route for structure prediction in large systems via the combination of state-of-art structure prediction methods and ML techniques. read less NOT USED (high confidence) J. S. Smith, B. Nebgen, N. Lubbers, O. Isayev, and A. Roitberg, “Less is more: sampling chemical space with active learning,” The Journal of chemical physics. 2018. link Times cited: 429 Abstract: The development of accurate and transferable machine learnin… read moreAbstract: The development of accurate and transferable machine learning (ML) potentials for predicting molecular energetics is a challenging task. The process of data generation to train such ML potentials is a task neither well understood nor researched in detail. In this work, we present a fully automated approach for the generation of datasets with the intent of training universal ML potentials. It is based on the concept of active learning (AL) via Query by Committee (QBC), which uses the disagreement between an ensemble of ML potentials to infer the reliability of the ensemble's prediction. QBC allows the presented AL algorithm to automatically sample regions of chemical space where the ML potential fails to accurately predict the potential energy. AL improves the overall fitness of ANAKIN-ME (ANI) deep learning potentials in rigorous test cases by mitigating human biases in deciding what new training data to use. AL also reduces the training set size to a fraction of the data required when using naive random sampling techniques. To provide validation of our AL approach, we develop the COmprehensive Machine-learning Potential (COMP6) benchmark (publicly available on GitHub) which contains a diverse set of organic molecules. Active learning-based ANI potentials outperform the original random sampled ANI-1 potential with only 10% of the data, while the final active learning-based model vastly outperforms ANI-1 on the COMP6 benchmark after training to only 25% of the data. Finally, we show that our proposed AL technique develops a universal ANI potential (ANI-1x) that provides accurate energy and force predictions on the entire COMP6 benchmark. This universal ML potential achieves a level of accuracy on par with the best ML potentials for single molecules or materials, while remaining applicable to the general class of organic molecules composed of the elements CHNO. read less NOT USED (high confidence) G. Pilania and X.-Y. Liu, “Machine learning properties of binary wurtzite superlattices,” Journal of Materials Science. 2018. link Times cited: 23 NOT USED (high confidence) A. Mannodi-Kanakkithodi et al., “Scoping the polymer genome: A roadmap for rational polymer dielectrics design and beyond,” Materials Today. 2017. link Times cited: 134 NOT USED (high confidence) F. Fracchia, G. D. Frate, G. Mancini, W. Rocchia, and V. Barone, “Force Field Parametrization of Metal Ions from Statistical Learning Techniques,” Journal of Chemical Theory and Computation. 2017. link Times cited: 33 Abstract: A novel statistical procedure has been developed to optimize… read moreAbstract: A novel statistical procedure has been developed to optimize the parameters of nonbonded force fields of metal ions in soft matter. The criterion for the optimization is the minimization of the deviations from ab initio forces and energies calculated for model systems. The method exploits the combination of the linear ridge regression and the cross-validation techniques with the differential evolution algorithm. Wide freedom in the choice of the functional form of the force fields is allowed since both linear and nonlinear parameters can be optimized. In order to maximize the information content of the data employed in the fitting procedure, the composition of the training set is entrusted to a combinatorial optimization algorithm which maximizes the dissimilarity of the included instances. The methodology has been validated using the force field parametrization of five metal ions (Zn2+, Ni2+, Mg2+, Ca2+, and Na+) in water as test cases. read less NOT USED (high confidence) J. Wu, L. Shen, and W. Yang, “Internal force corrections with machine learning for quantum mechanics/molecular mechanics simulations.,” The Journal of chemical physics. 2017. link Times cited: 28 Abstract: Ab initio quantum mechanics/molecular mechanics (QM/MM) mole… read moreAbstract: Ab initio quantum mechanics/molecular mechanics (QM/MM) molecular dynamics simulation is a useful tool to calculate thermodynamic properties such as potential of mean force for chemical reactions but intensely time consuming. In this paper, we developed a new method using the internal force correction for low-level semiempirical QM/MM molecular dynamics samplings with a predefined reaction coordinate. As a correction term, the internal force was predicted with a machine learning scheme, which provides a sophisticated force field, and added to the atomic forces on the reaction coordinate related atoms at each integration step. We applied this method to two reactions in aqueous solution and reproduced potentials of mean force at the ab initio QM/MM level. The saving in computational cost is about 2 orders of magnitude. The present work reveals great potentials for machine learning in QM/MM simulations to study complex chemical processes. read less NOT USED (high confidence) J. Behler, “Hochdimensionale neuronale Netze für Potentialhyperflächen großer molekularer und kondensierter Systeme,” Angewandte Chemie. 