Machine Learning of Accurate Energy-Conserving Molecular Force Fields

Abstract

Using conservation of energy - a fundamental property of closed classical and quantum mechanical systems - we develop an efficient gradient-domain machine learning (GDML) approach to construct accurate molecular force fields using a restricted number of samples from ab initio molecular dynamics (AIMD) trajectories. The GDML implementation is able to reproduce global potential energy surfaces of intermediate-sized molecules with an accuracy of 0.3 kcal ${\text{mol}}^{-1}$ for energies and 1 kcal ${\text{mol}}^{-1}$ ${\text{A˚}}^{-1}$ for atomic forces using only 1000 conformational geometries for training. We demonstrate this accuracy for AIMD trajectories of molecules, including benzene, toluene, naphthalene, ethanol, uracil, and aspirin. The challenge of constructing conservative force fields is accomplished in our work by learning in a Hilbert space of vector-valued functions that obey the law of energy conservation. The GDML approach enables quantitative molecular dynamics simulations for molecules at a fraction of cost of explicit AIMD calculations, thereby allowing the construction of efficient force fields with the accuracy and transferability of high-level ab initio methods.

References in corpus (1)

• Representing molecule-surface interactions with symmetry-adapted neural networks

Cited by in corpus (287)

• Methods for Interpreting and Understanding Deep Neural Networks
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• SchNet - a deep learning architecture for molecules and materials
• DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics
• E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials
• Machine Learning Force Fields
• Recent Advances and Applications of Deep Learning Methods in Materials Science
• PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
• Unmasking Clever Hans Predictors and Assessing What Machines Really Learn
• A Universal Graph Deep Learning Interatomic Potential for the Periodic Table
• A Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Machine learning for molecular simulation
• Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems
• DScribe: Library of Descriptors for Machine Learning in Materials Science
• By-passing the Kohn-Sham equations with machine learning
• Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields
• Big-Data Science in Porous Materials: Materials Genomics and Machine Learning
• DeePMD-kit v2: A software package for Deep Potential models
• Physics-inspired structural representations for molecules and materials
• Active Learning of Uniformly Accurate Inter-atomic Potentials for Materials Simulation
• Machine Learning Molecular Dynamics for the Simulation of Infrared Spectra
• Deep neural network solution of the electronic Schrödinger equation
• GPUMD: A package for constructing accurate machine-learned potentials and performing highly efficient atomistic simulations
• SchNetPack: A Deep Learning Toolbox For Atomistic Systems
• Towards Universal Neural Network Potential for Material Discovery Applicable to Arbitrary Combination of 45 Elements
• Hamiltonian Neural Networks
• Machine learning for electronically excited states of molecules
• SpookyNet: Learning Force Fields with Electronic Degrees of Freedom and Nonlocal Effects
• ANI-1: A data set of 20M off-equilibrium DFT calculations for organic molecules
• FCHL revisited: faster and more accurate quantum machine learning
• The Open Catalyst 2022 (OC22) Dataset and Challenges for Oxide Electrocatalysts
• Neural Network Based in Silico Simulation of Combustion Reactions
• Hierarchical modeling of molecular energies using a deep neural network
• Unified Representation of Molecules and Crystals for Machine Learning
• Embedded Atom Neural Network Potentials: Efficient and Accurate Machine Learning with a Physically Inspired Representation
• sGDML: Constructing Accurate and Data Efficient Molecular Force Fields Using Machine Learning
• 86 PFLOPS Deep Potential Molecular Dynamics simulation of 100 million atoms with ab initio accuracy
• A deep potential model with long-range electrostatic interactions
• Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems
• De novo exploration and self-guided learning of potential-energy surfaces
• Accurate Interatomic Force Fields via Machine Learning with Covariant Kernels
