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

Published in

Chemical Reviews(2021)

External Links:

Generate Graph DownloadPDF

TL;DR

A critical review of noteworthy applications that demonstrate how computational chemistry and machine learning can be used together to provide insightful predictions in molecular and materials modeling, retrosyntheses, catalysis, and drug design are reviewed.

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 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.

Authors

A. Tkatchenko

11 papers

M. Gastegger

4 papers

Stefan Chmiela

7 papers

References674 items

Morphological and molecular breast cancer profiling through explainable machine learning

Explaining Deep Neural Networks and Beyond: A Review of Methods and Applications

General-Purpose Machine Learning Potentials Capturing Nonlocal Charge Transfer.

When do short-range atomistic machine-learning models fall short?

Physics-Inspired Structural Representations for Molecules and Materials.

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

Published in

Chemical Reviews(2021)

External Links:

Generate Graph DownloadPDF

TL;DR

Abstract

Authors

A. Tkatchenko

11 papers

M. Gastegger

4 papers

Stefan Chmiela

7 papers

References674 items

Morphological and molecular breast cancer profiling through explainable machine learning

Explaining Deep Neural Networks and Beyond: A Review of Methods and Applications

General-Purpose Machine Learning Potentials Capturing Nonlocal Charge Transfer.

When do short-range atomistic machine-learning models fall short?

Physics-Inspired Structural Representations for Molecules and Materials.

V. Vassilev-Galindo

2 papers

Bingqing Cheng

1 papers

Klaus-Robert Müller

3 papers

Two-temperature warm dense hydrogen as a test of quantum protons driven by orbital-free density functional theory electronic forces

Transferable MP2-Based Machine Learning for Accurate Coupled-Cluster Energies.

Quantum-mechanical exploration of the phase diagram of water

Quantum chemical accuracy from density functional approximations via machine learning

Machine Learning Force Fields

An Introduction to Electrocatalyst Design using Machine Learning for Renewable Energy Storage

Quantifying Uncertainties in Solvation Procedures for Modeling Aqueous Phase Reaction Mechanisms.

Machine learning corrected alchemical perturbation density functional theory for catalysis applications

Retrospective on a decade of machine learning for chemical discovery

A Unifying Review of Deep and Shallow Anomaly Detection

An adaptive deep reinforcement learning framework enables curling robots with human-like performance in real-world conditions

A fourth-generation high-dimensional neural network potential with accurate electrostatics including non-local charge transfer

Quantum machine learning using atom-in-molecule-based fragments selected on the fly

Spectrum of deep learning algorithms in drug discovery

Multi-scale approach for the prediction of atomic scale properties

The 2020 plasma catalysis roadmap

Machine learning for chemical discovery

Mapping Materials and Molecules.

Machine Learning Approaches toward Orbital-free Density Functional Theory: Simultaneous Training on the Kinetic Energy Density Functional and Its Functional Derivative

Mapping the space of chemical reactions using attention-based neural networks

The Ground State Electronic Energy of Benzene.

Screening of Therapeutic Agents for COVID-19 Using Machine Learning and Ensemble Docking Studies

Challenge to scientists: does your ten-year-old code still run?

TeraChem: A graphical processing unit‐accelerated electronic structure package for large‐scale ab initio molecular dynamics

Machine learning the ropes: principles, applications and directions in synthetic chemistry.

The Advent of Generative Chemistry.

Chemiscope: interactive structure-property explorer for materials and molecules

Drug discovery with explainable artificial intelligence

QM7-X: A comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules

An accurate and transferable machine learning potential for carbon.

Extracting ice phases from liquid water: why a machine-learning water model generalizes so well.

Dynamical strengthening of covalent and non-covalent molecular interactions by nuclear quantum effects at finite temperature

Accurate Many-Body Repulsive Potentials for Density-Functional Tight Binding from Deep Tensor Neural Networks.

Data-Driven Approaches Can Overcome the Cost-Accuracy Trade-off in Multireference Diagnostics.