2017. link Times cited: 9 Abstract: Moderne Simulationstechniken haben heute ein Niveau erreicht… read moreAbstract: Moderne Simulationstechniken haben heute ein Niveau erreicht, das es ermoglicht, ein breites Spektrum von Problemen in der Chemie und in den Materialwissenschaften zu untersuchen. Trotz der Entwicklung immer leistungsfahigerer Hardware ist die Anwendung von Elektronenstrukturmethoden, die Vorhersagen ohne Ruckgriff auf experimentelle Daten ermoglichen, jedoch noch immer auf kleine Systeme begrenzt, und in absehbarer Zukunft ist keine Besserung dieser Situation zu erwarten. Um auch komplexe Systeme auf atomarer Ebene verstehen zu konnen, ist die Entwicklung von effizienteren und gleichzeitig zuverlassigen atomistischen Potentialen in den letzten Jahren immer mehr in den Fokus geruckt. Ein vielversprechender neuer Ansatz ist die Nutzung von maschinellem Lernen (ML) zur Beschreibung der atomaren Wechselwirkungen. Nach einem Trainingsprozess mit Elektronstrukturdaten konnen solche ML-Potentiale Computersimulationen um mehrere Grosenordnungen beschleunigen, wahrend die quantenmechanische Genauigkeit erhalten bleibt. Anhand einer wichtigen Klasse von ML-Potentialen, die auf kunstlichen neuronalen Netzen basiert, werden in diesem Aufsatz die Grundideen, die Anwendbarkeit und die offenen Fragen dieses Ansatzes diskutiert. read less NOT USED (high confidence) J. Behler, “First Principles Neural Network Potentials for Reactive Simulations of Large Molecular and Condensed Systems.,” Angewandte Chemie. 2017. link Times cited: 409 Abstract: Modern simulation techniques have reached a level of maturit… read moreAbstract: Modern simulation techniques have reached a level of maturity which allows a wide range of problems in chemistry and materials science to be addressed. Unfortunately, the application of first principles methods with predictive power is still limited to rather small systems, and despite the rapid evolution of computer hardware no fundamental change in this situation can be expected. Consequently, the development of more efficient but equally reliable atomistic potentials to reach an atomic level understanding of complex systems has received considerable attention in recent years. A promising new development has been the introduction of machine learning (ML) methods to describe the atomic interactions. Once trained with electronic structure data, ML potentials can accelerate computer simulations by several orders of magnitude, while preserving quantum mechanical accuracy. This Review considers the methodology of an important class of ML potentials that employs artificial neural networks. read less NOT USED (high confidence) D. Dragoni, T. Daff, G. Csányi, and N. Marzari, “Achieving DFT accuracy with a machine-learning interatomic potential: thermomechanics and defects in bcc ferromagnetic iron,” arXiv: Materials Science. 2017. link Times cited: 167 Abstract: We show that the Gaussian Approximation Potential machine le… read moreAbstract: We show that the Gaussian Approximation Potential machine learning framework can describe complex magnetic potential energy surfaces, taking ferromagnetic iron as a paradigmatic challenging case. The training database includes total energies, forces, and stresses obtained from density-functional theory in the generalized-gradient approximation, and comprises approximately 150,000 local atomic environments, ranging from pristine and defected bulk configurations to surfaces and generalized stacking faults with different crystallographic orientations. We find the structural, vibrational and thermodynamic properties of the GAP model to be in excellent agreement with those obtained directly from first-principles electronic-structure calculations. There is good transferability to quantities, such as Peierls energy barriers, which are determined to a large extent by atomic configurations that were not part of the training set. We observe the benefit and the need of using highly converged electronic-structure calculations to sample a target potential energy surface. The end result is a systematically improvable potential that can achieve the same accuracy of density-functional theory calculations, but at a fraction of the computational cost. read less NOT USED (high confidence) M. Gastegger, J. Behler, and P. Marquetand, “Machine learning molecular dynamics for the simulation of infrared spectra† †Electronic supplementary information (ESI) available. See DOI: 10.1039/c7sc02267k,” Chemical Science. 2017. link Times cited: 279 Abstract: Artificial neural networks are combined with molecular dynam… read moreAbstract: Artificial neural networks are combined with molecular dynamics to simulate molecular infrared spectra including anharmonicities and temperature effects. read less NOT USED (high confidence) E. Podryabinkin and A. Shapeev, “Active learning of linearly parametrized interatomic potentials,” Computational Materials Science. 