EigenFold: Generative Protein Structure Prediction with Diffusion Models

Protein structure prediction has reached revolutionary levels of accuracy on single structures, yet distributional modeling paradigms are needed to capture the conformational ensembles and flexibility that underlie biological function. We define a diffusion process that models the structure as a system of harmonic oscillators and which naturally induces a cascading-resolution generative process along the eigenmodes of the system. 's ability to model and predict conformational heterogeneity for fold-switching proteins and ligand-induced conformational change. The development of accurate methods for protein structure prediction such as AlphaFold2 (Jumper et al., 2021) has revolutionized in silico understanding of protein structure and function. However, while such methods are designed to model static experimental structures from crystallography or cryo-EM, proteins in vivo adopt dynamic structural ensembles featuring conformational flexibility, change, and even disorder to effect their biological functions (Teague, 2003; Wright & Dyson, 2015).