Machine Learning Force Fields: Recent Advances and Remaining Challenges

In Silico Chemical Experiments in the Age of AI: From Quantum Chemistry to Machine Learning and Back

Computational chemistry is an indispensable tool for understanding molecules and predicting chemical properties. However, traditional computational methods face significant challenges due to the difficulty of solving the Schrödinger equations and the increasing computational cost with the size of the molecular system. In response, there has been a surge of interest in leveraging artificial intelligence (AI) and machine learning (ML) techniques to in silico experiments. Integrating AI and ML into computational chemistry increases the scalability and speed of the exploration of chemical space. However, challenges remain, particularly regarding the reproducibility and transferability of ML models. This review highlights the evolution of ML in learning from, complementing, or replacing traditional computational chemistry for energy and property predictions. Starting from models trained entirely on numerical data, a journey set forth toward the ideal model incorporating or learning the physical laws of quantum mechanics. This paper also reviews existing computational methods and ML models and their intertwining, outlines a roadmap for future research, and identifies areas for improvement and innovation. Ultimately, the goal is to develop AI architectures capable of predicting accurate and transferable solutions to the Schrödinger equation, thereby revolutionizing in silico experiments within chemistry and materials science.

Molecular Dynamics Simulation of Complex Reactivity with the Rapid Approach for Proton Transport and Other Reactions (RAPTOR) Software Package

Simulating chemically reactive phenomena such as proton transport on nanosecond to microsecond and beyond time scales is a challenging task. Ab initio methods are unable to currently access these time scales routinely, and traditional molecular dynamics methods feature fixed bonding arrangements that cannot account for changes in the system's bonding topology. The Multiscale Reactive Molecular Dynamics (MS-RMD) method, as implemented in the Rapid Approach for Proton Transport and Other Reactions (RAPTOR) software package for the LAMMPS molecular dynamics code, offers a method to routinely sample longer time scale reactive simulation data with statistical precision. RAPTOR may also be interfaced with enhanced sampling methods to drive simulations toward the analysis of reactive rare events, and a number of collective variables (CVs) have been developed to facilitate this. Key advances to this methodology, including GPU acceleration efforts and novel CVs to model water wire formation are reviewed, along with recent applications of the method which demonstrate its versatility and robustness.

Machine learning interatomic potential: Bridge the gap between small-scale models and realistic device-scale simulations

Machine learning interatomic potential (MLIP) overcomes the challenges of high computational costs in density-functional theory and the relatively low accuracy in classical large-scale molecular dynamics, facilitating more efficient and precise simulations in materials research and design. In this review, the current state of the four essential stages of MLIP is discussed, including data generation methods, material structure descriptors, six unique machine learning algorithms, and available software. Furthermore, the applications of MLIP in various fields are investigated, notably in phase-change memory materials, structure searching, material properties predicting, and the pre-trained universal models. Eventually, the future perspectives, consisting of standard datasets, transferability, generalization, and trade-off between accuracy and complexity in MLIPs, are reported.

Accelerated chemical science with AI

In light of the pressing need for practical materials and molecular solutions to renewable energy and health problems, to name just two examples, one wonders how to accelerate research and development in the chemical sciences, so as to address the time it takes to bring materials from initial discovery to commercialization. Artificial intelligence (AI)-based techniques, in particular, are having a transformative and accelerating impact on many if not most, technological domains. To shed light on these questions, the authors and participants gathered in person for the ASLLA Symposium on the theme of 'Accelerated Chemical Science with AI' at Gangneung, Republic of Korea. We present the findings, ideas, comments, and often contentious opinions expressed during four panel discussions related to the respective general topics: 'Data', 'New applications', 'Machine learning algorithms', and 'Education'. All discussions were recorded, transcribed into text using Open AI's Whisper, and summarized using LG AI Research's EXAONE LLM, followed by revision by all authors. For the broader benefit of current researchers, educators in higher education, and academic bodies such as associations, publishers, librarians, and companies, we provide chemistry-specific recommendations and summarize the resulting conclusions.

Unlocking novel therapies: cyclic peptide design for amyloidogenic targets through synergies of experiments, simulations, and machine learning

Existing therapies for neurodegenerative diseases like Parkinson's and Alzheimer's address only their symptoms and do not prevent disease onset. Common therapeutic agents, such as small molecules and antibodies struggle with insufficient selectivity, stability and bioavailability, leading to poor performance in clinical trials. Peptide-based therapeutics are emerging as promising candidates, with successful applications for cardiovascular diseases and cancers due to their high bioavailability, good efficacy and specificity. In particular, cyclic peptides have a long in vivo stability, while maintaining a robust antibody-like binding affinity. However, the de novo design of cyclic peptides is challenging due to the lack of long-lived druggable pockets of the target polypeptide, absence of exhaustive conformational distributions of the target and/or the binder, unknown binding site, methodological limitations, associated constraints (failed trials, time, money) and the vast combinatorial sequence space. Hence, efficient alignment and cooperation between disciplines, and synergies between experiments and simulations complemented by popular techniques like machine-learning can significantly speed up the therapeutic cyclic-peptide development for neurodegenerative diseases. We review the latest advancements in cyclic peptide design against amyloidogenic targets from a computational perspective in light of recent advancements and potential of machine learning to optimize the design process. We discuss the difficulties encountered when designing novel peptide-based inhibitors and we propose new strategies incorporating experiments, simulations and machine learning to design cyclic peptides to inhibit the toxic propagation of amyloidogenic polypeptides. Importantly, these strategies extend beyond the mere design of cyclic peptides and serve as template for the de novo generation of (bio)materials with programmable properties.

