Title: Automated X-ray Crystallography for Structural Genomics Sponsor: NIH (National Institutes of Health) Amount: $1,440,764 Period: 7/1/01 - 6/30/06 PI: Thomas R. Ioerger (Dept. of Computer Science, Texas A&M) co-PI: James C. Sacchettini (Dept. of Biochem. and Biophysics, Texas A&M) Context: This grant was reviewed as a regular NIH R01 grant. It is a competitive renewel of our earlier 2-year NIH R21 (high-risk/high-impact) grant. However, the new grant is part of an NIH Program Project (P01), headed by Paul Adams at Lawrence Berkeley National Lab. The Program Project combines together 4 related R01 grants with a common goal. Our collaborators include Tom Terwilliger at Los Alamos National Lab, Randy Read at Cambridge University, England, and Axel Brunger at Stanford. Goals: The goals of this project are two-fold. First, we will continue developing and improving the TEXTAL system, a pattern recognition-based system for solving protein structures from X-ray crystallography data by automatically interpreting electron density maps. TEXTAL uses extraction of rotation-invariant features in 3D volumetric datasets to search through a database of previously solved protein structures to find local sets of atomic coordinates that fit the patterns in the density. These are combined to form a global model of the protein, and the predicted structure is subjected to a number of pre- and post-processing steps (using various artificial intelligence techniques to simulate intelligent decision-making by human crystallographers) to produce the final model. Second, we will integrate TEXTAL as the automated model-building component in the PHENIX software system, a comprehensive crystallographic computing environment (based on the Python scripting language), being designed by our collaborators at Lawrence Berkeley Lab. Other modules in PHENIX will be contributed by the other groups in the Program Project to form a complete package with a common interface and file formats to support all the computational steps in the process of crystallographic structure determination.