To achieve this goal, we will develop and implement:
we have developed a novel computational strategy to identify ligand binding profiles of proteins across gene families. The method is based on a novel algorithm to characterize protein structures using geometric potential, a sequence order independent profile-profile alignment (SOIPPA) algorithm and an extreme value distribution (EVD) model for statistical significance estimation of the match. These algorithms have been implemented in SMAP software package. SMAP has been successfully applied to investigate the adverse drug effect, to repurpose safe pharmaceuticals to treat different diseases, to identify multiple targets for polypharmacology drug designs, and to understand molecular mechanisms of phamacogenomics. By taking the mutli-scale representation of the protein and computional methods from biology and chemistry together, a high-throughput off-target pipeline has been built to study protein-ligand interactions across whole proteomes.
In collaboration with Prof. Patsy Babbitt in UCSF, whose group has developed Structure Function Linkage Database (SFLD) and an ontology to classify mechanistically diverse superfamilies of enzymes based on their partial reactions, we are building an ontology including all features involved in molecular interaction consisting of catalytic active residues, small organic molecule binding, metal binding, peptide binding, protein-protein, protein-DNA, and protein-RNA interactions to bridge biological and chemical space. These features will be represented at multiple levels including biological ensembles, molecules, domains, functional motifs, amino acid residues, atoms, and multi-scale physical profiles to facilitate data analysis and modeling at different scales. Ontology development will also bring to bear our experience in developing the ontology for the Protein Data Bank (PDB) and the immune epitope database (IEDB). As part of these collaborations we will also track appropriate developments of other biological ontologies and will incorporate them as appropriate.
Our experience with the PDB has taught us that usability is critical and hence deserving of a separate specific aim. With this in mind we are developing:
1) Visualization tools based on Molecular Biology Toolkit (MBT) to support comparative proteome analysis of multiple functional sites and protein structures simultaneously.
2) Graphical interfaces for querying, computing, analyzing, and visualizing functional site information.
3) Application programming interfaces (APIs) for algorithm development.
4) Web services to facilitate data and software integration.
We welcome collaborations from academia and industry to build a community-wide resources of protein-ligand interactions to facilatate our understanding of molecular basis of cellular functions.
We encourage students and postdoctors to join our team.
Please contact Prof. Phil Bourne (email: bourne@sdsc.edu) or Dr. Lei Xie (email: lxie@sdsc.edu) for more information.
 |
The FS project is supported by the National Institutes of Heath (NIH) grant number GM078596 and is located within the Skaggs School of Pharmacy and Pharmaceutical Sciences and the San Diego Supercomputer Center (SDSC) at the University of California San Diego (UCSD). © 2008, The Regents of the University of California |