Michael Feig

aSAM, a latent diffusion model trained on molecular dynamics, generates heavy-atom protein ensembles; its temperature-conditioned variant aSAMt reproduces mdCATH thermal unfolding and extrapolates beyond training temperatures.

This paper introduces GSnet, a graph neural network for predicting protein molecular properties, and shows it outperforms existing methods for solvent-accessible surface area prediction.

The paper studies non-native cargo loading into HO shell variants and finds that activity depends on cargo amount, not shell permeability, with simpler variants showing robust assembly.

The paper uses SAXS and PDDF to analyze BMC shell structures and enzyme encapsulation in Haliangium ochraceum BMCs expressed in E. coli, showing cytoplasmic protein capture during assembly.

COCOMO2 is a coarse-grained model that extends to folded proteins, enabling realistic simulations of biomolecular interactions and phase separation.

This study uses molecular dynamics simulations to investigate viscosity and diffusion in crowded protein solutions, showing that diffusion reduction exceeds viscosity effects due to cluster formation.

This paper introduces idpSAM, a deep generative model for transferable modeling of intrinsically disordered protein conformations, showing high performance on unseen sequences.

The paper uses molecular dynamics simulations to investigate how synthetic crowders affect the function of the hepatitis C virus NS3/4A protease and its interactions with substrates.