Perspective on integrating machine learning into computational chemistry and materials science

Abstract

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.

arXiv comment22 pages, 5 figures

References in corpus (111)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Automatic chemical design using a data-driven continuous representation of molecules
• Global Optimization by Basin-Hopping and the Lowest Energy Structures of Lennard-Jones Clusters Containing up to 110 Atoms
• Machine learning and the physical sciences
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• SchNet - a deep learning architecture for molecules and materials
• Solving the Quantum Many-Body Problem with Artificial Neural Networks
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Machine Learning Force Fields
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Ab initio Random Structure Searching
• Machine Learning of Accurate Energy-Conserving Molecular Force Fields
• Deep Reinforcement Learning for De-Novo Drug Design
• Machine learning for molecular simulation
• Big Data meets Quantum Chemistry Approximations: The $\Delta$-Machine Learning Approach
• Comparing molecules and solids across structural and alchemical space
• Deep learning for molecular design - a review of the state of the art
• By-passing the Kohn-Sham equations with machine learning
• Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields
• Finding Density Functionals with Machine Learning
• Machine Learning Molecular Dynamics for the Simulation of Infrared Spectra
• VAMPnets: Deep learning of molecular kinetics
• Deep neural network solution of the electronic Schrödinger equation
• Ab-Initio Solution of the Many-Electron Schrödinger Equation with Deep Neural Networks
• How to represent crystal structures for machine learning: towards fast prediction of electronic properties
• SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
• SchNetPack: A Deep Learning Toolbox For Atomistic Systems
• Optimization of Molecules via Deep Reinforcement Learning
• Accelerating crystal structure prediction by machine-learning interatomic potentials with active learning
• Machine learning for electronically excited states of molecules
• i-PI 2.0: A Universal Force Engine for Advanced Molecular Simulations
• Prediction model of band-gap for AX binary compounds by combination of density functional theory calculations and machine learning techniques
• Restricted-Boltzmann-Machine Learning for Solving Strongly Correlated Quantum Systems
• Alchemical and structural distribution based representation for improved QML
• Constrained Graph Variational Autoencoders for Molecule Design
• Chemical Shifts in Molecular Solids by Machine Learning
• Electronic Spectra from TDDFT and Machine Learning in Chemical Space
• A Transferable Machine-Learning Model of the Electron Density
• Equivariant message passing for the prediction of tensorial properties and molecular spectra
• sGDML: Constructing Accurate and Data Efficient Molecular Force Fields Using Machine Learning
• Symmetries and many-body excited states with neural-network quantum states
• OrbNet: Deep Learning for Quantum Chemistry Using Symmetry-Adapted Atomic-Orbital Features
• Transferability in Machine Learning for Electronic Structure via the Molecular Orbital Basis
• Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems
• Solving Many-Electron Schrödinger Equation Using Deep Neural Networks
• Machine learning enables long time scale molecular photodynamics simulations
• Combining SchNet and SHARC: The SchNarc machine learning approach for excited-state dynamics
• Low-Scaling Algorithm for Nudged Elastic Band Calculations Using a Surrogate Machine Learning Model
• A learning scheme to predict atomic forces and accelerate materials simulations
• DeePCG: constructing coarse-grained models via deep neural networks
• Unsupervised machine learning in atomistic simulations, between predictions and understanding
• Kohn-Sham equations as regularizer: building prior knowledge into machine-learned physics
• Efficient global structure optimization with a machine learned surrogate model
• Artificial Neural Network Methods in Quantum Mechanics
• Machine Learning in QM/MM Molecular Dynamics Simulations of Condensed-Phase Systems
• $\Delta$-Machine Learning for Potential Energy Surfaces: A PIP approach to bring a DFT-based PES to CCSD(T) Level of Theory
• Raman Spectrum and Polarizability of Liquid Water from Deep Neural Networks
• Learning the electronic density of states in condensed matter
• Nudged elastic band calculations accelerated with Gaussian process regression
• QM7-X: A comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules
• Efficient Bayesian Inference of Atomistic Structure in Complex Functional Materials
• Molecular geometry prediction using a deep generative graph neural network
• Solving the Bose-Hubbard model with machine learning
• Machine Learning Exciton Dynamics
• Operators in Machine Learning: Response Properties in Chemical Space
• Message-passing neural networks for high-throughput polymer screening
• A Universal Density Matrix Functional from Molecular Orbital-Based Machine Learning: Transferability across Organic Molecules
• Machine Learning the Physical Non-Local Exchange-Correlation Functional of Density-Functional Theory
• Generative Models for Automatic Chemical Design
• Gaussian Process Regression for Transition State Search
• A local Bayesian optimizer for atomic structures
• Fingerprints for spin-selection rules in the interaction dynamics of O2 at Al(111)
• Using Gaussian Process Regression to Simulate the Vibrational Raman Spectra of Molecular Crystals
• Breaking the Coupled Cluster Barrier for Machine Learned Potentials of Large Molecules: The Case of 15-atom Acetylacetone
• Efficient and Accurate Simulations of Vibrational and Electronic Spectra with Symmetry-Preserving Neural Network Models for Tensorial Properties
• Systematic Error Estimation for Chemical Reaction Energies
• Design and Analysis of Machine Learning Exchange-Correlation Functionals via Rotationally Invariant Convolutional Descriptors
• Signatures of nonadiabatic O2 dissociation at Al(111): First-principles fewest-switches study
• Ground state energy functional with Hartree-Fock efficiency and chemical accuracy
• Symmetry-adapted generation of 3d point sets for the targeted discovery of molecules
• Special Topic: Markov Models of Molecular Kinetics
• Deep Learning for UV Absorption Spectra with SchNarc: First Steps Towards Transferability in Chemical Compound Space
• Kinetic energy densities based on the fourth order gradient expansion: performance in different classes of materials and improvement via machine learning
• Physically inspired deep learning of molecular excitations and photoemission spectra
• Symmetry-Adapted High Dimensional Neural Network Representation of Electronic Friction Tensor of Adsorbates on Metals
• Machine learning and excited-state molecular dynamics
• Machine-learning accelerated geometry optimization in molecular simulation
• Ensemble Learning of Coarse-Grained Molecular Dynamics Force Fields with a Kernel Approach
• Inverse Quantum Chemistry: Concepts and Strategies for Rational Compound Design
• Automatically growing global reactive neural network potential energy surfaces: a trajectory free active learning strategy
• High-Dimensional Potential Energy Surfaces for Molecular Simulations
• ELSI -- An Open Infrastructure for Electronic Structure Solvers
• Atomic Permutationally Invariant Polynomials for Fitting Molecular Force Fields
• Improved accuracy and transferability of molecular-orbital-based machine learning: Organics, transition-metal complexes, non-covalent interactions, and transition states
• Machine learning density functional theory for the Hubbard model
• A deep neural network for molecular wave functions in quasi-atomic minimal basis representation
• Equivariant Flows: sampling configurations for multi-body systems with symmetric energies
• Atomistic structure learning
• ML Models of Vibrating H${}_2$CO: Comparing Reproducing Kernels, FCHL and PhysNet
• SAMPLE: Surface structure search enabled by coarse graining and statistical learning
• Atomistic Structure Learning Algorithm with surrogate energy model relaxation
• Long range intermolecular forces in triatomic systems: connecting the atom-diatom and atom-atom-atom representations
• The CECAM Electronic Structure Library and the modular software development paradigm
• Machine Learning Exchange-Correlation Potential in Time Dependent Density Functional Theory
• Machine-learning-corrected quantum dynamics calculations
• Non-parametric Local Pseudopotentials with Machine Learning: a Tin Pseudopotential Built Using Gaussian Process Regression
• Machine Learning with bond information for local structure optimizations in surface science
• Excitonic Wave Function Reconstruction from Near-Field Spectra Using Machine Learning Techniques
• Learning on-top: regressing the on-top pair density for real-space visualization of electron correlation
• DeepDFT: Neural Message Passing Network for Accurate Charge Density Prediction
• MP2 Is Not Good Enough: Transfer Learning ML Models for Accurate VPT2 Frequencies

Cited by in corpus (37)

• Inverse design of 3d molecular structures with conditional generative neural networks
• How to validate machine-learned interatomic potentials
• Advances of Machine Learning in Materials Science: Ideas and Techniques
• SchNetPack 2.0: A neural network toolbox for atomistic machine learning
• Physically inspired deep learning of molecular excitations and photoemission spectra
• Equivariant analytical mapping of first principles Hamiltonians to accurate and transferable materials models
• Permutationally invariant polynomial regression for energies and gradients, using reverse differentiation, achieves orders of magnitude speed-up with high precision compared to other machine learning methods
• Lifelong Machine Learning Potentials
• The MD17 Datasets from the Perspective of Datasets for Gas-Phase "Small" Molecule Potentials
• Towards fully ab initio simulation of atmospheric aerosol nucleation
• Machine Learning and Data-Driven Methods in Computational Surface and Interface Science
• UV-Visible Absorption Spectra of Solvated Molecules by Quantum Chemical Machine Learning
• Ligand additivity relationships enable efficient exploration of transition metal chemical space
• Reliable emulation of complex functionals by active learning with error control
• Exact constraints and appropriate norms in machine learned exchange-correlation functionals
• MolSets: Molecular Graph Deep Sets Learning for Mixture Property Modeling
• Efficient implementation and performance analysis of the independent electron surface hopping method for dynamics at metal surfaces
• Ab initio machine learning of phase space averages
• A machine learning potential for simulating infrared spectra of nanosilicate clusters
• Optimized Multifidelity Machine Learning for Quantum Chemistry
• Prediction rigidities for data-driven chemistry
• Classical and quantum machine learning applications in spintronics
• CoRe Optimizer: An All-in-One Solution for Machine Learning
• Decoding Energy Decomposition Analysis: Machine-Learned Insights on the Impact of the Density Functional on the Bonding Analysis
• Does Hessian Data Improve the Performance of Machine Learning Potentials?
• Stability and distortion of fcc-LaH${}_{10}$ with path-integral molecular dynamics
• Molecular Dipole Moment Learning via Rotationally Equivariant Gaussian Process Regression with Derivatives in Molecular-orbital-based Machine Learning
• Integrated workflows and interfaces for data-driven semi-empirical electronic structure calculations
• Investigating Data Hierarchies in Multifidelity Machine Learning for Excitation Energies
• Lifelong Machine Learning Potentials for Chemical Reaction Network Explorations
• Predicting trajectory behaviour via machine-learned invariant manifolds
• MOLPIPx: an end-to-end differentiable package for permutationally invariant polynomials in Python and Rust
• Translational eigenstates of He@C${}_{60}$ from four-dimensional \textit{ab initio} Potential Energy Surfaces interpolated using Gaussian Process Regression
• Assessing Non-Nested Configurations of Multifidelity Machine Learning for Quantum-Chemical Properties
• Theory of Moment Propagation for Quantum Dynamics in Single-Particle Description
• Exploring exotic configurations with anomalous features using deep learning: Application of classical and quantum-classical hybrid anomaly detection
• NQCDynamics.jl: A Julia Package for Nonadiabatic Quantum Classical Molecular Dynamics in the Condensed Phase