Predicting the Thermodynamic Properties of Proteins Using Computer Simulations

Protein molecules, sometimes referred to as the molecules of life, are the drivers of virtually every biological function. In this dissertation, we describe a series of computational studies to dissect the mystery of complex protein molecules. We consider a large collection of protein systems, ranging from globular proteins to Intrinsically Disordered Proteins (IDPs) with a focus on predicting thermodynamic observables that can be quantitatively compared with experimental data. In the first part of this dissertation, we study the effects of the phenomenon of macromolecular crowding and how it affects the properties of two different groups of proteins. First, we investigate the effects of crowding on globular proteins by calculating the free energy of all-atom proteins in crowded environments. Second, we study how crowding affect the conformational ensembles of disordered proteins with a focus on comparing computations with experiments. In the second part of this dissertation, we apply Monte Carlo simulation techniques to study protein droplet formation and Liquid Liquid Phase Separation in protein systems.