Detecting Population Substructure in Rare Variants and Common Variants Data

• Activity Number: 190
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Dandi Qiao*+
• Companies: Harvard University
• Keywords: Population substructure ; stratification ; rare variants

Abstract

In population-based association studies, population substructure gives rise to false positive results. Existing methods to adjust for population substructure are designed mostly to detect global stratification in common variants analysis. There is no formulated way to detect local stratification for either common variants or rare variants analysis. In this paper, we propose a novel test for population-based data to detect outliers and local population stratification in both common variants and rare variants analysis. We show by simulation that under few assumptions, this test has adequate type-I error and sufficient power in detecting global and local population stratification with both common variants data and sequence data. Also, we show that the test could be utilized in detecting subjects with unacceptable genotyping quality. We illustrate the test by applying it to a real datasets including German and Dutch subjects.