Identifying Genetic Factors for Predisposition in Polygenic Diseases

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): This proposal focuses on the development of an algorithm for determining the underlying factors responsible for predisposition to or protection from polygenic diseases. The algorithm relies on Markov chain Monte Carlo exploration of the space of possible genetic variants coupled to a Bayesian statistical test based on phenotypic ranks. The developed algorithm will improve our ability to reduce complex genetic interactions from the growing genotypic and single nucleotide polymorphism databases. The identification of interacting genetic variants placing individuals at risk for or providing protection from the development of polygenic diseases remains a problem both for our understanding of these diseases and for our ability to develop new treatments. Fundamentally, the problem involves the inability of standard statistical approaches to achieve power in the face of the enormous growth in our knowledge of genomics, brought about by the various genome projects and high throughput single nucleotide polymorphism (SNP) and genotype analyses. Similar "curse of dimensionality" problems have arisen in other fields, and Bayesian statistical approaches coupled to Markov chain Monte Carlo (MCMC) techniques have led to significant improvements in understanding, which has led to our focus on this technique here. Because polygenic diseases are much more widespread than single gene diseases, the potential impact on health is substantial, and many common diseases are believed to have a polygenic basis, including obesity, cardiac disease, and Type II diabetes. A method to dissect the complex genetic interactions underlying predisposition to or protection from polygenic diseases would have a substantial effect on improving health. We will disseminate the algorithm through publications and presentations at conferences. We will also through contact individuals in foundations focused on research in specific complex diseases, so that the algorithm can have the maximal impact on health. Upon successful completion of this project, we plan to develop the algorithm more fully. We would like to modularize the algorithm, allowing easier inclusion of different prior distributions and implement a more friendly interface with the ability to utilize the emerging bioinformatics standards for data exchange.