2016. link Times cited: 351 NOT USED (high confidence) Y. Yang et al., “Taking materials dynamics to new extremes using machine learning interatomic potentials,” Journal of Materials Informatics. 2021. link Times cited: 5 Abstract: Understanding materials dynamics under extreme conditions of… read moreAbstract: Understanding materials dynamics under extreme conditions of pressure, temperature, and strain rate is a scientific quest that spans nearly a century. Atomic simulations have had a considerable impact on this endeavor because of their ability to uncover materials’ microstructure evolution and properties at the scale of the relevant physical phenomena. However, this is still a challenge for most materials as it requires modeling large atomic systems (up to millions of particles) with improved accuracy. In many cases, the availability of sufficiently accurate but efficient interatomic potentials has become a serious bottleneck for performing these simulations as traditional potentials fail to represent the multitude of bonding. A new class of potentials has emerged recently, based on a different paradigm from the traditional approach. The new potentials are constructed by machinelearning with a high degree of fidelity from quantum-mechanical calculations. In this review, a brief introduction to the central ideas underlying machine learning interatomic potentials is given. In particular, the coupling of machine learning models with domain knowledge to improve accuracy, computational efficiency, and interpretability is highlighted. Subsequently, we demonstrate the effectiveness of the domain knowledge-based approach in certain select problems related to the kinetic response of warm dense materials. It is hoped that this review will inspire further advances in the understanding of matter under extreme conditions. read less NOT USED (high confidence) D. Fedorov, “Analyzing Interactions with the Fragment Molecular Orbital Method.,” Methods in molecular biology. 2020. link Times cited: 2 NOT USED (high confidence) Y. Litman, J. Behler, and M. Rossi, “Temperature dependence of the vibrational spectrum of porphycene: a qualitative failure of classical-nuclei molecular dynamics.,” Faraday discussions. 2019. link Times cited: 14 Abstract: The temperature dependence of vibrational spectra can provid… read moreAbstract: The temperature dependence of vibrational spectra can provide information about structural changes of a system and also serve as a probe to identify different vibrational mode couplings. Fully anharmonic temperature-dependent calculations of these quantities are challenging due to the cost associated with statistically converging trajectory-based methods, especially when accounting for nuclear quantum effects. Here, we train a high-dimensional neural network potential energy surface for the porphycene molecule based on data generated with DFT-B3LYP, including pairwise van der Waals interactions. In addition, we fit a kernel ridge regression model for the molecular dipole moment surface. The combination of this machinery with thermostatted path integral molecular dynamics (TRPMD) allows us to obtain well-converged, full-dimensional, fully-anharmonic vibrational spectra including nuclear quantum effects, without sacrificing the first-principles quality of the potential-energy surface or the dipole surface. Within this framework, we investigate the temperature and isotopologue dependence of the high-frequency vibrational fingerprints of porphycene. While classical-nuclei dynamics predicts a red shift of the vibrations encompassing the NH and CH stretches, TRPMD predicts a strong blue shift in the NH-stretch region and a smaller one in the CH-stretch region. We explain this behavior by analyzing the modulation of the effective potential with temperature, which arises from vibrational coupling between quasi-classical thermally activated modes and high-frequency quantized modes. read less NOT USED (definite) A. S. Christensen, L. A. Bratholm, F. A. Faber, and O. A. von Lilienfeld, “FCHL revisited: Faster and more accurate quantum machine learning.,” The Journal of chemical physics. 2019. link Times cited: 201 Abstract: We introduce the FCHL19 representation for atomic environmen… read moreAbstract: We introduce the FCHL19 representation for atomic environments in molecules or condensed-phase systems. Machine learning models based on FCHL19 are able to yield predictions of atomic forces and energies of query compounds with chemical accuracy on the scale of milliseconds. FCHL19 is a revision of our previous work [F. A. Faber et al., J. Chem. Phys. 148, 241717 (2018)] where the representation is discretized and the individual features are rigorously optimized using Monte Carlo optimization. Combined with a Gaussian kernel function that incorporates elemental screening, chemical accuracy is reached for energy learning on the QM7b and QM9 datasets after training for minutes and hours, respectively. The model also shows good performance for non-bonded interactions in the condensed phase for a set of water clusters with a mean absolute error (MAE) binding energy error of less than 0.1 kcal/mol/molecule after training on 3200 samples. For force learning on the MD17 dataset, our optimized model similarly displays state-of-the-art accuracy with a regressor based on Gaussian process regression. When the revised FCHL19 representation is combined with the operator quantum machine learning regressor, forces and energies can be predicted in only a few milliseconds per atom. The model presented herein is fast and lightweight enough for use in general chemistry problems as well as molecular dynamics simulations. read less NOT USED (definite) C. Brunken and M. Reiher, “Self-Parametrizing System-Focused Atomistic Models.,” Journal of chemical theory and computation. 