Generative deep learning for intrinsically disordered protein regions
Washington University School of Medicine
Washington University School of Medicine
Abstract:
Intrinsically disordered proteins and regions (IDPs/IDRs) lack stable structure yet play central roles in regulation, signaling, and molecular recognition. Despite their ubiquity and importance in eukaryotic cell biology, they often show weak sequence conservation, limiting the utility of conventional computational approaches that rely on homology or well-defined structural templates. These features also make IDRs difficult to systematically characterize and engineer. This seminar presents computational and deep learning methods for understanding how sequence encodes both conformational behavior and function in these proteins. My work combines sequence-to-ensemble modeling, disorder-specific deep learning, and software tool development to enable large-scale analysis and context-aware design of intrinsically disordered proteins and regions.
Biography -- Borna Novak:
Borna is an M.D./Ph.D. candidate at Washington University in St. Louis, where he completed his Ph.D. in Computational and Systems Biology in the lab of Alex Holehouse. His doctoral research focused on developing and applying machine learning tools to characterize the structural ensembles of intrinsically disordered proteins. He is currently completing the clinical phase of his medical training.
Biography -- Jeffrey Lotthammer:
Jeff completed his Ph.D. in Computational Systems Biology in the lab of Alex Holehouse at Washington University in St. Louis. His doctoral work focused on developing and applying computational tools to characterize and design disordered protein sequences. During his Ph.D., he was generously supported by the National Science Foundation Graduate Research Fellowship and the Frontera Computational Science Fellowship.