BPS2026 – Folding genomes at nucleosome resolution: Multiscale modeling and merging physicochemical potentials with machine learning approaches

Molecular modeling and simulation is now recognized as a field on its own right capable of making key mechanistic insights into biomolecular processes. Notable successes include prediction of conformational transitions and mechanisms; generation of new insights into biomolecular activity; and thriving collaborations between modeling and experimentation, including experiments driven by modeling. As examples, I will describe mesoscale and multiscale modeling applications of genome folding, showing how we develops models and tailored algorithms like equilibrium Monte Carlo sampling, DiSCO (discrete surface charge optimization), and Brownian dynamics to fold genes, recover chromosome conformations from Hi-C maps, and solve frontier problems in the epigenetic regulation of the genome and its relation to human disease.