Online Program Home
My Program

Abstract Details

Activity Number: 71 - Longitudinal/Correlated Data
Type: Contributed
Date/Time: Sunday, July 28, 2019 : 4:00 PM to 5:50 PM
Sponsor: Biometrics Section
Abstract #305337
Title: A Comparison of the Power of Generalized Linear Regressions (GLM) and Generalized Estimating Equations (GEE) in the Phenome –Wide Association Study (PheWAS) Setting
Author(s): Minh Chau Nguyen* and Erin Austin
Companies: University of Colorado Denver and University of Colorado Denver
Keywords: GLM; GEE; Multivariate; Correlation; Hypotheisis tests; Pleiotropy

One goal of PheWAS is to identify pleiotropy where a Single Nucleotide Polymorphism (SNP) controls multiple traits. Aspects that need the most consideration in detecting pleiotropy are the choice of method and the hypothesis tests. Because of its simplicity, the Generalized Linear Model with multiple testing correction is popular in the univariate approach. Generalized Estimation Equations is one favored method among the multivariate approaches since it takes correlations between traits into account. We apply both methods on a diverse range of simulated PheWAS data sets with varying minor allele frequencies at a SNP, different frequencies of traits associated and not associated with the SNP, varying phenotype prevalences, and varying correlation architectures between traits. We compare results from association tests using the traditional Bonferroni correction and the principal-component- based simpleM correction with those from the sequential test for pleiotropy developed by Schaid et al. in 2017. The project aims to compare power gained in each method while controlling the family-wide error rate to make conclusions on the best methods to use in different settings.

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

Back to the full JSM 2019 program