Inverse problems with experiment-guided AlphaFold

TL;DR

This paper introduces a framework that integrates experimental data with AlphaFold to generate diverse conformational ensembles of proteins, capturing their dynamic nature more accurately than existing methods.

Contribution

The authors develop a novel approach that treats AlphaFold as a prior and uses experimental data for posterior inference, enabling rapid and accurate modeling of protein conformational heterogeneity.

Findings

Uncovers previously unmodeled conformational heterogeneity from crystallographic data.

Generates high-accuracy NMR ensembles significantly faster than existing methods.

Ensembles outperform AlphaFold3 and sometimes better fit experimental data than PDB structures.

Abstract

Proteins exist as a dynamic ensemble of multiple conformations, and these motions are often crucial for their functions. However, current structure prediction methods predominantly yield a single conformation, overlooking the conformational heterogeneity revealed by diverse experimental modalities. Here, we present a framework for building experiment-grounded protein structure generative models that infer conformational ensembles consistent with measured experimental data. The key idea is to treat state-of-the-art protein structure predictors (e.g., AlphaFold3) as sequence-conditioned structural priors, and cast ensemble modeling as posterior inference of protein structures given experimental measurements. Through extensive real-data experiments, we demonstrate the generality of our method to incorporate a variety of experimental measurements. In particular, our framework uncovers previously unmodeled conformational heterogeneity from crystallographic densities, and generates high-accuracy NMR ensembles orders of magnitude faster than the status quo. Notably, we demonstrate that our ensembles outperform AlphaFold3 and sometimes better fit experimental data than publicly deposited structures to the Protein Data Bank (PDB). We believe that this approach will unlock building predictive models that fully embrace experimentally observed conformational diversity.