Abstract

Several traits in mammals, such as blood pressure and body temperature, are homeostatic and important to be regulated with little variation. However, most methods in genome-wide association studies (GWAS) are designed to search for markers where there are significant mean phenotypic differences between genotypes. We have developed a method to detect markers linked to genes controlling the variance rather than the mean. The method is based on double generalized linear models (DGLM) that simply iterates between weighted least squares and is computationally feasible for a large number of markers. Recently, it was shown that the method could also be a powerful tool to detect interaction effects bet-ween genes at different locations on the genome or between genes and the environment. Any trait distribution from the exponential family of distributions can be analyzed using DGLM, but the importance of measuring the trait on an appropriate scale should be noted. A common problem in GWAS is accounting for possible population structure. We propose a solution, in the context of detecting variance-controlling genes, where additional polygenic random effects are fitted using hierarchical generalized linear models implemented in the R package hglm on CRAN. Results from mice, humans and plants will be presented.

Published in

Title: Nordstat 2012, 24th Nordic Conference in Mathematical Statistics

UKÄ Subject classification

Genetics and Breeding


Permanent link to this page (URI)

https://res.slu.se/id/publ/42222