Development of a machine learning finite-range nonlocal density functional

Kohn-Sham density functional theory has been the most popular method in electronic structure calculations. To fulfill the increasing accuracy requirements, new approximate functionals are needed to address key issues in existing approximations. It is well known that nonlocal components are crucial. Current nonlocal functionals mostly require orbital dependence such as in Hartree-Fock exchange and many-body perturbation correlation energy, which, however, leads to higher computational costs. Deviating from this pathway, we describe functional nonlocality in a new approach. By partitioning the total density to atom-centered local densities, a many-body expansion is proposed. This many-body expansion can be truncated at one-body contributions, if a base functional is used and an energy correction is approximated. The contribution from each atom-centered local density is a single finite-range nonlocal functional that is universal for all atoms. We then use machine learning to develop this universal atom-centered functional. Parameters in this functional are determined by fitting to data that are produced by high-level theories. Extensive tests on several different test sets, which include reaction energies, reaction barrier heights, and non-covalent interaction energies, show that the new functional, with only the density as the basic variable, can produce results comparable to the best-performing double-hybrid functionals, (for example, for the thermochemistry test set selected from the GMTKN55 database, BLYP based machine learning functional gives a weighted total mean absolute deviations of 3.33 kcal/mol, while DSD-BLYP-D3(BJ) gives 3.28 kcal/mol) with a lower computational cost. This opens a new pathway to nonlocal functional development and applications.

NICE-FF: A non-empirical, intermolecular, consistent, and extensible force field for nucleic acids and beyond

A new non-empirical ab initio intermolecular force field (NICE-FF in buffered 14-7 potential form) has been developed for nucleic acids and beyond based on the dimer interaction energies (IEs) calculated at the spin component scaled-MI-second order Møller-Plesset perturbation theory. A fully automatic framework has been implemented for this purpose, capable of generating well-polished computational grids, performing the necessary ab initio calculations, conducting machine learning (ML) assisted force field (FF) parametrization, and extending existing FF parameters by incorporating new atom types. For the ML-assisted parametrization of NICE-FF, interaction energies of ∼18 000 dimer geometries (with IE < 0) were used, and the best fit gave a mean square deviation of about 0.46 kcal/mol. During this parametrization, atom types apparent in four deoxyribonucleic acid (DNA) bases have been first trained using the generated DNA base datasets. Both uracil and hypoxanthine, which contain the same atom types found in DNA bases, have been considered as test molecules. Three new atom types have been added to the DNA atom types by using IE datasets of both pyrazinamide and 9-methylhypoxanthine. Finally, the last test molecule, theophylline, has been selected, which contains already-fitted atom-type parameters. The performance of NICE-FF has been investigated on the S22 dataset, and it has been found that NICE-FF outperforms the well-known FFs by generating the most consistent IEs with the high-level ab initio ones. Moreover, NICE-FF has been integrated into our in-house developed crystal structure prediction (CSP) tool [called FFCASP (Fast and Flexible CrystAl Structure Predictor)], aiming to find the experimental crystal structures of all considered molecules. CSPs, which were performed up to 4 formula units (Z), resulted in NICE-FF being able to locate almost all the known experimental crystal structures with sufficiently low RMSD20 values to provide good starting points for density functional theory optimizations.

Kohn-Sham accuracy from orbital-free density functional theory via Δ-machine learning

We present a Δ-machine learning model for obtaining Kohn-Sham accuracy from orbital-free density functional theory (DFT) calculations. In particular, we employ a machine-learned force field (MLFF) scheme based on the kernel method to capture the difference between Kohn-Sham and orbital-free DFT energies/forces. We implement this model in the context of on-the-fly molecular dynamics simulations and study its accuracy, performance, and sensitivity to parameters for representative systems. We find that the formalism not only improves the accuracy of Thomas-Fermi-von Weizsäcker orbital-free energies and forces by more than two orders of magnitude but is also more accurate than MLFFs based solely on Kohn-Sham DFT while being more efficient and less sensitive to model parameters. We apply the framework to study the structure of molten Al0.88Si0.12, the results suggesting no aggregation of Si atoms, in agreement with a previous Kohn-Sham study performed at an order of magnitude smaller length and time scales.

Hybrid classical/machine-learning force fields for the accurate description of molecular condensed-phase systems

Electronic structure methods offer in principle accurate predictions of molecular properties, however, their applicability is limited by computational costs. Empirical methods are cheaper, but come with inherent approximations and are dependent on the quality and quantity of training data. The rise of machine learning (ML) force fields (FFs) exacerbates limitations related to training data even further, especially for condensed-phase systems for which the generation of large and high-quality training datasets is difficult. Here, we propose a hybrid ML/classical FF model that is parametrized exclusively on high-quality ab initio data of dimers and monomers in vacuum but is transferable to condensed-phase systems. The proposed hybrid model combines our previous ML-parametrized classical model with ML corrections for situations where classical approximations break down, thus combining the robustness and efficiency of classical FFs with the flexibility of ML. Extensive validation on benchmarking datasets and experimental condensed-phase data, including organic liquids and small-molecule crystal structures, showcases how the proposed approach may promote FF development and unlock the full potential of classical FFs.

