Unlocking hidden biomolecular conformational landscapes in diffusion models at inference time

TL;DR

ConforMix is an inference-time algorithm that enhances diffusion models to efficiently sample biomolecular conformations, capturing structural variability and dynamic features without prior knowledge of degrees of freedom.

Contribution

It introduces ConforMix, a novel method that improves conformational sampling in diffusion models at inference time, applicable to static and dynamic biomolecular structure prediction.

Findings

Successfully captures domain motion and cryptic pocket flexibility.

Avoids unphysical states during conformational sampling.

Demonstrates scalability and accuracy on critical proteins.

Abstract

The function of biomolecules such as proteins depends on their ability to interconvert between a wide range of structures or "conformations." Researchers have endeavored for decades to develop computational methods to predict the distribution of conformations, which is far harder to determine experimentally than a static folded structure. We present ConforMix, an inference-time algorithm that enhances sampling of conformational distributions using a combination of classifier guidance, filtering, and free energy estimation. Our approach upgrades diffusion models -- whether trained for static structure prediction or conformational generation -- to enable more efficient discovery of conformational variability without requiring prior knowledge of major degrees of freedom. ConforMix is orthogonal to improvements in model pretraining and would benefit even a hypothetical model that perfectly reproduced the Boltzmann distribution. Remarkably, when applied to a diffusion model trained for static structure prediction, ConforMix captures structural changes including domain motion, cryptic pocket flexibility, and transporter cycling, while avoiding unphysical states. Case studies of biologically critical proteins demonstrate the scalability, accuracy, and utility of this method.