Secondary Analysis in GWAS

• Activity Number: 186
• Type: Invited
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Huilin Li*+ and Mitchell Gail
• Companies: New York University and National Cancer Institute
• Keywords: secondary analysis ; case-control study ; GWAS ; maximum likelihood estimation

Abstract

Case-control genome-wide association studies provide a vast amount of genetic information that may be used to investigate secondary phenotypes. We study the situation in which the secondary phenotype and genetic markers are dichotomous. We first prove that with disease rate is known, the inverse-probability-of-sampling-weighted (IPW) regression method is exactly the maximum likelihood estimation method using the full disease model. Those two methods are the most robust methods in term of guarding the possibility of interaction effect of genetic variants and secondary phenotype on the disease. When there is no interaction effect, the maximum likelihood estimation method with the no interaction assumption is the most efficient method. To strike a balance of the above methods, we proposed an adaptively weighted method that combines the IPW and MLE with reduced disease model. Our adaptively weighted method is always unbiased and has reduced mean square error for estimation with a pre-specified gene and increase the power to discover a new association in a genome-wide study when non-zero interaction is possible. Case-control study with known population totals is also investigated rega