Multilocus Analysis in Genetic Studies
Nicholas
Schork
Scripps Genomic Medicine, Scripps Health &
Department of Molecular and Experimental Medicine
The Scripps Research Institute
Friday, March 28, 2008, 12:30–1:30 pm
GEMS classroom, 3rd Floor in
Shriner's Building
Coffee, tea, and cookies will be provided
Abstract
Most diseases of contemporary public health concern are multifactorial in nature, with many genetic and non-genetic determinants. Recent genome-wide association studies (GWAS) corroborate this to a great degree, as they have identified only one or a few genetic variations unequivocally associated with a few diseases that explain a very small fraction of the disease burden in the population at large. This suggests that either the diseases of interest are influenced by largely epigenomic or non-genetic factors, or the genetic factors that influence them have effects too small to be detected within the GWAS framework. In this talk, a discussion of statistical methods that might be more appropriate for detecting multiple genetic variations with small, but cumulative, effects on the disease of interest within the GWAS framework is provided. These methods take advantage of genomic similarity measures, regularized regression, bioinformatically-informed computational assessments of the function of genetic variations, and pathway analysis constructs.