2019. link Times cited: 20 Abstract: Computational studies of chemical reactions in complex envir… read moreAbstract: Computational studies of chemical reactions in complex environments such as proteins, nanostructures, or on surfaces require accurate and efficient atomistic models applicable to the nanometer scale. In general, an accurate parametrization of the atomistic entities will not be available for arbitrary system classes, but demands a fast automated system-focused parametrization procedure to be quickly applicable, reliable, flexible, and reproducible. Here, we develop and combine an automatically parametrizable quantum chemically derived molecular mechanics model with machine-learned corrections under autonomous uncertainty quantification and refinement. Our approach first generates an accurate, physically motivated model from a minimum energy structure and its corresponding Hessian matrix by a partial Hessian fitting procedure of the force constants. This model is then the starting point to generate a large number of configurations for which additional off-minimum reference data can be evaluated on the fly. A Delta-machine learning model is trained on these data to provide a correction to energies and forces including uncertainty estimates. During the procedure, the flexibility of the machine learning model is tailored to the amount of available training data. The parametrization of large systems is enabled by a fragmentation approach. Due to their modular nature, all model construction steps allow for model improvement in a rolling fashion. Our approach may also be employed for the generation of system-focused electrostatic molecular mechanics embedding environments in a quantum mechanical molecular-mechanical hybrid model for arbitrary atomistic structures at the nanoscale. read less NOT USED (definite) X. Lei, F. Hohman, D. H. Chau, and A. Medford, “ElectroLens: Understanding Atomistic Simulations through Spatially-Resolved Visualization of High-Dimensional Features,” 2019 IEEE Visualization Conference (VIS). 2019. link Times cited: 2 Abstract: In recent years, machine learning (ML) has gained significan… read moreAbstract: In recent years, machine learning (ML) has gained significant popularity in the field of chemical informatics and electronic structure theory. These techniques often require researchers to engineer abstract "features" that encode chemical concepts into a mathematical form compatible with the input to machine-learning models. However, there is no existing tool to connect these abstract features back to the actual chemical system, making it difficult to diagnose failures and to build intuition about the meaning of the features. We present ElectroLens, a new visualization tool for high-dimensional spatially-resolved features to tackle this problem. The tool visualizes high-dimensional data sets for atomistic and electron environment features by a series of linked 3D views and 2D plots. The tool is able to connect different derived features and their corresponding regions in 3D via interactive selection. It is built to be scalable, and integrate with existing infrastructure. read less NOT USED (definite) E. Kocer, J. Mason, and H. Erturk, “Continuous and optimally complete description of chemical environments using Spherical Bessel descriptors,” AIP Advances. 2019. link Times cited: 18 Abstract: Recently, machine learning potentials have been advanced as … read moreAbstract: Recently, machine learning potentials have been advanced as candidates to combine the high-accuracy of quantum mechanical simulations with the speed of classical interatomic potentials. A crucial component of a machine learning potential is the description of local atomic environments by some set of descriptors. These should ideally be continuous throughout the specified local atomic environment, twice-differentiable with respect to atomic positions and complete in the sense of containing all possible information about the neighborhood. An updated version of the recently proposed Spherical Bessel descriptors satisfies all three of these properties, and moreover is optimally complete in the sense of encoding all configurational information with the smallest possible number of descriptors. The Smooth Overlap of Atomic Position descriptors that are frequently visited in the literature and the Zernike descriptors that are built upon a similar basis are included into the discussion as being the natural counterparts of the Spherical Bessel descriptors, and shown to be incapable of satisfying the full list of core requirements for an accurate description. Aside being mathematically and physically superior, the Spherical Bessel descriptors have also the advantage of allowing machine learning potentials of comparable accuracy that require roughly an order of magnitude less computation time per evaluation than the Smooth Overlap of Atomic Position descriptors, which appear to be the common choice of descriptors in recent studies. read less NOT USED (definite) E. Kocer, J. Mason, and H. Erturk, “A novel approach to describe chemical environments in high-dimensional neural network potentials.,” The Journal of chemical physics. 