Yihang Wang (Case Western)

Date: Wed. October 29th, 2025, 4:30 pm-5:45 pm
Location: Rockefeller 301
Add to Google Calendar

Biophysics Colloquium

AI for Biomolecular Discovery: Overcoming Sampling and Modeling Challenges in Molecular Simulations

Molecular dynamics (MD) simulations are emerging as essential tools for understanding biomolecular systems. However, challenges such as the rare event problem and the development of accurate models remain. I will showcase, by combining ideas from statistical mechanics, how AI methods can be carefully adapted to tackle these challenges. First, I will present AI methods for learning optimal reaction coordinates (RCs), drawing on the information bottleneck framework, and demonstrate how these learned RCs accelerate rare-event sampling in MD. Second, I will introduce how diffusion models, trained to generate molecular ensembles, improve molecular configuration sampling. Furthermore, I will show that by incorporating external physical or chemical data, these generative models can predict unseen or novel molecular structures, thus expanding the predictive capacity of MD beyond traditional simulation limits. These works highlight how AI, when combined with principles of statistical mechanics and thermodynamics, provides a powerful new framework for studying complex molecular systems.