CHE 598 Seminar: Evolution and Pathways of Kinase Complex Formation

About the event

SPEAKER: Dr. Anton Persikov, Senior Data Scientist, Center for Computational Biology, Flatiron Institute Simons Foundation

BIOGRAPHY:

Anton V. Persikov, PhD, is a Senior Data Scientist in the Center for Computational Biology at the Flatiron Institute, Simons Foundation, in New York. He earned his undergraduate and master’s degrees in physics from Lomonosov Moscow State University and his PhD in biophysics and human physiology from the Institute of Developmental Biology, Russian Academy of Sciences. Before joining the Flatiron Institute in 2022, he held research appointments at Princeton University and Robert Wood Johnson Medical School.

His research combines experimental biophysics, molecular modeling, and machine learning to understand how protein sequences determine structure, stability, and function. A major focus is collagen, including how its amino acid sequence and disease-associated mutations affect the stability of its triple helix. He developed the Collagen Stability Calculator, a tool for predicting thermal stability directly from amino acid sequence. His broader work includes predicting how proteins recognize DNA and developing computational methods to detect engineered genomes. Dr. Persikov has authored more than 40 journal publications and three book chapters, contributed to patented technologies for protein design, and mentored undergraduate and graduate researchers.

ABSTRACT:

Protein-protein interactions in signaling networks must be specific enough to transmit information and transient enough to be regulated. The MEK-ERK module, a core kinase signaling cascade, illustrates this tension: ERK (signal-regulated kinase)  is a highly connected and strongly conserved hub, while its only activating kinase engages it through a short, intrinsically disordered docking motif. Phylogenetic analysis across metazoan species, structure prediction, and long-timescale molecular dynamics show that this motif diverges rapidly between evolutionary lineages while preserving the residue contacts required for ERK binding. Functional assays in Drosophila establish that these contacts are essential for downstream signaling and indicate a regulatory role beyond simple tethering. How the complex assembles in the first place remains inaccessible to conventional simulation, since association is a rare event on molecular dynamics timescales. Weighted ensemble simulations sample the association process directly and yield an ensemble of binding pathways rather than a single trajectory, making it possible to ask which interface engages first, which intermediate states the complex visits along the way, and how reproducibly these routes are recovered across independent runs. The resulting path ensembles inform both the mechanism of kinase complex formation and the practical limits of applying enhanced sampling to large protein assemblies.