Decoding and engineering free energy landscapes for mechanistic insight and functional protein design - Project summary/abstract. Proteins orchestrate cellular processes as dynamic ensembles of interconverting conformations, characterized by the underlying free energy landscape (FELs). Understanding these FELs is paramount for deciphering biological mechanisms, elucidating disease pathogenesis, and engineering novel therapeutics. However, resolving complete FELs, predicting how they respond to perturbations like mutations or ligand binding, and designing them de novo present formidable challenges, limiting our ability to rationally control protein function. This application seeks to bridge this critical gap by developing an integrated computational and experimental platform for the comprehensive decoding, modulation, and de novo design of protein FELs. I am a postdoctoral researcher in Dr. Anum Glasgow’s laboratory at Columbia University, with a strong background in computational biophysics, protein engineering, and advanced hydrogen-deuterium exchange mass spectrometry (HX/MS) analysis. My development of PIGEON-FEATHER, a state-of-the-art Bayesian framework for deriving site-resolved energetics from HX/MS data, exemplifies my commitment to advancing methods for studying protein ensembles. Building on this foundation, my K99 research will establish a transformative framework for resolving, manipulating, and designing protein FELs, providing fundamental insights and practical tools for protein science, drug discovery, and synthetic biology. Aim 1 will develop PF- MetaD, a novel enhanced sampling approach that incorporates HX/MS-derived protection factors (PFs) into meta dynamics simulations. This will be enabled by two deep learning tools I propose to develop—PFNet and PFBoost—for accurate, residue-level PF determination. Together, these will allow the reconstruction of complete protein FELs. Aim 2 will apply these landscape insights to a critical biomedical challenge by designing state- selective protein binders to modulate the FEL of BRAF kinase, aiming to rationally control its activity in cancer- associated mutants by reshaping its conformational ensemble. Aim 3 will push the boundaries of protein engineering by pursuing the de novo design of a universal, ligand-responsive allosteric protein switch based on the PAS domain scaffold, programming its FEL for custom molecular recognition and regulation. Under the primary mentorship of Dr. Anum Glasgow and Dr. Barry Honig, and with the support of collaborators and the rich research environment at Columbia University and affiliated New York City institutions, I will train in single- molecule FRET, high-throughput screening methodologies, advanced machine learning for integrating multimodal biophysical data, scientific leadership, and grant writing. These skills will enable my long-term goal: an independent multidisciplinary lab at a leading R1 institution, focusing on FEL-guided design of functional and therapeutic proteins. This K99/R00 award is critical for my transition to an independent investigator, transforming our ability to rationally program biomolecular behavior.
