Computational and theoretical studies of globular proteins
Protein crystallization is often achieved in experiment through a trial and error approach. To date, there exists a dearth of theoretical understanding of the initial conditions necessary to promote crystallization. While a better understanding of crystallization will help to create good crystals suitable for structure analysis, it will also allow us to prevent the onset of certain diseases. The core of this thesis is to model and, ultimately, understand the phase behavior of protein particles in solution. Toward this goal, we calculate the fluid-fluid coexistence curve in the vicinity of the metastable critical point of the modified Lennard-Jones potential, where it has been shown that nucleation is increased by many orders of magnitude. We use finite-size scaling techniques and grand canonical Monte Carlo simulation methods. This has allowed us to pinpoint the critical point and subcritical region with high accuracy in spite of the critical fluctuations that hinder sampling using other Monte Carlo techniques. We also attempt to model the phase behavior of the γ-crystallins, mutations of which have been linked to genetic cataracts. The complete phase behavior of the square well potential at the ranges of attraction λ = 1.15 and λ = 1.25 is calculated and compared with that of the γII-crystallin. The role of solvent is also important in the crystallization process and affects the phase behavior of proteins in solution. We study a model that accounts for the contribution of the solvent free-energy to the free-energy of globular proteins. This model allows us to model phase behavior that includes solvent.