High-Dimensional Neural Network Potentials for Accurate Prediction of Equation of State: A Case Study of Methane

Machine learning-based interatomic potentials, such as those provided by neural networks, are increasingly important in molecular dynamics simulations. In the present work, we consider the applicability and robustness of machine learning molecular dynamics to predict the equation of state properties of methane by using high-dimensional neural network potentials (HDNNPs). We investigate two different strategies for generating training data: one strategy based upon bulk representations using periodic cells and another strategy based upon clusters of molecules. We assess the accuracy of the trained potentials by predicting the equilibrium mass density for a wide range of thermodynamic conditions to characterize the liquid phase, supercritical fluid, and gas phase, as well as the liquid-vapor coexistence curve. Our results show an excellent agreement with reference phase diagrams, with an average error below ∼2% for all studied phases. Moreover, we confirm the applicability of models trained on cluster data sets for producing accurate and reliable results.

Stress and heat flux via automatic differentiation

Machine-learning potentials provide computationally efficient and accurate approximations of the Born-Oppenheimer potential energy surface. This potential determines many materials properties and simulation techniques usually require its gradients, in particular forces and stress for molecular dynamics, and heat flux for thermal transport properties. Recently developed potentials feature high body order and can include equivariant semi-local interactions through message-passing mechanisms. Due to their complex functional forms, they rely on automatic differentiation (AD), overcoming the need for manual implementations or finite-difference schemes to evaluate gradients. This study discusses how to use AD to efficiently obtain forces, stress, and heat flux for such potentials, and provides a model-independent implementation. The method is tested on the Lennard-Jones potential, and then applied to predict cohesive properties and thermal conductivity of tin selenide using an equivariant message-passing neural network potential.

Predicting crystal form stability under real-world conditions

The physicochemical properties of molecular crystals, such as solubility, stability, compactability, melting behaviour and bioavailability, depend on their crystal form1. In silico crystal form selection has recently come much closer to realization because of the development of accurate and affordable free-energy calculations2-4. Here we redefine the state of the art, primarily by improving the accuracy of free-energy calculations, constructing a reliable experimental benchmark for solid-solid free-energy differences, quantifying statistical errors for the computed free energies and placing both hydrate crystal structures of different stoichiometries and anhydrate crystal structures on the same energy landscape, with defined error bars, as a function of temperature and relative humidity. The calculated free energies have standard errors of 1-2 kJ mol-1 for industrially relevant compounds, and the method to place crystal structures with different hydrate stoichiometries on the same energy landscape can be extended to other multi-component systems, including solvates. These contributions reduce the gap between the needs of the experimentalist and the capabilities of modern computational tools, transforming crystal structure prediction into a more reliable and actionable procedure that can be used in combination with experimental evidence to direct crystal form selection and establish control5.

Combining Force Fields and Neural Networks for an Accurate Representation of Chemically Diverse Molecular Interactions

A key goal of molecular modeling is the accurate reproduction of the true quantum mechanical potential energy of arbitrary molecular ensembles with a tractable classical approximation. The challenges are that analytical expressions found in general purpose force fields struggle to faithfully represent the intermolecular quantum potential energy surface at close distances and in strong interaction regimes; that the more accurate neural network approximations do not capture crucial physics concepts, e.g., nonadditive inductive contributions and application of electric fields; and that the ultra-accurate narrowly targeted models have difficulty generalizing to the entire chemical space. We therefore designed a hybrid wide-coverage intermolecular interaction model consisting of an analytically polarizable force field combined with a short-range neural network correction for the total intermolecular interaction energy. Here, we describe the methodology and apply the model to accurately determine the properties of water, the free energy of solvation of neutral and charged molecules, and the binding free energy of ligands to proteins. The correction is subtyped for distinct chemical species to match the underlying force field, to segment and reduce the amount of quantum training data, and to increase accuracy and computational speed. For the systems considered, the hybrid ab initio parametrized Hamiltonian reproduces the two-body dimer quantum mechanics (QM) energies to within 0.03 kcal/mol and the nonadditive many-molecule contributions to within 2%. Simulations of molecular systems using this interaction model run at speeds of several nanoseconds per day.

Treating Semiempirical Hamiltonians as Flexible Machine Learning Models Yields Accurate and Interpretable Results

Quantum chemistry provides chemists with invaluable information, but the high computational cost limits the size and type of systems that can be studied. Machine learning (ML) has emerged as a means to dramatically lower the cost while maintaining high accuracy. However, ML models often sacrifice interpretability by using components such as the artificial neural networks of deep learning that function as black boxes. These components impart the flexibility needed to learn from large volumes of data but make it difficult to gain insight into the physical or chemical basis for the predictions. Here, we demonstrate that semiempirical quantum chemical (SEQC) models can learn from large volumes of data without sacrificing interpretability. The SEQC model is that of density-functional-based tight binding (DFTB) with fixed atomic orbital energies and interactions that are one-dimensional functions of the interatomic distance. This model is trained to ab initio data in a manner that is analogous to that used to train deep learning models. Using benchmarks that reflect the accuracy of the training data, we show that the resulting model maintains a physically reasonable functional form while achieving an accuracy, relative to coupled cluster energies with a complete basis set extrapolation (CCSD(T)*/CBS), that is comparable to that of density functional theory (DFT). This suggests that trained SEQC models can achieve a low computational cost and high accuracy without sacrificing interpretability. Use of a physically motivated model form also substantially reduces the amount of ab initio data needed to train the model compared to that required for deep learning models.