XAI for Graphs: Explaining Graph Neural Network Predictions by Identifying Relevant Walks

Hierarchically Structured Allotropes of Phosphorus from Data‐Driven Exploration

Ab initio phase diagram and nucleation of gallium

High-throughput experimentation meets artificial intelligence: a new pathway to catalyst discovery.

Ensemble learning of coarse-grained molecular dynamics force fields with a kernel approach.

Accelerated discovery of CO2 electrocatalysts using active machine learning

QSAR without borders.

Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems.

Multicomponent Quantum Chemistry: Integrating Electronic and Nuclear Quantum Effects via the Nuclear-Electronic Orbital Method.

The devil in the details: A tutorial review on some undervalued aspects of density functional theory calculations

REINVENT 2.0 – an AI Tool for De Novo Drug Design

First-principles modeling of chemistry in mixed solvents: Where to go from here?

Autonomous discovery in the chemical sciences part II: Outlook

Autonomous discovery in the chemical sciences part I: Progress

AiiDA 1.0, a scalable computational infrastructure for automated reproducible workflows and data provenance

Directional Message Passing for Molecular Graphs

Chemical product design – recent advances and perspectives

Advances and challenges in modeling solvated reaction mechanisms for renewable fuels and chemicals

Autonomous robotic nanofabrication with reinforcement learning

Data-Driven Collective Variables for Enhanced Sampling.

Predictive Multivariate Models for Bioorthogonal Inverse-Electron Demand Diels-Alder Reactions.

Machine Learning for Catalysis Informatics: Recent Applications and Prospects

Intelligent-ReaxFF: Evaluating the reactive force field parameters with machine learning

Conceptual density functional theory: status, prospects, issues

On the Completeness of Atomic Structure Representations

Machine learning approaches for ELNES/XANES.

Wasserstein metric for improved quantum machine learning with adjacency matrix representations

Discovery of novel chemical reactions by deep generative recurrent neural network

Improved protein structure prediction using potentials from deep learning

Automation of Active Space Selection for Multireference Methods via Machine Learning on Chemical Bond Dissociation.

Fast predictions of liquid-phase acid-catalyzed reaction rates using molecular dynamics simulations and convolutional neural networks†

Exploring chemical compound space with quantum-based machine learning

Machine learning for molecular simulation

The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules

The Lennard-Jones potential: when (not) to use it.

PiNN: A Python Library for Building Atomic Neural Networks of Molecules and Materials

Machine Learning in Computational Chemistry: An Evaluation of Method Performance for Nudged Elastic Band Calculations.

Predicting the phase diagram of titanium dioxide with random search and pattern recognition.

Gaussian Mixture Based Enhanced Sampling For Statics And Dynamics.

Deep-neural-network solution of the electronic Schrödinger equation

Machine learning analysis of DNA methylation profiles distinguishes primary lung squamous cell carcinomas from head and neck metastases

Incorporating long-range physics in atomic-scale machine learning.

Sharing Data from Molecular Simulations

Ab-Initio Solution of the Many-Electron Schrödinger Equation with Deep Neural Networks

Boltzmann generators: Sampling equilibrium states of many-body systems with deep learning

Machine Learning Interatomic Potentials as Emerging Tools for Materials Science

FCHL revisited: Faster and more accurate quantum machine learning.

Unfolding adsorption on metal nanoparticles: Connecting stability with catalysis

Deep learning enables rapid identification of potent DDR1 kinase inhibitors

Machine Learning the Physical Nonlocal Exchange-Correlation Functional of Density-Functional Theory.

Multireference Electron Correlation Methods: Journeys along Potential Energy Surfaces.

Promoting transparency and reproducibility in enhanced molecular simulations

Theory and practice of modeling van der Waals interactions in electronic-structure calculations.

Embedded Atom Neural Network Potentials: Efficient and Accurate Machine Learning with a Physically Inspired Representation.

Efficient Global Structure Optimization with a Machine-Learned Surrogate Model.