• Incorporating long-range physics in atomic-scale machine learning
• Perspective on integrating machine learning into computational chemistry and materials science
• A practical guide to machine learning interatomic potentials -- Status and future
• On the Completeness of Atomic Structure Representations
• Coarse Graining Molecular Dynamics with Graph Neural Networks
• DeePCG: constructing coarse-grained models via deep neural networks
• Committee neural network potentials control generalization errors and enable active learning
• Inclusion of machine learning kernel ridge regression potential energy surfaces in on-the-fly nonadiabatic molecular dynamics simulation
• Atom-Density Representations for Machine Learning
• Coarse-Graining Auto-Encoders for Molecular Dynamics
• Non-covalent interactions across organic and biological subsets of chemical space: Physics-based potentials parametrized from machine learning
• Efficient global structure optimization with a machine learned surrogate model
• Structure-based Sampling and Self-correcting Machine Learning for Accurate Calculations of Potential Energy Surfaces and Vibrational Levels
• Machine Learning in QM/MM Molecular Dynamics Simulations of Condensed-Phase Systems
• Deep neural network for the dielectric response of insulators
• Raman Spectrum and Polarizability of Liquid Water from Deep Neural Networks
• Feature Optimization for Atomistic Machine Learning Yields A Data-Driven Construction of the Periodic Table of the Elements
• Automated Fitting of Neural Network Potentials at Coupled Cluster Accuracy: Protonated Water Clusters as Testing Ground
• Neural Network Potentials for Chemistry: Concepts, Applications and Prospects
• DPA-2: a large atomic model as a multi-task learner
• Quantum-Accurate Spectral Neighbor Analysis Potential Models for Ni-Mo Binary Alloys and FCC Metals
• AdsorbML: A Leap in Efficiency for Adsorption Energy Calculations using Generalizable Machine Learning Potentials
• A Deep Dive into Machine Learning Density Functional Theory for Materials Science and Chemistry
• Operators in Machine Learning: Response Properties in Chemical Space
• Molecular Force Fields with Gradient-Domain Machine Learning: Construction and Application to Dynamics of Small Molecules with Coupled Cluster Forces
• Toward Explainable AI for Regression Models
• The Many-Body Expansion Combined with Neural Networks
• Training neural nets to learn reactive potential energy surfaces using interactive quantum chemistry in virtual reality
• Classification of Local Chemical Environments from X-ray Absorption Spectra using Supervised Machine Learning
• Gaussian Moments as Physically Inspired Molecular Descriptors for Accurate and Scalable Machine Learning Potentials
• Diffusion Models in $\text{De Novo}$ Drug Design
• OrbNet Denali: A machine learning potential for biological and organic chemistry with semi-empirical cost and DFT accuracy
• Learning the exchange-correlation functional from nature with fully differentiable density functional theory
• Informing Geometric Deep Learning with Electronic Interactions to Accelerate Quantum Chemistry
• Strategies for the Construction of Machine-Learning Potentials for Accurate and Efficient Atomic-Scale Simulations
• Single-model uncertainty quantification in neural network potentials does not consistently outperform model ensembles
• DeePKS: a comprehensive data-driven approach towards chemically accurate density functional theory
• SchNetPack 2.0: A neural network toolbox for atomistic machine learning
• Error-Controlled Exploration of Chemical Reaction Networks with Gaussian Processes
• First principles simulations of dense hydrogen
• Solid Harmonic Wavelet Scattering for Predictions of Molecule Properties
• Sparse Gaussian Process Potentials: Application to Lithium Diffusivity in Superionic Conducting Solid Electrolytes
• Design and Analysis of Machine Learning Exchange-Correlation Functionals via Rotationally Invariant Convolutional Descriptors
• Deep learning inter-atomic potential model for accurate irradiation damage simulations
• Graph Neural Networks for Molecules
• Metadynamics for Training Neural Network Model Chemistries: a Competitive Assessment
• Ground state energy functional with Hartree-Fock efficiency and chemical accuracy
• Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules
• BIGDML: Towards Exact Machine Learning Force Fields for Materials
