Committors without Descriptors

TL;DR

This paper introduces a novel, automated committor-based sampling method that leverages graph neural networks to directly process atomic coordinates, improving the study of rare events in atomistic simulations.

Contribution

It combines a variational, neural-network-based committor approach with graph neural networks to enhance sampling of rare events without relying on predefined descriptors.

Findings

Effective in benchmark systems

Improves sampling of transition states

Captures solvent effects explicitly

Abstract

The study of rare events is one of the major challenges in atomistic simulations, and several enhanced sampling methods towards its solution have been proposed. Recently, it has been suggested that the use of the committor, which provides a precise formal description of rare events, could be of use in this context. We have recently followed up on this suggestion and proposed a committor-based method that promotes frequent transitions between the metastable states of the system and allows extensive sampling of the process transition state ensemble. One of the strengths of our approach is being self-consistent and semi-automatic, exploiting a variational criterion to iteratively optimize a neural-network-based parametrization of the committor, which uses a set of physical descriptors as input. Here, we further automate this procedure by combining our previous method with the expressive power of graph neural networks, which can directly process atomic coordinates rather than descriptors. Besides applications on benchmark systems, we highlight the advantages of a graph-based approach in describing the role of solvent molecules in systems, such as ion pair dissociation or ligand binding.