Enhanced Sampling and Machine Learning of Biomolecular Dynamics

Molecular dynamics (MD) is a powerful computational technique for simulating biomolecular dynamics at an atomistic level. Conventional MD (cMD) is often limited to microsecond simulation timescale, thus cannot sufficiently sample many biological processes of interest. Enhanced sampling techniques have been developed to overcome this challenge. Gaussian accelerated MD (GaMD) is an unconstrained enhanced sampling technique that adds a harmonic boost potential to smooth the potential energy surface and reduce system energy barriers. It allows for simultaneous enhanced sampling and free energy calculations. GaMD simulations have successfully captured complex biological processes, including ligand binding, protein folding, protein conformational changes, and protein-protein/nucleic acid interactions. Given its robustness, GaMD and its variant methods are suitable to explore the structural and functional mechanisms of important biomolecules (such as G-protein-coupled receptors (GPCRs), γ-secretase, and RNAs). In chapter 1, I apply GaMD to uncover two different binding pathways and dissociation mechanism of caffeine in the human A2A adenosine receptor (A2AAR). In chapter 2, I perform sequence coevolutionary and structural analysis to uncover important residue contacts for class A and B GPCR activation and allosteric modulation. In chapters 3 and 4, I develop GLOW – a workflow integrating GaMD and Deep Learning (DL) for free energy profiling and apply GLOW to uncover the dynamic mechanism of class A and B GPCR allosteric modulation. In chapter 5, I apply GaMD to uncover the effects of six representative presenilin-1 (PS1) familial Alzheimer’s disease (FAD) mutations in the amyloid precursor protein (APP)-bound γ-secretase complex. In chapter 6, I apply the GaMD in OpenMM to explore the folding mechanisms and ligand binding to RNAs. In chapter 7, I develop a new Deep Boosted Molecular Dynamics (DBMD) method to capture multiple folding-unfolding events of fast-folding protein and RNA structures.