• Global optimization of atomistic structure enhanced by machine learning
• Isotope Effects in Liquid Water via Deep Potential Molecular Dynamics
• AI-driven inverse design of materials: Past, present and future
• Neural networks and kernel ridge regression for excited states dynamics of CH${}_2$NH${}_2^{+}$: From single-state to multi-state representations and multi-property machine learning models
• Building and Interpreting Deep Similarity Models
• Physically inspired deep learning of molecular excitations and photoemission spectra
• Direct prediction of phonon density of states with Euclidean neural networks
• The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
• Graph Neural Networks Accelerated Molecular Dynamics
• Universal QM/MM Approaches for General Nanoscale Applications
• Ensemble Learning of Coarse-Grained Molecular Dynamics Force Fields with a Kernel Approach
• JARVIS-Leaderboard: A Large Scale Benchmark of Materials Design Methods
• TorchMD-Net 2.0: Fast Neural Network Potentials for Molecular Simulations
• DIG: A Turnkey Library for Diving into Graph Deep Learning Research
• A Perspective on Deep Learning for Molecular Modeling and Simulations
• High-Dimensional Potential Energy Surfaces for Molecular Simulations
• Deep Ensembles vs. Committees for Uncertainty Estimation in Neural-Network Force Fields: Comparison and Application to Active Learning
• Permutationally Invariant, Reproducing Kernel-Based Potential Energy Surfaces for Polyatomic Molecules: From Formaldehyde to Acetone
• Simulations meet Machine Learning in Structural Biology
• Molecular Force Fields with Gradient-Domain Machine Learning (GDML): Comparison and Synergies with Classical Force Fields
• Atomic Permutationally Invariant Polynomials for Fitting Molecular Force Fields
• Challenges for Machine Learning Force Fields in Reproducing Potential Energy Surfaces of Flexible Molecules
• Development of Interatomic Potential for Al-Tb Alloy by Deep Neural Network Learning Method
• Rotation Invariant Graph Neural Networks using Spin Convolutions
• Group-theoretical high-order rotational invariants for structural representations: Application to linearized machine learning interatomic potential
• Dynamical Strengthening of Covalent and Non-Covalent Molecular Interactions by Nuclear Quantum Effects at Finite Temperature
• Convolutional neural networks for atomistic systems
• Efficient implementation of atom-density representations
• A deep neural network for molecular wave functions in quasi-atomic minimal basis representation
• Transfer learning for chemically accurate interatomic neural network potentials
• Scalable Hybrid Deep Neural Networks/Polarizable Potentials Biomolecular Simulations including long-range effects
• Fast Neural Network Approach for Direct Covariant Forces Prediction in Complex Multi-Element Extended Systems
• Forecasting the outcome of spintronic experiments with Neural Ordinary Differential Equations
• Predicting the failure of two-dimensional silica glasses
• Predicting Properties of Periodic Systems from Cluster Data: A Case Study of Liquid Water
• Global optimization of atomic structures with gradient-enhanced Gaussian process regression
• Combining phonon accuracy with high transferability in Gaussian approximation potential models
• Universal Machine Learning Kohn-Sham Hamiltonian for Materials
• Fast and Sample-Efficient Interatomic Neural Network Potentials for Molecules and Materials Based on Gaussian Moments
• Overcoming the Barrier of Orbital-Free Density Functional Theory for Molecular Systems Using Deep Learning
• Towards Linearly Scaling and Chemically Accurate Global Machine Learning Force Fields for Large Molecules
• Atomistic structure learning
• Simulating solvation and acidity in complex mixtures with first-principles accuracy: the case of CH${}_3$SO${}_3$H and H${}_2$O${}_2$ in phenol
• A Tungsten Deep Neural-Network Potential for Simulating Mechanical Property Degradation Under Fusion Service Environment
• Automated construction of quantum-classical hybrid models
• Non-classical nucleation pathways in stacking-disordered crystals
• Gaussian process model of 51-dimensional potential energy surface for protonated imidazole dimer
• Self-Parametrizing System-Focused Atomistic Models
• Exploring a potential energy surface by machine learning for characterizing atomic transport
