Our research is aimed at development of computational tools for the integrated analysis of newly available genomic-wide databases with existing structural databases. The rationale for this effort is based on the premise that detailed examinations of structural and functional biological databases will reveal new classes of molecular targets and potentially novel small ligands that specifically recognize these targets. Our suite of computational tools is divided into a) those aimed at developing a better undertanding of the processes involved in molecular recognition, and b) those aimed at the analysis of large amounts of drug screening data, microarray data and biochemical pathway data, relevant for novel drug discovery. The integration of these two sets of computational tools will facilitate hypothesis generation about chemically important compounds and their putative targets and/or mechanism of action. To date, progress on the development of computational tools for molecular recognition include a) the detailed characterization of important atoms involved in strong and specific ligand binding, b) the development of detailed quantum chemical descriptors that characterize the electronic structure of reactive partners, with emphasis on classes of proteins that contain zinc-finger domains, c) the development of computational tools that permit explorations of the roles of ligand and target flexibility in molecular interactions, and d) the construction of a novel class of shape descriptors that permit the rapid determination of shape-compatible ligands and putative target binding sites.Our future efforts continue to explore the relationships between molecular structure and biological function. These efforts are focused on new drug discovery as it relates to agents effective against cancer and related opportunistic infections. Analyses are currently in progress to examine all of the publically available data collected in the Developmental Therapeutics Program's drug screening efforts. These data include chemical screens as well as microarray datasets. A web-based utility is now available for detailed examinations of a) screening and microarray datasets, b) chemical classifications of tested compounds, c) relationships between molecular structures that exhibit similar activities in selected screens and d) molecular targets and their role in biochemical pathways. Efforts to integrate and mine these new datasets will improve the pace of drug discovery and provide a foundation for applications of basic science towards rational drug discovery. Z01-BC-10281-02
