Soumendranath Bhakat

Benchmarking generative AI and physics based molecular simulation for sampling conformational heterogeneity in T4 Lysozyme

The paper benchmarks generative AI and AI-accelerated molecular simulation for sampling conformational states of T4 lysozyme, finding that AMS recovers all states and matches experimental data better than generative AI methods.

Accelerated sampling of protein dynamics using BioEmu augmented molecular simulation

This study introduces BioEmu, a workflow combining generative AI with molecular simulations to sample protein conformations, showing both broad coverage and specific limitations in capturing conformational heterogeneity.

CryoPhold: CryoEM meets AlphaFold and molecular simulation to reveal protein dynamics

CryoPhold integrates AlphaFold, cryo-EM, and molecular simulations to predict protein dynamics and structural hotspots in human proteins like GlyT1 and BRAF mutants.

Generalizable Protein Dynamics in Kinases: Physics is the key

This paper introduces a computational framework to study kinase conformational dynamics by integrating AlphaFold, machine learning, and physics-based simulations to explore metastable states beyond traditional binary classifications.