Gaussian Process Regression for Materials and Molecules

2021/08/16 by Volker L. Deringer, Albert P. Bartók, Noam Bernstein +3 · 51 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · Materials Science · #Computational Drug Discovery Methods #Machine Learning in Materials Science #Protein Structure and Dynamics

paper · pdf · doi:10.1021/acs.chemrev.1c00022

openalex publication_date 2021/08/16 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/29

Abstract

We provide an introduction to Gaussian process regression (GPR) machine-learning methods in computational materials science and chemistry. The focus of the present review is on the regression of atomistic properties: in particular, on the construction of interatomic potentials, or force fields, in the Gaussian Approximation Potential (GAP) framework; beyond this, we also discuss the fitting of arbitrary scalar, vectorial, and tensorial quantities. Methodological aspects of reference data generation, representation, and regression, as well as the question of how a data-driven model may be validated, are reviewed and critically discussed. A survey of applications to a variety of research questions in chemistry and materials science illustrates the rapid growth in the field. A vision is outlined for the development of the methodology in the years to come.

Citations

• Gaussian approximation potentials: A brief tutorial introduction
• Nonphysical sampling distributions in Monte Carlo free-energy estimation: Umbrella sampling
• Efficient iterative schemes forab initiototal-energy calculations using a plane-wave basis set
• Taking the Human Out of the Loop: A Review of Bayesian Optimization
• The farthest point strategy for progressive image sampling
• SchNet – A deep learning architecture for molecules and materials
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• Operators in quantum machine learning: Response properties in chemical space
• Machine learning and the physical sciences
• DScribe: Library of descriptors for machine learning in materials science
• Machine Learning for Molecular Simulation
• A Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Predicting molecular dipole moments by combining atomic partial charges and atomic dipoles
• Four Generations of High-Dimensional Neural Network Potentials
• Multi-body effects in a coarse-grained protein force field

Cited by

• Electrostatic embedding machine learning for ground and excited state molecular dynamics of solvated molecules
• The Analysis of Electron Densities: From Basics to Emergent Applications
• Uncertainty Quantification for In Silico Chemistry
• Synergizing chemical and AI communities for advancing laboratories of the future
• Cross-disciplinary perspectives on the potential for artificial intelligence across chemistry
• Challenges and opportunities for machine learning potentials in transition path sampling: alanine dipeptide and azobenzene studies
• Efficient small-cell sampling for machine-learning potentials of multi-principal element alloys
• Reciprocal Space Attention for Learning Long-Range Interactions
• Active learning of alchemical adsorption simulations; towards a universal adsorption model
• An Accurate and Efficient Machine-Learned Potential for SiC from Ambient to Extreme Environments
• SiC-TGAP: A machine learning interatomic potential for radiation damage simulations in 3C-SiC
• Neural Network Surrogates for Free Energy Computation of Complex Chemical Systems
• Adaptive Pruning for Increased Robustness and Reduced Computational Overhead in Gaussian Process Accelerated Saddle Point Searches
• Investigating Anharmonicities in Polarization-Orientation Raman Spectra of Acene Crystals with Machine Learning
• Improved capabilities of the TurboGAP code for radiation induced cascade simulations: an illustration with silicon
• Active and transfer learning with partially Bayesian neural networks for materials and chemicals
• AI‐Guided Inverse Design and Discovery of Recyclable Vitrimeric Polymers
• Nested sampling for physical scientists
• Application of pretrained universal machine-learning interatomic potential for physicochemical simulation of liquid electrolytes in Li-ion batteries
• Active learning driven prioritisation of compounds from on-demand libraries targeting the SARS-CoV-2 main protease
• Tuning of atomic layer deposition pulse time through physics-informed Bayesian active learning
• A foundation model for atomistic materials chemistry
• Vibrational Fingerprints of Strained Polymers: A Spectroscopic Pathway to Mechanical State Prediction
• Density-Aware Farthest Point Sampling
• Unsupervised Atomic Data Mining via Multi-Kernel Graph Autoencoders for Machine Learning Force Fields
• Medium-range structural order in amorphous arsenic
• Node-Equivariant Message Passing for Efficient and Accurate Machine Learning Interatomic Potentials
• Metatensor and metatomic: foundational libraries for interoperable atomistic machine learning
• Data for Physics-informed Hamiltonian learning for large-scale optoelectronic property prediction
• MACE4IRmol: An uncertainty-aware foundation model for molecular infrared spectroscopy
• Hydrogen Diffusion in Garnet: Insights From Atomistic Simulations
• Physics-Informed Gaussian Process Inference of Liquid Structure from Scattering Data
• Phase Stability and Transformations in Lead Mixed Halide Perovskites from Machine Learning Force Fields
• Machine Learning Accelerates Raman Computations from Molecular Dynamics for Materials Science
• An Atomic Cluster Expansion Potential for Twisted Multilayer Graphene
• Leveraging active learning-enhanced machine-learned interatomic potential for efficient infrared spectra prediction
• Machine learning model for efficient nonthermal tuning of the charge density wave in monolayer NbSe₂
• Enhancing Experimental Efficiency in Materials Design: A Comparative Study of Taguchi and Machine Learning Methods
• High-throughput viscometry via machine-learning from videos of inverted vials
• Machine Learning the Energetics of Electrified Solid/Liquid Interfaces
• Recovering Hidden Degrees of Freedom Using Gaussian Processes
• Electrostatics as a Guiding Principle in Understanding and Designing Enzymes
• Finite-Sample-Based Reachability for Safe Control with Gaussian Process Dynamics
• The evolution of machine learning potentials for molecules, reactions and materials
• Pushing the limits of unconstrained machine-learned interatomic potentials
• A Padding Method for Enhanced Encoding of Inorganic Structures with Varying Chemical Compositions
• Scalable Quantum Molecular Generation via GPU-Accelerated Tensor-Network Simulation
• Exploring Charge Density Waves in two-dimensional NbSe2 with Machine Learning
• Interface Engineering of Helium Confinement in Argon-Preplated MCM-41 Nanopores
• Agentic self-driving microscopy benchmarks support qualification but do not necessarily generalize to unseen tasks
• Reconstructions and Dynamics of β-Lithium Thiophosphate Surfaces

Related

• Efficient Implementation of Gaussian Process Regression Accelerated Saddle Point Searches with Application to Molecular Reactions
• Learning inducing points and uncertainty on molecular data by scalable variational Gaussian processes
• Atomic-scale representation and statistical learning of tensorial properties
• Tensor Basis Gaussian Process Models of Hyperelastic Materials
• Gaussian Approximation Potentials: theory, software implementation and application examples