Efficient Modeling of Quantum Dynamics of Charge Carriers in Materials Using Short Nonequilibrium Molecular Dynamics

Nonadiabatic molecular dynamics provides essential insights into excited-state processes, but it is computationally intense and simplifications are needed. The classical path approximation provides critical savings. Still, long heating and equilibration steps are required. We demonstrate that practical results can be obtained with short, partially equilibrated ab initio trajectories. Once the system's structure is adequate and essential fluctuations are sampled, the nonadiabatic Hamiltonian can be constructed. Local structures require only 1-2 ps trajectories, as demonstrated with point defects in metal halide perovskites. Short trajectories represent anharmonic motions common in defective structures, an essential improvement over the harmonic approximation around the optimized geometry. Glassy systems, such as grain boundaries, require different simulation protocols, e.g., involving machine learning force fields. 10-fold shorter trajectories generate 10-20% time scale errors, which are acceptable, given experimental uncertainties and other approximations. The practical NAMD protocol enables fast screening of excited-state dynamics for rapid exploration of new materials.

Thermodynamic and kinetic modeling of electrocatalytic reactions using a first-principles approach

The computational modeling of electrochemical interfaces and their applications in electrocatalysis has attracted great attention in recent years. While tremendous progress has been made in this area, however, the accurate atomistic descriptions at the electrode/electrolyte interfaces remain a great challenge. The Computational Hydrogen Electrode (CHE) method and continuum modeling of the solvent and electrolyte interactions form the basis for most of these methodological developments. Several posterior corrections have been added to the CHE method to improve its accuracy and widen its applications. The most recently developed grand canonical potential approaches with the embedded diffuse layer models have shown considerable improvement in defining interfacial interactions at electrode/electrolyte interfaces over the state-of-the-art computational models for electrocatalysis. In this Review, we present an overview of these different computational models developed over the years to quantitatively probe the thermodynamics and kinetics of electrochemical reactions in the presence of an electrified catalyst surface under various electrochemical environments. We begin our discussion by giving a brief picture of the different continuum solvation approaches, implemented within the ab initio method to effectively model the solvent and electrolyte interactions. Next, we present the thermodynamic and kinetic modeling approaches to determine the activity and stability of the electrocatalysts. A few applications to these approaches are also discussed. We conclude by giving an outlook on the different machine learning models that have been integrated with the thermodynamic approaches to improve their efficiency and widen their applicability.

Exploring Configurations of Nanocrystal Ligands Using Machine-Learned Force Fields

Semiconducting nanocrystals passivated with organic ligands have emerged as a powerful platform for light harvesting, light-driven chemical reactions, and sensing. Due to their complexity and size, little structural information is available from experiments, making these systems challenging to model computationally. Here, we develop a machine-learned force field trained on DFT data and use it to investigate the surface chemistry of a PbS nanocrystal interfaced with acetate ligands. In doing so, we go beyond considering individual local minimum energy geometries and, importantly, circumvent a precarious issue associated with the assumption of a single assigned atomic partial charge for each element in a nanocrystal, independent of its structural position. We demonstrate that the carboxylate ligands passivate the metal-rich surfaces by adopting a very wide range of "tilted-bridge" and "bridge" geometries and investigate the corresponding ligand IR spectrum. This work illustrates the potential of machine-learned force fields to transform computational modeling of these materials.

Computational Framework Combining Quantum Mechanics, Molecular Dynamics, and Deep Neural Networks to Evaluate the Intrinsic Properties of Materials

The design and evaluation of future nanomaterials with specific properties is a challenging task as the current traditional methods rely on trial and error approaches that are time-consuming and expensive. On the computational front, design tools such as molecular dynamics (MD) simulations help us reduce the costs and times. However, nonbonded potential parameters, the key input parameters for an MD simulation, are usually not available for designing and studying new materials. Resolving this, quantum mechanics (QM) calculations could be used to evaluate the system's energy as a function of the nonbonded distances, and the resulting data set could be fit to a generic potential equation to obtain the fitting constants (potential parameters). However, fitting this massive data set containing thousands of unknown parameters using traditional mathematical formulations is not feasible. Hence, most computational frameworks in the literature utilize several simplifications, leading to a severe loss of accuracy. Addressing this deficiency, in this work, we propose a multi-scale framework that couples QM calculations and MD with advanced deep neural networks to determine the potential parameters. This advanced framework has been extensively validated by employing it to predict properties such as the density, boiling point, and melting point of five different types of molecules that are well-understood, namely, the polar molecule H2O, ionic compound LiPF6, ethanol (C2H5OH), long-chain molecule C8H18, and the complex molecular system ethylene carbonate (EC).

