The determinants of genetic susceptibility for most common traits are complex, likely to involve contributions from multiple gene variants. These variations in gene function that affect critical phenotypes can represent amino acid replacements as well as extragenic differences that might affect expression levels; in most instances, these are single nucleotide polymorphisms (SNPs). Although one would like to assay for these gene variants in a completely unbiased fashion, ultimately measuring the contribution of every variant of every gene, this is obviously impractical at this point in time. An alternative strategy is to identify those genes most likely to make contributions to disease variation, identify the variations within this group of genes, and then conduct association studies to link gene variants with disease phenotype. Using cardiovascular disease as a model system, this proposal describes a multi-dimensional approach to this overall problem. In particular, we will use multiple methods for the identification of candidate genes most likely to make contributions to disease variation within populations of patients. This work will take advantage of three unique clinical resources here at Duke. First, we have begun the collection of a large series of aorta samples from heart transplant donors as a source of vascular tissue for gene expression analysis. These samples are unique in the volume of the samples (hundreds) as well as the range of phenotype: early stages of atherosclerosis to advanced forms of the disease. As such, it provides an opportunity to match gene expression profiles with the development of disease in a very unique way. This will be the focus of work in Component 1 as well as statistical efforts in Component 5. Second, a large study of the genetics of early onset cardiovascular disease, representing a collaboration between Duke investigators and GlaxoSmith-Kline, offers the opportunity to identify loci that are linked with the development of disease. This provides a mechanism for the identification of additional candidate genes without any bias whatsoever, including whether the gene actually functions within cardiovascular tissue or not. This represents the focus of Component 2, and bioinformatic efforts in Component 6. The combined efforts of Component I and 2, taking different approaches to the identification of candidate genes, will then be the source of substrate to discover SNPs within this group of genes (Component 3). Much of this work will take advantage of existing information regarding SNPs as well as other major studies directed at cardiovascular disease. But, it will also necessitate efforts within this component to identify as exhaustively as possible those sequence variants that can then be the subject for assays in clinical populations. Third, and possibly the most important asset of this program, is the Duke Cardiovascular Database, an effort initiated some thirty years ago at Duke to follow the clinical course of every cardiovascular disease patient. As such, we now have access to over 40,000 patients who are being followed on a regular basis, creating a clinical dataset that is unmatched. This clinical dataset provides a completely unique resource for this study, both from the quantity as well as the quality of patient clinical data to allow the validation of candidate gene variants with disease variation. Thus, the SNPs identified in Component 3 will go into an expansive genotyping program (Component 4), to bring these candidate genes to a point of validation. A major challenge in an undertaking of such magnitude will be the statistical power to find associations in complex situations. Component 5, will develop the methodologies for understanding the complex gene expression datasets, will also develop the statistical approaches to the analysis of the complex genotyping studies. The program will also enhance and integrate with existing and developing educational programs in bioinformatics and genome technology at Duke (Component 6). This synergy will be of clear benefit to the program, bringing in talented investigators from multiple disciplines in each area critical for the program. Hence, our project will advance the frontiers of genome sciences and technology in the field of common traits. The culmination of our program will provide essential tools to clinicians to improve risk stratification of patients and to design novel preventive and therapeutic strategies.
Showing the most recent 10 out of 41 publications
