Combining Dependent P-Values Using Generalizations of Gamma Distribution with Applications to Multi-Trait Association

• Activity Number: 410
• Type: Topic Contributed
• Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Gang Zheng*+ and Qizhai Li
• Companies: National Heart, Lung and Blood Institute and Academy of Mathematics and Systems Science, CAS
• Keywords: dependent tests ; Fisher's combination ; gamma distributions ; pleiotrophic associations ; genetic association ; GWAS

Abstract

Fisher's combination of p-values is a classical approach to combine independent test statistics. When the test statistics are dependent, the gamma distribution is most commonly used to fit the Fisher's combination test. We propose to use two generalizations of the gamma distribution for the Fisher's combination test. Our results show that both generalizations have better control type I error rates than the gamma distribution at more extreme tails. Applications of the results to genetic pleiotrophic associations are described, where multiple traits are tested for association with a marker.