Transferability evaluation of the deep potential model for simulating water-graphene confined system

Machine learning potentials (MLPs) are poised to combine the accuracy of ab initio predictions with the computational efficiency of classical molecular dynamics (MD) simulation. While great progress has been made over the last two decades in developing MLPs, there is still much to be done to evaluate their model transferability and facilitate their development. In this work, we construct two deep potential (DP) models for liquid water near graphene surfaces, Model S and Model F, with the latter having more training data. A concurrent learning algorithm (DP-GEN) is adopted to explore the configurational space beyond the scope of conventional ab initio MD simulation. By examining the performance of Model S, we find that an accurate prediction of atomic force does not imply an accurate prediction of system energy. The deviation from the relative atomic force alone is insufficient to assess the accuracy of the DP models. Based on the performance of Model F, we propose that the relative magnitude of the model deviation and the corresponding root-mean-square error of the original test dataset, including energy and atomic force, can serve as an indicator for evaluating the accuracy of the model prediction for a given structure, which is particularly applicable for large systems where density functional theory calculations are infeasible. In addition to the prediction accuracy of the model described above, we also briefly discuss simulation stability and its relationship to the former. Both are important aspects in assessing the transferability of the MLP model.

Toward a general neural network force field for protein simulations: Refining the intramolecular interaction in protein

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.

Machine Learning Methods for Small Data Challenges in Molecular Science

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.

Efficient interatomic descriptors for accurate machine learning force fields of extended molecules

Machine learning force fields (MLFFs) are gradually evolving towards enabling molecular dynamics simulations of molecules and materials with ab initio accuracy but at a small fraction of the computational cost. However, several challenges remain to be addressed to enable predictive MLFF simulations of realistic molecules, including: (1) developing efficient descriptors for non-local interatomic interactions, which are essential to capture long-range molecular fluctuations, and (2) reducing the dimensionality of the descriptors to enhance the applicability and interpretability of MLFFs. Here we propose an automatized approach to substantially reduce the number of interatomic descriptor features while preserving the accuracy and increasing the efficiency of MLFFs. To simultaneously address the two stated challenges, we illustrate our approach on the example of the global GDML MLFF. We found that non-local features (atoms separated by as far as 15 Å in studied systems) are crucial to retain the overall accuracy of the MLFF for peptides, DNA base pairs, fatty acids, and supramolecular complexes. Interestingly, the number of required non-local features in the reduced descriptors becomes comparable to the number of local interatomic features (those below 5 Å). These results pave the way to constructing global molecular MLFFs whose cost increases linearly, instead of quadratically, with system size.

Fitting Force Field Parameters to NMR Relaxation Data

We present an approach to optimize force field parameters using time-dependent data from NMR relaxation experiments. To do so, we scan parameters in the dihedral angle potential energy terms describing the rotation of the methyl groups in proteins and compare NMR relaxation rates calculated from molecular dynamics simulations with the modified force fields to deuterium relaxation measurements of T4 lysozyme. We find that a small modification of C^γ methyl groups improves the agreement with experiments both for the protein used to optimize the force field and when validating using simulations of CI2 and ubiquitin. We also show that these improvements enable a more effective a posteriori reweighting of the MD trajectories. The resulting force field thus enables more direct comparison between simulations and side-chain NMR relaxation data and makes it possible to construct ensembles that better represent the dynamics of proteins in solution.

An Ecosystem for Digital Reticular Chemistry

The vastness of the materials design space makes it impractical to explore using traditional brute-force methods, particularly in reticular chemistry. However, machine learning has shown promise in expediting and guiding materials design. Despite numerous successful applications of machine learning to reticular materials, progress in the field has stagnated, possibly because digital chemistry is more an art than a science and its limited accessibility to inexperienced researchers. To address this issue, we present mofdscribe, a software ecosystem tailored to novice and seasoned digital chemists that streamlines the ideation, modeling, and publication process. Though optimized for reticular chemistry, our tools are versatile and can be used in nonreticular materials research. We believe that mofdscribe will enable a more reliable, efficient, and comparable field of digital chemistry.

Application of molecular dynamics simulation for exploring the roles of plant biomolecules in promoting environmental health

Understanding the dynamic changes of plant biomolecules is vital for exploring their mechanisms in the environment. Molecular dynamics (MD) simulation has been widely used to study structural evolution and corresponding properties of plant biomolecules at the microscopic scale. Here, this review (i) outlines structural properties of plant biomolecules, and the crucial role of MD simulation in advancing studies of the biomolecules; (ii) describes the development of MD simulation in plant biomolecules, determinants of simulation, and analysis parameters; (iii) introduces the applications of MD simulation in plant biomolecules, including the response of the biomolecules to multiple stresses, their roles in corrosive environments, and their contributions in improving environmental health; (iv) reviews techniques integrated with MD simulation, such as molecular biology, quantum mechanics, molecular docking, and machine learning modeling, which bridge gaps in MD simulation. Finally, we make suggestions on determination of force field types, investigation of plant biomolecule mechanisms, and use of MD simulation in combination with other techniques. This review provides comprehensive summaries of the mechanisms of plant biomolecules in the environment revealed by MD simulation and validates it as an applicable tool for bridging gaps between macroscopic and microscopic behavior, providing insights into the wide application of MD simulation in plant biomolecules.

