Diffusion Methods for Generating Transition Paths

TL;DR

This paper introduces two novel score-based generative methods for efficiently simulating rare transition paths between metastable states, aiding molecular system analysis especially when data is limited.

Contribution

The paper presents two new path generation techniques, a chain-based and a midpoint-based approach, for improved simulation of rare transitions in molecular systems.

Findings

Effective in data-rich regimes

Works well in data-scarce regimes

Demonstrated on Muller potential and Alanine dipeptide

Abstract

In this work, we seek to simulate rare transitions between metastable states using score-based generative models. An efficient method for generating high-quality transition paths is valuable for the study of molecular systems since data is often difficult to obtain. We develop two novel methods for path generation in this paper: a chain-based approach and a midpoint-based approach. The first biases the original dynamics to facilitate transitions, while the second mirrors splitting techniques and breaks down the original transition into smaller transitions. Numerical results of generated transition paths for the M\"uller potential and for Alanine dipeptide demonstrate the effectiveness of these approaches in both the data-rich and data-scarce regimes.