• Pushing the limit of molecular dynamics with ab initio accuracy to 100 million atoms with machine learning
• The MD17 Datasets from the Perspective of Datasets for Gas-Phase "Small" Molecule Potentials
• Hydration free energies from kernel-based machine learning: Compound-database bias
• Learning from the Density to Correct Total Energy and Forces in First Principle Simulations
• On the design space between molecular mechanics and machine learning force fields
• Exponential data encoding for quantum supervised learning
• Progress, challenges and perspectives of computational studies on glassy superionic conductors for solid-state batteries
• Spherical Message Passing for 3D Graph Networks
• Symmetrical Graph Neural Network for Quantum Chemistry, with Dual R/K Space
• Transferability of machine learning potentials: Protonated water neural network potential applied to the protonated water hexamer
• Multilayer atomic cluster expansion for semi-local interactions
• PhysNet Meets CHARMM: A Framework for Routine Machine Learning / Molecular Mechanics Simulations
• Arrested phase separation in double-exchange models: machine-learning enabled large-scale simulation
• Thermally Averaged Magnetic Anisotropy Tensors via Machine Learning Based on Gaussian Moments
• An Atomistic Fingerprint Algorithm for Learning Ab Initio Molecular Force Fields
• Ab Initio Molecular Dynamics on Quantum Computers
• Model-free quantification of completeness, uncertainties, and outliers in atomistic machine learning using information theory
• Metadynamics sampling in atomic environment space for collecting training data for machine learning potentials
• An Extendible, Graph-Neural-Network-Based Approach for Accurate Force Field Development of Large Flexible Organic Molecules
• Machine Learning and Data-Driven Methods in Computational Surface and Interface Science
• Neural Canonical Transformation with Symplectic Flows
• Denoise Pretraining on Nonequilibrium Molecules for Accurate and Transferable Neural Potentials
• Training atomic neural networks using fragment-based data generated in virtual reality
• Navigating chemical reaction space with a steering wheel
• ForceNet: A Graph Neural Network for Large-Scale Quantum Calculations
• Atomistic structure search using local surrogate mode
• Machine learning electron correlation in a disordered medium
• Geometry-enhanced Pre-training on Interatomic Potentials
• Reliable emulation of complex functionals by active learning with error control
• Modeling intermolecular and intramolecular modes of liquid water using multiple heat baths: Machine learning approach
• Neural Network Potential with Multi-Resolution Approach Enables Accurate Prediction of Reaction Free Energies in Solution
• Construction of coarse-grained molecular dynamics with many-body non-Markovian memory
• Machine learning nonequilibrium electron forces for adiabatic spin dynamics
• Uncertainty quantification and estimation in differential dynamic microscopy
• 3DReact: Geometric deep learning for chemical reactions
• Machine learning density functionals from the random-phase approximation
• Kernel based quantum machine learning at record rate : Many-body distribution functionals as compact representations
• Measuring transferability issues in machine-learning force fields: The example of Gold-Iron interactions with linearized potentials
• Data efficiency and extrapolation trends in neural network interatomic potentials
• Atomic structure optimization with machine-learning enabled interpolation between chemical elements
• Hybrid Classical/Machine-Learning Force Fields for the Accurate Description of Molecular Condensed-Phase Systems
• An Introduction to Electrocatalyst Design using Machine Learning for Renewable Energy Storage
• Determining electronic properties from L-edge X-ray absorption spectra of transition metal compounds with artificial neural networks
• MGNN: Moment Graph Neural Network for Universal Molecular Potentials
• Accurate quantum Monte Carlo forces for machine-learned force fields: Ethanol as a benchmark
• Machine Learning with bond information for local structure optimizations in surface science
• Adaptive coupling of a deep neural network potential to a classical force field
• Molecular Quantum Chemical Data Sets and Databases for Machine Learning Potentials
• Ab initio machine learning of phase space averages