Machine Learning Assisted Simulations of Electrochemical Interfaces: Recent Progress and Challenges

The electrochemical interface, where the adsorption of reactants and electrocatalytic reactions take place, has long been a focus of attention. Some of the important processes on it tend to possess relatively slow kinetic characteristics, which are usually beyond the scope of ab initio molecular dynamics. The newly emerging technique, machine learning methods, provides an alternative approach to achieve thousands of atoms and nanosecond time scale while ensuring precision and efficiency. In this Perspective, we summarize in detail the recent progress and achievements made by the introduction of machine learning to simulate electrochemical interfaces, and focus on the limitations of current machine learning models, such as accurate descriptions of long-range electrostatic interactions and the kinetics of the electrochemical reactions occurring at the interface. Finally, we further point out the future directions for machine learning to expand in the field of electrochemical interfaces.

Machine learning in computational chemistry: interplay between (non)linearity, basis sets, and dimensionality

Machine learning (ML) based methods and tools have now firmly established themselves in physical chemistry and in particular in theoretical and computational chemistry and in materials chemistry. The generality of popular ML techniques such as neural networks or kernel methods (Gaussian process and kernel ridge regression and their flavors) permitted their application to diverse problems from prediction of properties of functional materials (catalysts, solid state ionic conductors, etc.) from descriptors to the building of interatomic potentials (where ML is currently routinely used in applications) and electron density functionals. These ML techniques are assumed to have superior expressive power of nonlinear methods, and are often used "as is", with concepts such as "non-parametric" or "deep learning" used without a clear justification for their need or advantage over simpler and more robust alternatives. In this Perspective, we highlight some interrelations between popular ML techniques and traditional linear regressions and basis expansions and demonstrate that in certain regimes (such as a very high dimensionality) these approximations might collapse. We also discuss ways to recover the expressive power of a nonlinear approach and to help select hyperparameters with the help of high-dimensional model representation and to obtain elements of insight while preserving the generality of the method.

Computer-Aided Drug Design: An Update

Computer-aided drug design (CADD) approaches are playing an increasingly important role in understanding the fundamentals of ligand-receptor interactions and helping medicinal chemists design therapeutics. About 5 years ago, we presented a chapter devoted to an overview of CADD methods and covered typical CADD protocols including structure-based drug design (SBDD) and ligand-based drug design (LBDD) approaches that were frequently used in the antibiotic drug design process. Advances in computational hardware and algorithms and emerging CADD methods are enhancing the accuracy and ability of CADD in drug design and development. In this chapter, an update to our previous chapter is provided with a focus on new CADD approaches from our laboratory and other peers that can be employed to facilitate the development of antibiotic therapeutics.

Fluctuations at Metal Halide Perovskite Grain Boundaries Create Transient Trap States: Machine Learning Assisted Ab Initio Analysis

All-inorganic perovskites are promising candidates for solar energy and optoelectronic applications, despite their polycrystalline nature with a large density of grain boundaries (GBs) due to facile solution-processed fabrication. GBs exhibit complex atomistic structures undergoing slow rearrangements. By studying evolution of the Σ5(210) CsPbBr₃ GB on a nanosecond time scale, comparable to charge carrier lifetimes, we demonstrate that GB deformations appear every ∼100 ps and increase significantly the probability of deep charge traps. However, the deep traps form only transiently for a few hundred femtoseconds. In contrast, shallow traps appear continuously at the GB. Shallow traps are localized in the GB layer, while deep traps are in a sublayer, which is still distorted from the pristine structure and can be jammed in unfavorable conformations. The GB electronic properties correlate with bond angles, with notable exception of the Br-Br distance, which provides a signature of halide migration along GBs. The transient nature of trap states and localization of electrons and holes at different parts of GBs indicate that charge carrier lifetimes should be long. At the same time, charge mobility can be reduced. The complex, multiscale evolution of geometric and electronic structures of GBs rationalize the contradictory statements made in the literature regarding both benign and detrimental roles of GBs in perovskite performance and provide new atomistic insights into perovskite properties.

Effects of External Perturbations on Protein Systems: A Microscopic View

Protein folding can be viewed as the origami engineering of biology resulting from the long process of evolution. Even decades after its recognition, research efforts worldwide focus on demystifying molecular factors that underlie protein structure-function relationships; this is particularly relevant in the era of proteopathic disease. A complex co-occurrence of different physicochemical factors such as temperature, pressure, solvent, cosolvent, macromolecular crowding, confinement, and mutations that represent realistic biological environments are known to modulate the folding process and protein stability in unique ways. In the current review, we have contextually summarized the substantial efforts in unveiling individual effects of these perturbative factors, with major attention toward bottom-up approaches. Moreover, we briefly present some of the biotechnological applications of the insights derived from these studies over various applications including pharmaceuticals, biofuels, cryopreservation, and novel materials. Finally, we conclude by summarizing the challenges in studying the combined effects of multifactorial perturbations in protein folding and refer to complementary advances in experiment and computational techniques that lend insights to the emergent challenges.

Computational design of magnetic molecules and their environment using quantum chemistry, machine learning and multiscale simulations

Having served as a playground for fundamental studies on the physics of d and f electrons for almost a century, magnetic molecules are now becoming increasingly important for technological applications, such as magnetic resonance, data storage, spintronics and quantum information. All of these applications require the preservation and control of spins in time, an ability hampered by the interaction with the environment, namely with other spins, conduction electrons, molecular vibrations and electromagnetic fields. Thus, the design of a novel magnetic molecule with tailored properties is a formidable task, which does not only concern its electronic structures but also calls for a deep understanding of the interaction among all the degrees of freedom at play. This Review describes how state-of-the-art ab initio computational methods, combined with data-driven approaches to materials modelling, can be integrated into a fully multiscale strategy capable of defining design rules for magnetic molecules.

