PREDICTING FUNCTIONAL EFFECTS OF NSSNPS USING MULTIPLE INFORMATION SOURCES

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. The broad goal is to contribute to the functional characterization of proteins by integrating and cross-linking a variety of information sources and making them available via a public web server. We are developing a computational method for predicting the functional effects of non-synonymous SNPs, using a combination of physical and statistical potentials, protein sequence alignments, know and predicted structures, protein-protein interactions, and clinical studies. The specific aims are: Aim 1: Quantitatively identify the most informative sequence- and structure-based features that can be used to predict whether an nsSNP has a functional effect. Aim 2: Develop and validate an algorithm for combining the features so as to best classify nsSNPs. The input to the algorithm will be a protein sequence and one for more candidate nsSNPs. The output will be a prediction of whether the nsSNPs have a functional effect (quantified as a probability and a statistical significance measure) and an explanation of the prediction, which features were used and their relative importance. The method will be validated computationally with several data sets: proteins for which comprehensive mutagenesis experiments and functional assays have been performed, and nsSNPs identified as being neutral or disease-associated in SNP databases. Aim 3: Implement the method in a software package and make it accessible as a web server. Aim 4: Apply the method in collaboration with the UCSF Pharmocogenetics of Membrane Transporters (PMT) project (Leabman et al., 2003) including the laboratories of Kathy Giacomini and Deanna Kroetz. Aim 5: Use the successful predictions made by the classifier to understand why certain feature combinations are effective. Aim 6: Create a database of functional predictions for all available nsSNPs and keep it up to date. Aim 7: Cross-link the web resource with two larger web databases: the Modbase collection of protein homology models and the U.C. Santa Cruz Gene Family Browser. We use the Resource for Biocomputing, Visualization, and Informatics (RBVI) to access machine-readable formats of PMT SNP data. This data is used to develop predictive features useful in characterizing non-synonymous SNPs.