Project 2

Computational Approaches to Identify Disease-Associated Polymorphisms (PI: H. Garner)

The objective of the PGA program is to provide valuable resources to the NHLBI research community. The Southwestern PGA, along with others, has been compiling SNPs found in various patient (sample) populations of various size and disease state, and the annotating that data to make it of high utility. One fundamental approach has been to re-sequence small populations to discover frequently appearing SNPs which can then be evaluated en masse using htp technologies for larger populations. The computer group in the Southwestern PGA has supported this effort by establishing the software needed for the htp re-sequencing pipeline, along with automated annotation and publication to the web data warehouse. In a different but complementary approach for identifying DNA variants that may play a role in heart disease, the computation group has devised computational methods to select and rank potential phenotype-causing DNA polymorphisms, of both the Single Nucleotide Polymorphism (SNP) and repeat (microsatellite) types. In this next phase of the PGA we will emphasize the refinement of these programs and the creation of a validated list of polymorphisms relative to heart disease. Our specific aims will be:

Specific Aim 1:
Refine our polymorphism prediction methods (SNPs and repeats) and produce ready-to-use ranked lists of genes and gene regions, especially for transcription factors and genes involved in lipid metabolism, enabling a directed experimental search for important new alleles.

Specific Aim 2:
Create a verified genotype resource of 1,000 predicted and ranked SNPs using the Perlegen assay 600 patients and an additional 1,000 predicted SNPs measured by Perlegen assay in the Reynold’s cardiac study population of 3,554 individuals with all high-impact alleles being verified by TaqMan assay in our lab. We will also genotype all repeat-containing exons and test the computed reagents in collaboration with Project 1. We will analyze 150 regions annually in cardiomyopathy and congenital heart disease populations and a normal population, and thus screen for polymorphism for every known coding region repeat by year 4. Those results will be annotated, analyzed for impact and made available via our data warehouse.

Specific Aim 3:
Create a variety of computational tools to advance the goals of all projects, including computational resources and reagents for the design and analysis of experiments for the identification of transcription factors and genes involved in lipid metabolism (Project 1), and a relational database on serum proteins (Project 3). We will also create education modules to effectively instruct code and data users.

Specific Aim 4:
Maintain, update and organize all data for easy access via the web (http://pga.swmed.edu) while submitting all data to the appropriate public repositories.