Electron-Volt Fluctuation of Defect Levels in Metal Halide Perovskites on a 100 ps Time Scale

Metal halide perovskites (MHPs) have gained considerable attention due to their excellent optoelectronic performance, which is often attributed to unusual defect properties. We demonstrate that midgap defect levels can exhibit very large and slow energy fluctuations associated with anharmonic acoustic motions. Therefore, care should be taken classifying MHP defects as deep or shallow, since shallow defects may become deep and vice versa. As a consequence, charges from deep levels can escape into bands, and light absorption can be extended to longer wavelengths, improving material performance. The phenomenon, demonstrated with iodine vacancy in CH₃NH₃PbI₃ using a machine learning force field, can be expected for a variety of defects and dopants in many MHPs and other soft inorganic semiconductors. Since large-scale anharmonic motions can be precursors to chemical decomposition, a known problem with MHPs, we propose that materials that are stiffer than MHPs but softer than traditional inorganic semiconductors, such as Si and TiO₂, may simultaneously exhibit excellent performance and stability.

Materials Data toward Machine Learning: Advances and Challenges

Machine learning (ML) is believed to have enabled a paradigm shift in materials research, and in practice, ML has demonstrated its power in speeding up the cost-efficient discovery of new materials and autonomizing materials laboratories. In this Perspective, current research progress in materials data which are the backbones of ML are reviewed, focusing on high-throughput data generation, standardized data storage, and data representation. More importantly, the challenging issues in materials data that should be overcome to unlock the full potential of ML in materials research and development, including classic 5V (volume, velocity, variety, veracity, and value) issues, 3M (multicomponent, multiscale, and multistage) challenges, co-mining of experimental and computational data, and materials data toward transferable/explainable ML or causal ML, are discussed.

Accurate, affordable, and generalizable machine learning simulations of transition metal x-ray absorption spectra using the XANESNET deep neural network

The affordable, accurate, and generalizable prediction of spectroscopic observables plays a key role in the analysis of increasingly complex experiments. In this article, we develop and deploy a deep neural network-XANESNET-for predicting the lineshape of first-row transition metal K-edge x-ray absorption near-edge structure (XANES) spectra. XANESNET predicts the spectral intensities using only information about the local coordination geometry of the transition metal complexes encoded in a feature vector of weighted atom-centered symmetry functions. We address in detail the calibration of the feature vector for the particularities of the problem at hand, and we explore the individual feature importance to reveal the physical insight that XANESNET obtains at the Fe K-edge. XANESNET relies on only a few judiciously selected features-radial information on the first and second coordination shells suffices along with angular information sufficient to separate satisfactorily key coordination geometries. The feature importance is found to reflect the XANES spectral window under consideration and is consistent with the expected underlying physics. We subsequently apply XANESNET at nine first-row transition metal (Ti-Zn) K-edges. It can be optimized in as little as a minute, predicts instantaneously, and provides K-edge XANES spectra with an average accuracy of ∼±2%-4% in which the positions of prominent peaks are matched with a >90% hit rate to sub-eV (∼0.8 eV) error.

Molecular Simulations to Elucidate Transport Phenomena in Polymeric Membranes

Despite decades of dominance in separation technology, progress in the design and development of high-performance polymer-based membranes has been incremental. Recent advances in materials science and chemical synthesis provide opportunities for molecular-level design of next-generation membrane materials. Such designs necessitate a fundamental understanding of transport and separation mechanisms at the molecular scale. Molecular simulations are important tools that could lead to the development of fundamental structure-property-performance relationships for advancing membrane design. In this Perspective, we assess the application and capability of molecular simulations to understand the mechanisms of ion and water transport across polymeric membranes. Additionally, we discuss the reliability of molecular models in mimicking the structure and chemistry of nanochannels and transport pathways in polymeric membranes. We conclude by providing research directions for resolving key knowledge gaps related to transport phenomena in polymeric membranes and for the construction of structure-property-performance relationships for the design of next-generation membranes.

Interpolating Nonadiabatic Molecular Dynamics Hamiltonian with Inverse Fast Fourier Transform

Nonadiabatic (NA) molecular dynamics (MD) allows one to investigate far-from-equilibrium processes in nanoscale and molecular materials at the atomistic level and in the time domain, mimicking time-resolved spectroscopic experiments. Ab initio NAMD is limited to about 100 atoms and a few picoseconds, due to computational cost of excitation energies and NA couplings. We develop a straightforward methodology that can extend ab initio quality NAMD to nanoseconds and thousands of atoms. The ab initio NAMD Hamiltonian is sampled and interpolated along a trajectory using a Fourier transform, and then, it is used to perform NAMD with known algorithms. The methodology relies on the classical path approximation, which holds for many materials and processes. To achieve a complete ab initio quality description, the trajectory can be obtained using an ab initio trained machine learning force field. The method is demonstrated with charge carrier trapping and relaxation in hybrid organic-inorganic and all-inorganic metal halide perovskites that exhibit complex dynamics and are actively studied for optoelectronic applications.

