A New Class of Test Statistics for Detecting Rare Disease-Associated Genetic Variants

• Activity Number: 588
• Type: Contributed
• Date/Time: Wednesday, August 3, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): John Ferguson*+ and Hongyu Zhao and Judy H. Cho and Joshua Sampson
• Companies: Yale University and Yale University and Yale University and National Cancer Institute
• Keywords: rare variant ; association ; Score statistic ; Data adaptive ; GWAS ; exon sequencing

Abstract

For many complex genetic diseases, the common disease associated genetic variation has largely been identified via large scale Genome Wide Association Studies. Unfortunately, rare variants that are disease associated are more difficult to find since the number of individuals carrying the rare allele, which is proportional to the effective sample size in a case control study, will be relatively small when the allele is rare. Recently, many techniques that collectively analyze multiple rare variants within a genomic region have been proposed in an effort to increase statistical power. Here we present a new class of test statistics designed for rare variant association analysis that unifies many of these existing approaches under a common framework. In particular, this class of test statistic encompasses and generalizes a large variety of both uni-variate and multivariate Score tests. Data-adaptive techniques are suggested for choosing a particular member of this class in real data situations. We demonstrate the application of this methodology to find Crohn's disease associated regions using data from a special rare-variant enriched genotyping platform.