Artificial Intelligence Assists Discovery of Reaction Coordinates and Mechanisms from Molecular Dynamics Simulations

TL;DR

This paper presents an AI-driven framework that guides molecular dynamics simulations and extracts complex reaction mechanisms autonomously, leveraging advanced sampling, neural networks, and deep learning for efficient analysis of large simulation data.

Contribution

It introduces an adaptive, autonomous AI framework combining sampling, inference, and deep learning to discover molecular mechanisms from MD simulations, suitable for high-performance computing.

Findings

Framework effectively guides simulations and extracts mechanisms.

Neural networks are made interpretable for molecular insights.

Applicable to massively parallel computing architectures.

Abstract

Exascale computing holds great opportunities for molecular dynamics (MD) simulations. However, to take full advantage of the new possibilities, we must learn how to focus computational power on the discovery of complex molecular mechanisms, and how to extract them from enormous amounts of data. Both aspects still rely heavily on human experts, which becomes a serious bottleneck when a large number of parallel simulations have to be orchestrated to take full advantage of the available computing power. Here, we use artificial intelligence (AI) both to guide the sampling and to extract the relevant mechanistic information. We combine advanced sampling schemes with statistical inference, artificial neural networks, and deep learning to discover molecular mechanisms from MD simulations. Our framework adaptively and autonomously initializes simulations and learns the sampled mechanism, and is thus suitable for massively parallel computing architectures. We propose practical solutions to make the neural networks interpretable, as illustrated in applications to molecular systems.