Challenges and frontiers of computational modelling of biomolecular recognition

Biomolecular recognition including binding of small molecules, peptides and proteins to their target receptors plays a key role in cellular function and has been targeted for therapeutic drug design. However, the high flexibility of biomolecules and slow binding and dissociation processes have presented challenges for computational modelling. Here, we review the challenges and computational approaches developed to characterise biomolecular binding, including molecular docking, molecular dynamics simulations (especially enhanced sampling) and machine learning. Further improvements are still needed in order to accurately and efficiently characterise binding structures, mechanisms, thermodynamics and kinetics of biomolecules in the future.

Accurate Simulations of the Reaction of H2 on a Curved Pt Crystal through Machine Learning

Theoretical studies on molecule-metal surface reactions have so far been limited to small surface unit cells due to computational costs. Here, for the first time molecular dynamics simulations on very large surface unit cells at the level of density functional theory are performed, allowing a direct comparison to experiments performed on a curved crystal. Specifically, the reaction of D2 on a curved Pt crystal is investigated with a neural network potential (NNP). The developed NNP is also accurate for surface unit cells considerably larger than those that have been included in the training data, allowing dynamical simulations on very large surface unit cells that otherwise would have been intractable. Important and complex aspects of the reaction mechanism are discovered such as diffusion and a shadow effect of the step. Furthermore, conclusions from simulations on smaller surface unit cells cannot always be transfered to larger surface unit cells, limiting the applicability of theoretical studies of smaller surface unit cells to heterogeneous catalysts with small defect densities.

Significance of the Chemical Environment of an Element in Nonadiabatic Molecular Dynamics: Feature Selection and Dimensionality Reduction with Machine Learning

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.

Modeling Non-adiabatic Dynamics in Nanoscale and Condensed Matter Systems

Rapid, far-from-equilibrium processes involving excitation of electronic, vibrational, spin, photon, topological, and other degrees of freedom form the basis of modern technologies, including electronics and optoelectronics, solar energy harvesting and conversion to electrical and chemical energy, quantum information processing, spin- and valleytronics, chemical detection, and medical therapies. Such processes are studied experimentally with various time-resolved spectroscopies that allow scientists to track system's evolution on ultrafast time scales and at close to atomistic level of detail. The availability of various forms of lasing has made such measurements easily accessible to many experimental groups worldwide, to study atoms and small molecules, nanoscale and condensed matter systems, proteins, cells, and mesoscopic materials. The experimental work necessitates parallel theoretical efforts needed to interpret the experiments and to provide insights that cannot be gained through measurements due to experimental limitations.Non-adiabatic (NA) molecular dynamics (MD) allows one to study processes at the atomistic level and in the time domain most directly mimicking the time-resolved experiments. Atomistic modeling takes full advantage of chemical intuition and principles that guide design and fabrication of molecules and materials. It provides atomistic origins of quasi-particles, such as holes, excitons, trions, plasmons, phonons, polarons, polaritons, spin-waves, momentum-resolved and topological states, electrically and magnetically polarized structures, and other abstract concepts. An atomistic description enables one to study realistic aspects of materials, which necessarily contain defects, dopants, surfaces, interfaces, passivating ligands, and solvent layers. Often, such realistic features govern material properties and are hard to account for phenomenologically. NA-MD requires few approximations and assumptions. It does not need to assume that atomic motions are harmonic, that electrons are Drude oscillators, that coupling between different degrees of freedom is weak, that dynamics is Markovian or has short memory, or that evolution occurs by exponential kinetics of transitions between few states. The classical or semiclassical treatment of atomic motions constitutes the main approximation of NA-MD and is used because atoms are 3-5 orders of magnitude heavier than electrons. NA-MD is limited by system size, typically hundreds or thousands of atoms, and time scale, picoseconds to nanoseconds. The quality of NA-MD simulations depends on the electronic structure method used to obtain excited state energies and NA couplings.NA-MD has been largely popularized and advanced in the chemistry community that focuses on molecules. Modeling far-from-equilibrium dynamics in nanoscale and condensed matter systems often has to account for other types of physics. At the same time, condensed phase NA-MD allows for approximations that may not work in molecules. Focusing on the recent NA-MD developments aimed at studying excited state processes in nanoscale and condensed phases, this Account considers how the phenomena important on the nanoscale can be incorporated into NA-MD and what approximations can be made to increase its efficiency with complex systems and processes.

Theoretical Study of Chemical Vapor Deposition Synthesis of Graphene and Beyond: Challenges and Perspectives

Two-dimensional (2D) materials have attracted great attention in recent years because of their unique dimensionality and related properties. Chemical vapor deposition (CVD), a crucial technique for thin-film epitaxial growth, has become the most promising method of synthesizing 2D materials. Different from traditional thin-film growth, where strong chemical bonds are involved in both thin films and substrates, the interaction in 2D materials and substrates involves the van der Waals force and is highly anisotropic, and therefore, traditional thin-film growth theories cannot be applied to 2D material CVD synthesis. During the last 15 years, extensive theoretical studies were devoted to the CVD synthesis of 2D materials. This Perspective attempts to present a theoretical framework for 2D material CVD synthesis as well as the challenges and opportunities in exploring CVD mechanisms. We hope that this Perspective can provide an in-depth understanding of 2D material CVD synthesis and can further stimulate 2D material synthesis.