We will carry out calculations in four main areas. First, we describe a pattern-matching approach to prediction of protein secondary structure. This program can be coupled to the tertiary packing algorithms of Cohen et.al. to produce protein-like structures. In the best case with proteins that have been studied crystallographically, discrepancies of ca. 4 angstroms are observed. Methods of reducing these errors are evaluated. Second, we seek a detailed model for the motions of ligands in proteins. Conventional molecular dynamics and trajectory calculations will be compared to methods that reveal packing defects and channels in proteins in a direct way. Third, we wish to extend our docking calculations to macromolecular interactions and to systematic structural comparison of sets of bioactive compounds. Fourth, methods are discussed to evaluate the potential functions used in molecular mechanics and molecular dynamics.
Showing the most recent 10 out of 56 publications