A Comparison of Methods for Association Mapping of Nominal Responses with Rare Variants

• Activity Number: 253
• Type: Contributed
• Date/Time: Monday, July 30, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Janae Nicholson*+ and Lan Zhu
• Companies: Oklahoma State University and Oklahoma State University
• Keywords: genome association ; rare variants ; nonparametric association method ; nominal response

Abstract

An allele with a minor allele frequency of 5% or less is termed a rare variant. Methods for identifying rare variants that are responsible for nominal phenotypes with more than two categories are lacking in the statistical literature. The few published methods that could be used in this case have not been adequately evaluated. In this study, we propose a nonparametric approach to detect the association of multi-level nominal responses with rare variants by extending Madsen and Browning's method (2009) where only case control data were considered. The performance of our proposed method is evaluated and compared with existing methods for nominal responses through simulations under different scenarios. Factors that may affect the power and type I error of the methods are explored, including sample size, heritability, and number of phenotypic categories in the population. Results of this study provide a guideline for researchers to select the most appropriate method for their own data analysis.