Sampling Strategies for Rare Variant Tests

• Activity Number: 640
• Type: Invited
• Date/Time: Thursday, August 4, 2011 : 10:30 AM to 12:20 PM
• Sponsor: ENAR
• Author(s): Sebastian Zoellner*+
• Companies: University of Michigan
• Keywords: Rare variants ; Family ; Genetics ; Transmission

Abstract

Recent advances in sequencing technology allow assessing the contribution of rare variation to the heritable risk of common disorders. Methods have been developed to combine functional rare variants across a gene and compare this aggregate statistic between cases and controls. However, it is presently still costly to sequence many individuals. Hence it is necessary to select case samples that result in powerful tests under realistic model assumptions. Power can be increased by select cases that are highly likely to carry risk variants. Here we consider the strategy of sequencing unrelated cases ascertained to have affected family members and comparing their aggregate statistic with the aggregate test statistic of random control individuals. . We quantify this gain in power and provide criteria for sample selection under different models of inheritance. We compare the sample sizes required to achieve 80% power between samples of cases with one affected family member and samples of random cases. For realistic effect sizes the sample of random cases has to be 2-5 fold larger to achieve the same power as a sample of cases ascertained to have an affected family member.