• A transferable artificial neural network model for atomic forces in nanoparticles
• Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
• Optimized Multifidelity Machine Learning for Quantum Chemistry
• Machine learning for phase ordering dynamics of charge density waves
• GPU-Accelerated Approximate Kernel Method for Quantum Machine Learning
• Analyzing Atomic Interactions in Molecules as Learned by Neural Networks
• Hydrogen-induced degradation dynamics in silicon heterojunction solar cells via machine learning
• Wavelet Scattering Networks for Atomistic Systems with Extrapolation of Material Properties
• Algorithmic Differentiation for Automated Modeling of Machine Learned Force Fields
• Analytical Gradients for Molecular-Orbital-Based Machine Learning
• No Headache for PIPs: A PIP Potential for Aspirin Outperforms Other Machine-Learned Potentials
• Machine-learning of atomic-scale properties based on physical principles
• Mean-Field Density Matrix Decompositions
• Deep Density: circumventing the Kohn-Sham equations via symmetry preserving neural networks
• Super-resolution in Molecular Dynamics Trajectory Reconstruction with Bi-Directional Neural Networks
• Beyond MD17: the reactive xxMD dataset
• Transferable empirical pseudopotenials from machine learning
• Improved decision making with similarity based machine learning: Applications in chemistry
• Construction of Machine Learned Force Fields with Quantum Chemical Accuracy: Applications and Chemical Insights
• Building nonparametric $n$-body force fields using Gaussian process regression
• Training models using forces computed by stochastic electronic structure methods
• Low-data deep quantum chemical learning for accurate MP2 and coupled-cluster correlations
• Deep-learning-based prediction of the tetragonal$\to$cubic transition in davemaoite
• Machine learning modeling of high entropy alloy: the role of short-range order
• Efficient force field and energy emulation through partition of permutationally equivalent atoms
• Accurate Molecular Dynamics Enabled by Efficient Physically-Constrained Machine Learning Approaches
• Machine learning potentials for complex aqueous systems made simple
• SPA${}^{\mathrm{H}}$M(a,b): encoding the density information from guess Hamiltonian in quantum machine learning representations
• Machine learning at the atomic-scale
• Gradient domain machine learning with composite kernels: improving the accuracy of PES and force fields for large molecules
• Machine learning for accuracy in density functional approximations
• Equivariant neural network for Green's functions of molecules and materials
• Crystalline Material Discovery in the Era of Artificial Intelligence
• Does Hessian Data Improve the Performance of Machine Learning Potentials?
• Gated Graph Recursive Neural Networks for Molecular Property Prediction
• Robust data-driven approach for predicting the configurational energy of high entropy alloys
• Tell machine learning potentials what they are needed for: Simulation-oriented training exemplified for glycine
• Compressing physical properties of atomic species for improving predictive chemistry
• End-to-end AI framework for interpretable prediction of molecular and crystal properties
• Machine Learning Framework for Modeling Exciton-Polaritons in Molecular Materials
• Molecular Dipole Moment Learning via Rotationally Equivariant Gaussian Process Regression with Derivatives in Molecular-orbital-based Machine Learning
• Aqueous Solution Chemistry In Silico and the Role of Data Driven Approaches
• Molecular relaxation by reverse diffusion with time step prediction
• HamNet: Conformation-Guided Molecular Representation with Hamiltonian Neural Networks
• Compressing and forecasting atomic material simulations with descriptors
• Scaling Graph Neural Networks to Large Proteins
• Role of long-range dipolar interactions in the simulation of the properties of polar crystals using effective atomic potentials
• A neural network potential with self-trained atomic fingerprints: a test with the mW water potential
• Deep Learning Inter-atomic Potential for Thermal and Phonon Behaviour of Silicon Carbide with Quantum Accuracy
• Learning Potentials of Quantum Systems using Deep Neural Networks
• Inversion of the chemical environment representations
• Steerable Wavelet Scattering for 3D Atomic Systems with Application to Li-Si Energy Prediction
