Online Program Home
  My Program

Abstract Details

Activity Number: 584 - Statistical Methods for Genetic Association Analysis
Type: Contributed
Date/Time: Wednesday, August 2, 2017 : 2:00 PM to 3:50 PM
Sponsor: Section on Statistics in Genomics and Genetics
Abstract #324681
Title: Improved Rank Truncated Product Method for Combining P-Values
Author(s): Dmitri Zaykin* and Olga Vsevolozhskaya
Companies: NIH/NIEHS and
Keywords: rank truncated product ; P-values ; combined P-value
Abstract:

Dudbridge and Koleman (2003) studied the rank truncated product method (RTP) for combining P-values in genetic research, where the smallest K values of potentially large L number of P-values are combined. RTP was independently developed by us and gained popularity in various extensions, including adaptive variations where K is chosen in a way to maximize power. However, the theoretical form of the RTP distribution is unwieldy and involves a complicated multiple integration. We recently developed new ways of evaluating its distribution that simplifies substantially the analytic form of the combined P-value, down to a single line of R code. We will discuss implications of this development including potential directions for future research.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association