Pathway-PDT: A Family-Based Pathway Analysis Method

• Activity Number: 337
• Type: Topic Contributed
• Date/Time: Tuesday, August 2, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Ren-Hua Chung*+ and YoSon Park and Margaret A. Pericak-Vance and Eden R. Martin and Richard W. Morris
• Companies: University of Miami Miller School of Medicine and University of Miami Miller School of Medicine and University of Miami Miller School of Medicine and University of Miami Miller School of Medicine and University of Miami Miller School of Medicine
• Keywords: Pathway analysis ; PDT ; Family-based association

Abstract

Pathway analysis is useful for identifying the joint effects of genes grouped into biologically-based pathways on disease. Pathway analysis may be more powerful than single-marker association tests to identify variants with modest individual effects on a disease but accumulating across genes in a pathway. The development of pathway analysis methods has focused on using unrelated case-control datasets or p-values from genome-wide association tests. We developed Pathway-PDT, a family-based pathway analysis method and an extension of the Pedigree Disequilibrium Test (PDT) (Martin. AJHG 2000). Pathway-PDT defines a score for each gene based on the most significant PDT statistic from PDT statistics for SNPs within a gene (and 20 kb flanking region). Then the weighted Kolmogorov-Smirnov-like running-sum statistic (Wang. AJHG 2007) is calculated for each pathway. A permutation procedure is used in Pathway-PDT to approximate the distribution of the running-sum statistic. We used simulations to verify that Pathway-PDT maintains correct type I error rates under different scenarios. Our simulation results also suggested that Pathway-PDT can have more power than methods using p-values only.