Nudged elastic band calculations accelerated with Gaussian process regression based on inverse inter-atomic distances.

DFT+U in Dudarev's formulation with corrected interactions between the electrons with opposite spins: The form of Hamiltonian, calculation of forces, and bandgap adjustments.

Advances in Density-Functional Calculations for Materials Modeling

Unsupervised word embeddings capture latent knowledge from materials science literature

ChemML: A machine learning and informatics program package for the analysis, mining, and modeling of chemical and materials data

100

Unifying machine learning and quantum chemistry with a deep neural network for molecular wavefunctions

101

A Performance and Cost Assessment of Machine Learning Interatomic Potentials.

102

Machine Learning Guided Approach for Studying Solvation Environments.

103

Developing an improved crystal graph convolutional neural network framework for accelerated materials discovery

104

Evidence for supercritical behaviour of high-pressure liquid hydrogen

105

The digitization of organic synthesis

106

Holistic Prediction of Enantioselectivity in Asymmetric Catalysis

107

On the Inductive Bias of Neural Tangent Kernels

108

Machine Learning Accelerates the Discovery of Design Rules and Exceptions in Stable Metal–Oxo Intermediate Formation

109

De novo exploration and self-guided learning of potential-energy surfaces

110

Comparing Molecular Patterns Using the Example of SMARTS: Applications and Filter Collection Analysis

111

Comparing Molecular Patterns Using the Example of SMARTS: Theory and Algorithms

112

End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography

113

Functional form of the superconducting critical temperature from machine learning

114

A critical comparison of neural network potentials for molecular reaction dynamics with exact permutation symmetry.

115

Neural-Network-Based Path Collective Variables for Enhanced Sampling of Phase Transformations.

116

A novel approach to describe chemical environments in high-dimensional neural network potentials.

117

PubChemQC PM6: Data Sets of 221 Million Molecules with Optimized Molecular Geometries and Electronic Properties

118

Isotope effects in liquid water via deep potential molecular dynamics

119

Optimization of the facet structure of transition-metal catalysts applied to the oxygen reduction reaction

120

New frontiers for the materials genome initiative

121

Neural networks-based variationally enhanced sampling

122

Molecular Geometry Prediction using a Deep Generative Graph Neural Network

123

Describing Molecular Polarizability by a Bond Capacity Model.

124

Data-Mining for Processes in Chemistry, Materials, and Engineering

125

A coordination-based model for transition metal alloy nanoparticles.

126

Completing density functional theory by machine learning hidden messages from molecules

127

Unmasking Clever Hans predictors and assessing what machines really learn

128

Resonant catalysis of thermally activated chemical reactions with vibrational polaritons

129

PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments, and Partial Charges.

130

Unsupervised machine learning in atomistic simulations, between predictions and understanding.

131

Mol-CycleGAN: a generative model for molecular optimization

132

Origins of complex solvent effects on chemical reactivity and computational tools to investigate them: a review

133

Qresp, a tool for curating, discovering and exploring reproducible scientific papers

134

Making Better Decisions During Synthetic Route Design: Leveraging Prediction to Achieve Greenness-by-Design

135

Molecular force fields with gradient-domain machine learning: Construction and application to dynamics of small molecules with coupled cluster forces.

136

Prediction of higher-selectivity catalysts by computer-driven workflow and machine learning

137

A Universal Density Matrix Functional from Molecular Orbital-Based Machine Learning: Transferability across Organic Molecules

138

Effect of Crystallization Modes in TIPS-pentacene/Insulating Polymer Blends on the Gas Sensing Properties of Organic Field-Effect Transistors

139

Atomic cluster expansion for accurate and transferable interatomic potentials

140

Automatic Prediction of Surface Phase Diagrams Using Ab Initio Grand Canonical Monte Carlo

141

Solvation Free Energy Calculations with Quantum Mechanics/Molecular Mechanics and Machine Learning Models.