2019. link Times cited: 24 Abstract: A central concern of molecular dynamics simulations is the p… read moreAbstract: A central concern of molecular dynamics simulations is the potential energy surfaces that govern atomic interactions. These hypersurfaces define the potential energy of the system and have generally been calculated using either predefined analytical formulas (classical) or quantum mechanical simulations (ab initio). The former can accurately reproduce only a selection of material properties, whereas the latter is restricted to short simulation times and small systems. Machine learning potentials have recently emerged as a third approach to model atomic interactions, and are purported to offer the accuracy of ab initio simulations with the speed of classical potentials. However, the performance of machine learning potentials depends crucially on the description of a local atomic environment. A set of invariant, orthogonal, and differentiable descriptors for an atomic environment is proposed, implemented in a neural network potential for solid-state silicon, and tested in molecular dynamics simulations. Neural networks using the proposed descriptors are found to outperform ones using the Behler-Parinello and smooth overlap of atomic position descriptors in the literature. read less NOT USED (definite) J. D. de Pablo et al., “New frontiers for the materials genome initiative,” npj Computational Materials. 2019. link Times cited: 292 NOT USED (definite) A. Chandrasekaran, D. Kamal, R. Batra, C. Kim, L. Chen, and R. Ramprasad, “Solving the electronic structure problem with machine learning,” npj Computational Materials. 2019. link Times cited: 181 NOT USED (definite) M. Haghighatlari and J. Hachmann, “Advances of machine learning in molecular modeling and simulation,” Current Opinion in Chemical Engineering. 2019. link Times cited: 76 NOT USED (definite) M. Gastegger and P. Marquetand, “Molecular Dynamics with Neural Network Potentials,” Machine Learning Meets Quantum Physics. 2018. link Times cited: 20 NOT USED (definite) D. Stephenson, J. Kermode, and D. Lockerby, “Accelerating multiscale modelling of fluids with on-the-fly Gaussian process regression,” Microfluidics and Nanofluidics. 2018. link Times cited: 7 NOT USED (definite) L. Zhang, D.-Y. Lin, H. Wang, R. Car, and E. Weinan, “Active Learning of Uniformly Accurate Inter-atomic Potentials for Materials Simulation,” ArXiv. 2018. link Times cited: 246 Abstract: An active learning procedure called Deep Potential Generator… read moreAbstract: An active learning procedure called Deep Potential Generator (DP-GEN) is proposed for the construction of accurate and transferable machine learning-based models of the potential energy surface (PES) for the molecular modeling of materials. This procedure consists of three main components: exploration, generation of accurate reference data, and training. Application to the sample systems of Al, Mg and Al-Mg alloys demonstrates that DP-GEN can produce uniformly accurate PES models with a minimal number of reference data. read less NOT USED (definite) H. Zong, G. Pilania, X. Ding, G. Ackland, and T. Lookman, “Developing an interatomic potential for martensitic phase transformations in zirconium by machine learning,” npj Computational Materials. 2018. link Times cited: 78 NOT USED (definite) J. Heinen and D. Dubbeldam, “On flexible force fields for metal–organic frameworks: Recent developments and future prospects,” Wiley Interdisciplinary Reviews. Computational Molecular Science. 2018. link Times cited: 36 Abstract: Classical force field simulations can be used to study struc… read moreAbstract: Classical force field simulations can be used to study structural, diffusion, and adsorption properties of metal–organic frameworks (MOFs). To account for the dynamic behavior of the material, parameterization schemes have been developed to derive force constants and the associated reference values by fitting on ab initio energies, vibrational frequencies, and elastic constants. Here, we review recent developments in flexible force field models for MOFs. Existing flexible force field models are generally able to reproduce the majority of experimentally observed structural and dynamic properties of MOFs. The lack of efficient sampling schemes for capturing stimuli‐driven phase transitions, however, currently limits the full predictive potential of existing flexible force fields from being realized. read less NOT USED (definite) P. Li, X. Jia, X. Pan, Y. Shao, and Y. Mei, “Accelerated Computation of Free Energy Profile at ab Initio Quantum Mechanical/Molecular Mechanics Accuracy via a Semi-Empirical Reference Potential. I. Weighted Thermodynamics Perturbation.,” Journal of chemical theory and computation. 