Generative methods for sampling transition paths in molecular dynamics
This work addresses a computational bottleneck in molecular dynamics simulations, but it appears incremental as it explores existing machine learning techniques without claiming major breakthroughs.
The authors tackled the problem of efficiently simulating transition paths between metastable states in molecular dynamics, which is difficult with direct numerical methods, by exploring generative models and reinforcement learning approaches.
Molecular systems often remain trapped for long times around some local minimum of the potential energy function, before switching to another one -- a behavior known as metastability. Simulating transition paths linking one metastable state to another one is difficult by direct numerical methods. In view of the promises of machine learning techniques, we explore in this work two approaches to more efficiently generate transition paths: sampling methods based on generative models such as variational autoencoders, and importance sampling methods based on reinforcement learning.