Statistical Methods for High Throughput Genomic Data

点击翻译按钮获取中文摘要

This research program will develop and validate new statistical methods for the analysis of high throughput genomic data. Advances in technology over the past decade have made it possible to examine multiple genes or genetic variants simultaneously. These new technologies have facilitated many important discoveries in the genetics of human diseases. There has been some disappointment, however, that more of the genetic heritability has not been explained. In this work my research group and I will examine two possible reasons for that; the presence of interactions among genes or between genes and the environment, and the inherent lack of power in single-marker analyses. We shall propose methods to assist in finding gene-gene and gene-environment interactions as well as methods to use aggregated information from many related markers to improve the power of these analyses. Technological advances have not stopped, however, and recent advances make it possible to directly sequence RNA fragments and so give much better measures of gene expression. This new technology produces quite different data from the array based data of the last decade and new methods are required to properly analyze the data. The few methods currently proposed are good starting points but preliminary work with an undergraduate student suggest that they are all somewhat flawed as the assumptions underlying them do not hold for real data. My research team will develop new methods, with fewer restrictive assumptions, which can be used to reliably analyze this type of data. We will also examine ways of incorporating the new data with existing data collected using gene expression microarrays to leverage maximum power from the large amount of data available. This research will address some of the major statistical issues which arise in the analysis of high-throughput genetic data and thus facilitate further further discoveries on the etiology of human disease in Canada. Training a new generation of statisticians to understand the data from genetic experiments and their analysis will also provide Canadian scientists with a rich pool of talent on which to draw when they conduct these powerful but expensive experiments.