| Structure-based design offers insight that cannot be made with new methods that are in use for drug discovery and development, such as genomics, high-throughput screening and combinatorial chemistry. By advancing structure-based methods, it may be possible to make predictions about the structural aspects of macromolecular interactions from sequence information. These predictions may be used to identify new targets and to optimize existing targets.; This dissertation is a first step toward the long-term goal of using macromolecular docking to predict interactions and function. Chapter 2 details a method that describes the shape of complex macromolecular surfaces for docking. In chapter 3, these descriptors are used to explore the interactions of known systems, and predict a geometry of human growth hormone receptor that is not seen in x-ray crystal structures. Chapter 4 describes the discovery of small molecule inhibitors of HIV-1 integrase by using molecular docking methods to target a site at the dimer interface, more than 15A from the active site with no known function. Chapter 5 is a preliminary study of the use of DOCK to screen a database of small molecules with the eventual goal of building novel, larger molecules from the small molecules. Our model is FK506 binding protein, and the inherent difficulties of working with this type of site are revealed in the work. |
