DESIGN AND TESTING OF DOCKING ALGORITHMS

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. Our long term goal is to address the "docking problem," that of predicting the structures and ligand identities of protein-ligand complexes. Methods to do so would have wide application in structure-function studies. The specific aims are: 1. To develop algorithms that dock ligands in hierarchies of increasing geometrical and chemical complexity. Although the configuration space for ligand-protein complexes is enormous, it is also highly redundant and constrained by the excluded volume of the complex. Docking molecules as ensembles of states allow for methods that take advantage of these features. By representing molecules in hierarchies of increasing complexity, it may be possible to exclude most unfavorable conformations early in the calculation. Such a hierarchical algorithm should be applicable to related chemistries in the same way as to related conformations. These methods would explore many more states and chemistries than can now be considered in docking calculations. Improvements to solvation energy models will also be investigated. 2. To test the new algorithms in a well-behaved experimental system. The new algorithms will be tested for their ability to predict new ligands and geometries for two model systems: AmpC beta-lactamase and a cavity site in lysozyme. Both proteins are easy to work with and provide well defined sites for docking studies. At the simplest level, the experimental tests will evaluate "hit-rates" for the algorithms and their accuracy through structure determination. More fundamentally, this will alllow for investigation of the thermodynamic driving forces in complex formation. Extensive prelimary results suggest that the new algorithms are feasible. See the following articles for instance: a. Flexible Ligand Docking Using Conformational Ensembles b. Docking Molecules by Families to Increase the Diversity of Hits in Database Screens c. Protein-Protein Docking with Multiple Ligand Residue Conformations and Multiple Residue Identities The model system seems to be well suited to testing the new algorithms. See the following articles for instance: a. Mapping the Active Site of AmpC beta-Lactamase for Hot Spots b. Structure-based Discovery of a Novel, Non-Covalent Inhibitor of AmpC beta-Lactamase c. A Model Binding Site for Testing Scoring Functions in Molecular Docking Interactive computer graphics available through the RBVI are key to visualizing and evaluating how well docking calculations have performed, and to the structure-based discovery of new ligands. They are also very useful for crystallographic modeling and structure determination, which is part of the experimental testing aspect of this project.