Using Molecular Simulations and Statistical Models to Understand Biomolecular Conformational Dynamics

Abstract

Conformational dynamics are important to the function of biological molecules. While many experimental techniques (e.g. X-ray crystallography and NMR spectroscopy) have been developed for providing the structure of functional conformations, it is exceptionally challenging to understand conformational dynamics from experimental characterization. Molecular dynamics (MD) simulations is a powerful tool for probing conformational dynamics. The timescale resolution of MD simulations enables people to investigate intermediate conformations and transition pathways in atomic detail. Recent advancements in computer hardware have increased the timescales accessible to MD simulations. Meanwhile, more accurate and specific force fields have been developed to accurately model a variety biological system of different sizes. My graduate research has been focused on using MD simulations to study the conformational dynamics of proteins. Markov State Model (MSM) based approaches are extensively applied to investigate a variety of folding and/or binding mechanisms in atomic detail. Another focus of my work has been developing a Bayesian inference-based approach called BICePs to reconcile experimental measurements with simulation data to determine conformational ensembles and to validate force fields.