• Investigation of Machine Learning-based Coarse-Grained Mapping Schemes for Organic Molecules
• Kinetics of orbital ordering in cooperative Jahn-Teller models: Machine-learning enabled large-scale simulations
• Molecular Hessian matrices from a machine learning random forest regression algorithm
• Free energy profiles for chemical reactions in solution from high-dimensional neural network potentials: The case of the Strecker synthesis
• Physical machine learning outperforms "human learning" in Quantum Chemistry
• Selecting the independent coordinates of manifolds with large aspect ratios
• Symmetry- and Gradient-enhanced Gaussian Process Regression for the Active Learning of Potential Energy Surfaces in Porous Materials
• Ab initio machine learning in chemical compound space
• A Bond-Based Machine Learning Model for Molecular Polarizabilities and A Priori Raman Spectra
• MOLPIPx: an end-to-end differentiable package for permutationally invariant polynomials in Python and Rust
• Enhancing Brain Source Reconstruction by Initializing 3D Neural Networks with Physical Inverse Solutions
• Unifying O(3) Equivariant Neural Networks Design with Tensor-Network Formalism
• Recurrent Neural Network-based Model for Accelerated Trajectory Analysis in AIMD Simulations
• Deep quantum Monte Carlo approach for polaritonic chemistry
• Modeling and simulating the excited-state dynamics of a system with condensed phases: A machine learning approach
• Towards Improved Quantum Machine Learning for Molecular Force Fields
• GeoT: A Geometry-aware Transformer for Reliable Molecular Property Prediction and Chemically Interpretable Representation Learning
• Flexible dual-branched message passing neural network for quantum mechanical property prediction with molecular conformation
• Orbital Mixer: Using Atomic Orbital Features for Basis Dependent Prediction of Molecular Wavefunctions
• Machine-Learned Potentials for Solvation Modeling
• Selecting fitted models under epistemic uncertainty using a stochastic process on quantile functions
• Reconstructing Kernel-based Machine Learning Force Fields with Super-linear Convergence
• An Asympirical Smoothing Parameters Selection Approach for Smoothing Spline ANOVA Models in Large Samples
• Emulating the First Principles of Matter: A Probabilistic Roadmap
• Machine-learning force-field models for dynamical simulations of metallic magnets
• Fast and Accurate Explanations of Distance-Based Classifiers by Uncovering Latent Explanatory Structures
• Scalable Training of Neural Network Potentials for Complex Interfaces Through Data Augmentation
• Efficient interpolation of molecular properties across chemical compound space with low-dimensional descriptors
• Self-consistency error correction for accurate machine learning potentials from variational Monte Carlo
• The decomposition of the higher-order homology embedding constructed from the $k$-Laplacian
• Enhancing ab initio diffusion calculations in materials through Gaussian process regression
• Machine learning dynamics of phase separation in correlated electron magnets
• Towards Symmetry-Aware Efficient Simulation of Quantum Systems and Beyond
• Charge Transfer in Classical Molecular Dynamics Simulations of Met-enkephalin: Improving Traditional Force Field with Data Driven Models
• Geometric Algebra Attention Networks for Small Point Clouds
• Simplifying inverse material design problems for fixed lattices with alchemical chirality
• Machine Learning the Effective Hamiltonian in High Entropy Alloys
• Bond Energies from a Diatomics-in-Molecules Neural Network
• Distribution free optimality intervals for clustering
• Neural Polarization: Toward Electron Density for Molecules by Extending Equivariant Networks
• Detect the Interactions that Matter in Matter: Geometric Attention for Many-Body Systems
• Manifold learning for coarse-graining atomistic simulations: Application to amorphous solids
• On the forbidden graphene's ZO (out-of-plane optic) phononic band-analog vibrational modes in fullerenes
• Machine learning approach for vibronically renormalized electronic band structures
• Machine-learning modeling of magnetization dynamics in quasi-equilibrium and driven metallic spin systems
• Improving Molecular Force Fields Across Configurational Space by Combining Supervised and Unsupervised Machine Learning
• Learning Atomic Multipoles: Prediction of the Electrostatic Potential with Equivariant Graph Neural Networks