| Despite the abundance of structural data, we still cannot accurately predict the structural and energetic changes resulting from mutations at protein interfaces. The inadequacy of current computational approaches to the analysis and design of protein-protein interactions has hampered the development of novel therapeutic and diagnostic agents. Given the importance of side-chain packing in specifying the stability of protein-protein interfaces and protein cores, it is essential to quantitatively understand (and predict) the side-chain dihedral angle distributions observed in proteins of known structure, for computational protein design to be successful.;Following upon the pioneering work of Ramachandran and colleagues on predicting the backbone dihedral angle (&phis; and psi) distributions with hard sphere plus stereochemical constraints models, we first use an analysis of the tau angle dependence of the distribution of &phis;/psi angle observed in proteins of known structure to show that steric constraints alone (i.e. without invoking any additional energetic contributions, such as hydrogen bonding) are sufficient to explain the backbone dihedral angle distributions.;We then employ a hard-sphere plus stereochemical constraints only model to comprehensively study (predict) the side-chain dihedral angle distributions of all non-charged amino acid (Val, Thr, Ser, Cys, Leu, Ile, Phe, Tyr, Trp, Met, Mse, Nle). Our results emphasize that, in many cases, modeling steric, local and stereochemical constraints alone can quantitatively describe side-chain conformational statistics.;In addition, we illustrate how our simple hard sphere models can be applied to rank pretein stabilities and binding affinities. This thesis work will not only lead to a fundamental understanding of protein-protein interactions, but also to the development of efficient computational methods to rationally design protein interfaces with tunable specificity and affinity, and numerous applications in biomedicine. |
