| The binding of proteins to other proteins is an important event in many biochemical processes. Therefore, the mechanisms by which protein-protein recognition occurs have received considerable attention from computational biologists, both in the analysis of known complexes and the prediction and design of their structures. Protein design uses current knowledge of structural biology to predict, by computer simulation, an amino acid sequence that would produce a protein with targetable properties. Conversely, protein structure prediction aims to identify the lowest energy conformation of a protein given a defined sequence, or in the case of the protein docking, the lowest energy conformation for two structures with defined amino acid sequences. This dissertation reports the utilization of computational protein design and protein dock methods to further understand recognition in protein-protein interactions and assess the potential of proteins to bind synthetic ligands.;Chapter 2 describes a study to identify and analyze helix-mediated protein-protein interactions from experimentally determined structures in the Protein Databank and identify potential targets for small molecules and helix mimetics in this dataset using computational protein design methods. This study predicts 159 helix-mediated protein-protein interactions as small molecule targets and 252 helix-mediated protein-protein interactions as targets for helix mimetic compounds.;Chapter 3 is a study of protein ligand recognition and specifically assesses the potential of triazolamer beta-strand mimetic compounds to inhibit HIV protease using protein dock methods. The docking studies predict that the triazolamer side chains bind to hydrophobic pockets in the HIV active site and provide insight into the differential activities of triazolamer compounds. |
