Introducing AlphaFold-SFA, a novel method combining AlphaFold, slow feature analysis (SFA), and metadynamics to accelerate the sampling of rare events like cryptic pocket openings, protein-ligand bindings, and allostery in proteins. This method overcomes limitations of traditional molecular dynamics by using SFA as a dimensionality reduction tool, thus enabling more efficient exploration of protein dynamics and interactions. SFA trained on AlphaFold seeded molecular dynamics mange to capture rare conformational events which plays a key role in conformational dynamics and biomolecular recognition. SFA augmented metadynamics allowed us to proposed a novel mechanism by which RIPK2 kinase interacts with E3 ligase, thus regulate of innate immune responses, which are crucial in the pathogenesis of inflammatory bowel disease (IBD), including Crohn’s disease. This approach could significantly advance structure-based drug discovery by providing deeper insights into protein functions and interactions.
Paper: Vats, S., Bobrovs, R., Soderhjelm, P., Bhakat, S* (2023) “AlphaFold-SFA: accelerated sampling of cryptic pocket opening, protein-ligand binding and allostery by AlphaFold, slow feature analysis and metadynamics”, bioRxiv
SFA code: Github
Analysis: Github
Data: OSF
Life in Computational Biology 