2018. link Times cited: 43 Abstract: Free energy profile (FE Profile) is an essential quantity fo… read moreAbstract: Free energy profile (FE Profile) is an essential quantity for the estimation of reaction rate and the validation of reaction mechanism. For chemical reactions in condensed phase or enzymatic reactions, the computation of FE profile at the ab initio (ai) quantum mechanical/molecular mechanics (QM/MM) level is still far too expensive. Although semiempirical (SE) method can be hundreds or thousands of times faster than the ai methods, the accuracy of SE methods is often unsatisfactory due to the approximations that have been adopted in these methods. In this work, we propose a new method termed MBAR+wTP in which the ai QM/MM free energy profile is computed by a weighted thermodynamic perturbation (TP) correction to the SE profile generated by the multistate Bennett acceptance ratio (MBAR) analysis of the trajectories from umbrella samplings (US). The weight factors used in the TP calculations are a byproduct of the MBAR analysis in the postprocessing of the US trajectories, which are often discarded after the free energy calculations. The raw ai QM/MM free energy profile is then smoothed using Gaussian process regression in which the noise of each datum is set to be inversely proportional to the exponential of the reweighting entropy. The results show that this approach can enhance the efficiency of ai FE profile calculations by several orders of magnitude with only a slight loss of accuracy. This method can significantly enhance the applicability of ai QM/MM methods in the studies of chemical reactions in condensed phase and enzymatic reactions. read less NOT USED (definite) T. D. Huan, R. Batra, J. Chapman, S. Krishnan, L. Chen, and R. Ramprasad, “A universal strategy for the creation of machine learning-based atomistic force fields,” npj Computational Materials. 2017. link Times cited: 201 NOT USED (definite) C. Kim, T. D. Huan, S. Krishnan, and R. Ramprasad, “A hybrid organic-inorganic perovskite dataset,” Scientific Data. 2017. link Times cited: 109 NOT USED (definite) “Charge Transfer in Classical Molecular Dynamics Simulations of Met-enkephalin: Improving Traditional Force Field with Data Driven Models,” arXiv: Chemical Physics. 2018. link Times cited: 0 Abstract: The charge transfer and polarization effects are important c… read moreAbstract: The charge transfer and polarization effects are important components in the molecular mechanism description of bio-molecules. Classical force field with fixed point charge cannot take into the account of the non-negligible correlation between atomic charge and structure changes. In this work, high throughput ab initio calculations for the pentapeptide Met-enkephalin (MetEnk) reveal that geometric dependent charge transfer among residues is significant among tens of thousands of conformations. And we suggest a data driven model with machine learning algorithms to solve the geometric dependent charge fluctuations problem. This data driven model can directly provide ab initio level atomic charges of any structure for MetEnk, and avoids self-consistent iteration in polarizable force field. Molecular dynamics simulations demonstrated that the data driven model provides a possible choice to describe the explicit charge flux with minor modification of available classical force fields. This work provides us an alternative molecular mechanism model for future dynamics simulation of oligopeptides. read less NOT USED (definite) “Linking plastic heterogeneity of bulk metallic glasses to quench-in structural defects with machine learning,” arXiv: Materials Science. 2019. link Times cited: 0 Abstract: When metallic glasses are subjected to mechanical loads, the… read moreAbstract: When metallic glasses are subjected to mechanical loads, the plastic response of atoms is heterogeneous. However, the degree to which the plastic units are correlated with the structural defects frozen in the quenched glass structure is still elusive. Here, we introduce a machine learning framework to predict the plastic heterogeneity of atoms in Cu-Zr metallic glasses solely from the undeformed, quenched configuration. We propose that an atomic-scale quantity, "quench-in softness", calibrated from a gradient boosted decision tree model trained on a set of short- and medium-range site features, can identify plastically susceptible sites at various strain levels with high accuracy. The predictive ability is further confirmed in that a model trained on a single composition and quench rate retains high accuracy on other compositions and quench rates without any further training. We also quantitatively assess historical site descriptors against our method, demonstrating that the regularity-related features introduced in this work are more predictive and may play an important role in future glass characterization. Our work presents a general, data-centric framework that could potentially be used to address the structural origin of any site-specific property in metallic glasses. read less |
