From data to noise to data: mixing physics across temperatures with generative artificial intelligence

TL;DR

This paper introduces a novel framework combining statistical mechanics and generative AI, specifically denoising diffusion models, to predict biomolecular behavior across unsimulated temperatures by treating temperature as a fluctuating variable.

Contribution

It develops a new method that enables sampling of molecular configurations at arbitrary temperatures without prior simulations, improving understanding of biomolecular energy landscapes.

Findings

Successfully applied to peptide and RNA systems

Discovered unseen transition and metastable states

Eliminated need for additional temperature-specific simulations

Abstract

Using simulations or experiments performed at some set of temperatures to learn about the physics or chemistry at some other arbitrary temperature is a problem of immense practical and theoretical relevance. Here we develop a framework based on statistical mechanics and generative Artificial Intelligence that allows solving this problem. Specifically, we work with denoising diffusion probabilistic models, and show how these models in combination with replica exchange molecular dynamics achieve superior sampling of the biomolecular energy landscape at temperatures that were never even simulated without assuming any particular slow degrees of freedom. The key idea is to treat the temperature as a fluctuating random variable and not a control parameter as is usually done. This allows us to directly sample from the joint probability distribution in configuration and temperature space. The results here are demonstrated for a chirally symmetric peptide and single-strand ribonucleic acid undergoing conformational transitions in all-atom water. We demonstrate how we can discover transition states and metastable states that were previously unseen at the temperature of interest, and even bypass the need to perform further simulations for wide range of temperatures. At the same time, any unphysical states are easily identifiable through very low Boltzmann weights. The procedure while shown here for a class of molecular simulations should be more generally applicable to mixing information across simulations and experiments with varying control parameters.