Statistical Issues in the Analysis of Disease Association with Next-Generation Sequence Data

• Activity Number: 275
• Type: Invited
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: ENAR
• Author(s): Danyu Lin*+
• Companies: The University of North Carolina
• Keywords: next-generation sequencing ; rare variants ; copy number variants ; single nucleotide polymorphisms ; quantitative traits ; selective genotyping

Abstract

Next-generation sequencing technologies are being employed in an increasing number of genetic associations studies. There are many unique statistical challenges in the designs and analysis of such studies. In this talk, I will present an overview of these challenges, focusing on three different issues: (1) analysis of disease association with rare variants; (2) analysis of disease association with copy number variations; (3) analysis of quantitative traits with outcome-dependent sampling. There are strong motivations for addressing these three issues: (1) because we are particularly interested in rare variants detected by next-generation sequencing technologies, standard statistical methods based on asymptotic distributions are inappropriate; (2) the copy number variations are not directly measured, and the intensity data from next-generation sequencing technologies are quite different from genome-wide SNP array data; (3) genotyping individuals with extreme phenotype values can be very cost-effective for large cohort studies. We will describe our solutions to these problems and illustrate them with simulated and real data.