A transferable active-learning strategy for reactive molecular force fields. Issue 32 (21st July 2021)

• Record Type: Journal Article
• Authors: Young, Tom A.; Johnston-Wood, Tristan; Deringer, Volker L.; Duarte, Fernanda
• Is Part Of: Chemical science. Volume 12:Issue 32(2021)
• Page Start: 10944
• Page End: 10955
• Publication Date: 2021-07-21
• Subjects: Chemistry -- Periodicals; 540.5
• DOI: 10.1039/d1sc01825f
• Languages: English
• ISSNs: 2041-6520

Abstract

Abstract : An efficient strategy for training Gaussian Approximation Potential (GAP) models to study chemical reactions using hierarchical and active learning. Abstract : Predictive molecular simulations require fast, accurate and reactive interatomic potentials. Machine learning offers a promising approach to construct such potentials by fitting energies and forces to high-level quantum-mechanical data, but doing so typically requires considerable human intervention and data volume. Here we show that, by leveraging hierarchical and active learning, accurate Gaussian Approximation Potential (GAP) models can be developed for diverse chemical systems in an autonomous manner, requiring only hundreds to a few thousand energy and gradient evaluations on a reference potential-energy surface. The approach uses separate intra- and inter-molecular fits and employs a prospective error metric to assess the accuracy of the potentials. We demonstrate applications to a range of molecular systems with relevance to computational organic chemistry: ranging from bulk solvents, a solvated metal ion and a metallocage onwards to chemical reactivity, including a bifurcating Diels–Alder reaction in the gas phase and non-equilibrium dynamics (a model SN 2 reaction) in explicit solvent. The method provides a route to routinely generating machine-learned force fields for reactive molecular systems.