Location
Mathematics/Psychology : 103
Date & Time
September 12, 2014, 11:00 am – 12:00 pm
Description
Speaker
Dr. Chuanhua Julia Xing
Medimmune
Title
Weighted Estimating Equations for Analysis of Secondary Phenotypes in Case-Control Genetic Association Studies
Abstract
The case-control study is a popular approach to assess association between genetic exposures and disease by measuring statistical differences between cases (diseased) and controls (disease-free). There is considerable interest in assessing the association of multiple phenotypes in genome-wide sequencing data in case-control studies for gaining power. However, the case-control sample represents a biased sample for estimating the effect of exposures on secondary phenotypes. We propose a general approach for estimating and testing the population effect of a genetic variant on secondary phenotypes. Our approach is based on weighted estimating equations using a conditional probability of an individual being a case or a control as the corresponding weight. Our model is substantially more robust to model misspecification, and out-performs traditional inverse probability weighting and likelihood based methods, both in terms of validity and power. Our model is also much more computationally efficient. The advantages made our approach a practical tool for genome-wide genetic association studies. The talk extends to our recent explorations in multiple secondary phenotypes and sequencing data analysis.
Abstract
The case-control study is a popular approach to assess association between genetic exposures and disease by measuring statistical differences between cases (diseased) and controls (disease-free). There is considerable interest in assessing the association of multiple phenotypes in genome-wide sequencing data in case-control studies for gaining power. However, the case-control sample represents a biased sample for estimating the effect of exposures on secondary phenotypes. We propose a general approach for estimating and testing the population effect of a genetic variant on secondary phenotypes. Our approach is based on weighted estimating equations using a conditional probability of an individual being a case or a control as the corresponding weight. Our model is substantially more robust to model misspecification, and out-performs traditional inverse probability weighting and likelihood based methods, both in terms of validity and power. Our model is also much more computationally efficient. The advantages made our approach a practical tool for genome-wide genetic association studies. The talk extends to our recent explorations in multiple secondary phenotypes and sequencing data analysis.
Tags: