All ETDs from UAB

Advisory Committee Chair

Nengjun Yi

Advisory Committee Members

Sadeep Shrestha

Marguerite Irvin

Xiangqin Cui

Document Type


Date of Award


Degree Name by School

Doctor of Philosophy (PhD) School of Public Health


In the last two decades, thousands of genome-wide association (GWA) analyses have been published. These studies have been primarily conducted using data from Caucasians and many of the reported findings have not been replicated in other populations, perhaps reflecting that the underlying architecture of effects may vary across ethnically diverse groups. Similarly, the possibility that the genetic architecture of traits may vary between males and females has also been overlooked. We propose a methodology for estimation and testing of effect heterogeneity between groups such as sex or ethnicity that involves modeling genotype-by-group random interactions. Using simulations, we show that the proposed methods yield nearly unbiased estimates of the correlation of effects between groups with moderately large sample sizes, in contrast to the simple correlation of estimated effects from stratified GWA analyses. For hypothesis testing, we show that the proposed methods yield the desired type-I error rates and higher power than tests based on fixed-effect interaction models. Using data from the Atherosclerosis Risk In Communities Study (N=8,228 subjects, p=828,822 SNPs) we assessed the extent of effect heterogeneity between European Americans and African Americans for four complex traits (human height, serum urate, LDL and HDL cholesterol). We found that the correlation of effects varied substantially among traits; it was lower for traits affected by lifestyle (LDL cholesterol and serum urate) and higher for height. Interestingly, the correlation of effects was far smaller for low-variance effects than for high-variance effects suggesting that epistasis may play a role in effect heterogeneity. Finally, using data from the interim release of the UK-Biobank (N=102,643 Caucasians, p=580,028 SNPs) we studied sex differences for anthropometric and obesity-related traits. We again found substantial differences in effect correlations between sexes among the considered traits and report novel sexually dimorphic genes for waist-to-hip ratio.

Included in

Public Health Commons



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.