Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems

2021/02/12 by John A. Keith, Valentin Vassilev-Galindo, Valentín Vassilev-Galindo +6 · 40 citations
Chemical Engineering · Computer Science · Materials Science · Physics and Astronomy · #Catalysis and Oxidation Reactions #Computational Drug Discovery Methods #Machine Learning in Materials Science #physics.chem-ph #physics.comp-ph

paper · pdf · doi:10.1021/acs.chemrev.1c00107

169 pages, 24 figures

arxiv created 2021/02/12 · openalex created_date 2021/03/01 · openalex publication_date 2021/07/07 · arxiv updated 2021/10/14 · openalex updated_date 2026/07/30

Abstract

Machine learning models are poised to make a transformative impact on chemical sciences by dramatically accelerating computational algorithms and amplifying insights available from computational chemistry methods. However, achieving this requires a confluence and coaction of expertise in computer science and physical sciences. This review is written for new and experienced researchers working at the intersection of both fields. We first provide concise tutorials of computational chemistry and machine learning methods, showing how insights involving both can be achieved. We then follow with a critical review of noteworthy applications that demonstrate how computational chemistry and machine learning can be used together to provide insightful (and useful) predictions in molecular and materials modeling, retrosyntheses, catalysis, and drug design.

Citations

• Statistical Mechanics: Theory and Molecular Simulation
• Nonphysical sampling distributions in Monte Carlo free-energy estimation: Umbrella sampling
• Watson: Beyond Jeopardy!
• The perceptron: A probabilistic model for information storage and organization in the brain.
• Deep Learning in Neural Networks: An Overview
• Fully optimized contracted Gaussian basis sets for atoms Li to Kr
• Ab initioeffective potentials for use in molecular quantum mechanics
• TRANSITIONPATHSAMPLING: Throwing Ropes Over Rough Mountain Passes, in the Dark
• Optimizing Spatial filters for Robust EEG Single-Trial Analysis
• A Practical Bayesian Framework for Backpropagation Networks
• Multilayer feedforward networks are universal approximators
• Progress in Time-Dependent Density-Functional Theory
• Density-functional thermochemistry. III. The role of exact exchange
• Development and testing of a general amber force field
• Zur Quantentheorie der Molekeln
• DeepWalk
• Crystal structure representations for machine learning models of formation energies
• Quantum-chemical insights from deep tensor neural networks
• Automatic chemical design using a data-driven continuous representation of molecules
• Methods for Interpreting and Understanding Deep Neural Networks
• SchNet – A deep learning architecture for molecules and materials
• UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
• DNA methylation-based classification of central nervous system tumours
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• Operators in quantum machine learning: Response properties in chemical space
• Learning Multimodal Graph-to-Graph Translation for Molecular Optimization
• On the Inductive Bias of Neural Tangent Kernels
• Machine Learning for Molecular Simulation
• Improved protein structure prediction using potentials from deep learning
• A Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Conceptual density functional theory: status, prospects, issues
• QSAR without borders
• Accelerated discovery of CO2 electrocatalysts using active machine learning
• The devil in the details: A tutorial review on some undervalued aspects of density functional theory calculations
• A Unifying Review of Deep and Shallow Anomaly Detection
• Cohesion

Cited by

• A machine-learning framework for accelerating spin-lattice relaxation simulations
• Molecular Simulations with a Pretrained Neural Network and Universal Pairwise Force Fields
• Uncertainty Quantification for In Silico Chemistry
• Condensed‐Phase Quantum Chemistry
• Uncertainty Calibration in Molecular Machine Learning: Comparing Evidential and Ensemble Approaches
• Lost Data in Electron Microscopy
• Cross-disciplinary perspectives on the potential for artificial intelligence across chemistry
• Three decades of quantum science: how quantum chemistry transformed thermochemical database generation for benchmarking DFT and machine learning
• Mechanical properties of AlCoCrCuFeNi high-entropy alloys using molecular dynamics and machine learning
• High-Throughput Microscale MOF Catalysis via Solid Dispersion Reactor Dosing
• Assessing the performance of quantum-mechanical descriptors in physicochemical and biological property prediction
• In Silico Therapeutic Study: The Next Frontier in the Fight against SARS-CoV-2 and Its Variants
• Are neural scaling laws leading quantum chemistry astray?
• Discovery of hydrogen storage molecules using large language models and machine learning
• Optimal Sampling Density for Nonparametric Regression
• Machine learning-driven breakthroughs in water electrolysis and supercapacitors
• Electronic structure simulations in the cloud computing environment
• Response to Promises and Pitfalls of Deep Kernel Learning
• Biomolecular dynamics with machine-learned quantum-mechanical force fields trained on diverse chemical fragments
• Ion-modulated structure, proton transfer, and capacitance in the Pt(111)/water electric double layer
• Meta-Learning Linear Models for Molecular Property Prediction
• Fast and Interpretable Machine Learning Modelling of Atmospheric Molecular Clusters
• Neural Network Solution of Non-Markovian Quantum State Diffusion and Operator Construction of Quantum Stochastic Process
• Representative Random Sampling of Chemical Space
• Equivariant Electronic Hamiltonian Prediction with Many-Body Message Passing
• Topological Feature Compression for Molecular Graph Neural Networks
• The emergence of machine learning force fields in drug design
• A universal augmentation framework for long-range electrostatics in machine learning interatomic potentials
• Implicit Machine Learning Force Fields Accelerate Molecular Dynamics Simulations
• Intrinsic Dimensionality of Molecular Properties
• Machine learning prediction of a chemical reaction over 8 decades of energy
• Quantifying Gibbs measures of disordered crystals up to the solid-liquid phase transition
• Farthest Point Sampling in Property Designated Chemical Feature Space as a General Strategy for Enhancing the Machine Learning Model Performance for Small Scale Chemical Dataset
• Towards Desiderata-Driven Design of Visual Counterfactual Explainers
• A Multi-Scale Quantum Framework for Evaluating Metal-Organic Frameworks in Carbon Capture
• Self-Refining Training for Amortized Density Functional Theory
• Machine-Learned Potentials for Solvation Modeling
• The informativeness of the gradient revisited
• Advancing the Scientific Method with Large Language Models: From Hypothesis to Discovery
• Learning Hamiltonian Flow Maps: Mean Flow Consistency for Large-Timestep Molecular Dynamics

Related

• Advances of Machine Learning in Molecular Modeling and Simulation
• From 2015 to 2023: How Machine Learning Aids Natural Product Analysis
• Molecular Machine Learning in Chemical Process Design
• On the use of chemical bonding descriptors in machine learning
• Learning to Make Chemical Predictions: the Interplay of Feature Representation, Data, and Machine Learning Algorithms