142

sGDML: Constructing accurate and data efficient molecular force fields using machine learning

143

Machine Learning of Coarse-Grained Molecular Dynamics Force Fields

144

A regression approach for explaining manifold embedding coordinates

145

Ab initio thermodynamics of liquid and solid water

146

Low-Scaling Algorithm for Nudged Elastic Band Calculations Using a Surrogate Machine Learning Model.

147

Comment on “Predicting reaction performance in C–N cross-coupling using machine learning”

148

Response to Comment on “Predicting reaction performance in C–N cross-coupling using machine learning”

149

Graph Theory Approach to High-Throughput Surface Adsorption Structure Generation.

150

Molecular Transformer: A Model for Uncertainty-Calibrated Chemical Reaction Prediction

151

Generating equilibrium molecules with deep neural networks

152

GFN2-xTB-An Accurate and Broadly Parametrized Self-Consistent Tight-Binding Quantum Chemical Method with Multipole Electrostatics and Density-Dependent Dispersion Contributions.

153

Reactive SINDy: Discovering governing reactions from concentration data

154

Scoring of tumor-infiltrating lymphocytes: From visual estimation to machine learning.

155

Learning Multimodal Graph-to-Graph Translation for Molecular Optimization

156

Quantifying Density Errors in DFT.

157

Accurate molecular polarizabilities with coupled cluster theory and machine learning

158

Transferable Machine-Learning Model of the Electron Density

159

Towards operando computational modeling in heterogeneous catalysis

160

Alchemical perturbation density functional theory

161

SchNetPack: A Deep Learning Toolbox For Atomistic Systems.

162

Entangled Conditional Adversarial Autoencoder for de Novo Drug Discovery.

163

Deep learning in biomedicine

164

Review of force fields and intermolecular potentials used in atomistic computational materials research

165

Local Bayesian optimizer for atomic structures

166

Machine Learning Configuration Interaction.

167

A Density Functional Tight Binding Layer for Deep Learning of Chemical Hamiltonians.

168

Boosting Quantum Machine Learning Models with a Multilevel Combination Technique: Pople Diagrams Revisited.

169

Physically informed artificial neural networks for atomistic modeling of materials

170

Machine learning in chemoinformatics and drug discovery.

171

From Machine Learning to Explainable AI

172

Inverse molecular design using machine learning: Generative models for matter engineering

173

Operators in quantum machine learning: Response properties in chemical space.

174

Perspective: Multireference coupled cluster theories of dynamical electron correlation.

175

Machine learning meets volcano plots: computational discovery of cross-coupling catalysts

176

Machine learning enhanced global optimization by clustering local environments to enable bundled atomic energies.

177

Machine learning for predicting product distributions in catalytic regioselective reactions.

178

Atom-density representations for machine learning.

179

Machine learning for molecular and materials science

180

Interaction trends between single metal atoms and oxide supports identified with density functional theory and statistical learning

181

Machine learning for heterogeneous catalyst design and discovery

182

A Data-Driven Construction of the Periodic Table of the Elements

183

Can exact conditions improve machine-learned density functionals?

184

Extracting Knowledge from Data through Catalysis Informatics

185

Molecular generative model based on conditional variational autoencoder for de novo molecular design

186

Machine learning-based screening of complex molecules for polymer solar cells.

187

The rise of deep learning in drug discovery.

188

Bridging homogeneous and heterogeneous catalysis by heterogeneous single-metal-site catalysts

189

Neural Message Passing with Edge Updates for Predicting Properties of Molecules and Materials

190

MolGAN: An implicit generative model for small molecular graphs

191

Many-Body Descriptors for Predicting Molecular Properties with Machine Learning: Analysis of Pairwise and Three-Body Interactions in Molecules.

192

Machine Learning in Computer-Aided Synthesis Planning.

193

Predicting reaction performance in C–N cross-coupling using machine learning

194

Quantum Machine Learning in Chemical Compound Space.

195

Automatic selection of atomic fingerprints and reference configurations for machine-learning potentials.

196

Machine learning in catalysis

197

Orbital-free density functional theory for materials research

198