Genetic Interactions with Sex Make a Relatively Small Contribution to the Heritability of Complex Traits in Mice

Tatjana Adamovic,Jon Krohn,Doug Speed,Rupert Palme,Chadi Touma,Richard Mott,Jonathan Flint

doi:10.1371/journal.pone.0096450

Abstract

The extent to which sex-specific genetic effects contribute to phenotypic variation is largely unknown. We applied a novel Bayesian method, sparse partitioning, to detect gene by sex (GxS) and gene by gene (GxG) quantitative loci (QTLs) in 1,900 outbred heterogeneous stock mice. In an analysis of 55 phenotypes, we detected 16 GxS and 6 GxG QTLs. The increase in the amount of phenotypic variance explained by models including GxS was small, ranging from 0.14% to 4.30%. We conclude that GxS rarely make a large overall contribution to the heritability of phenotypes, however there are cases where these will be individually important.

Highlights

Genome-wide association studies (GWAS) typically seek only main effects of genetic variation on phenotypes
It is possible that sex effects are manifest in a subset of the main effect quantitative trait loci (QTL), but it is possible that they represent a completely different set of loci whose biological function is restricted to the sex specific features of the phenotype
Our principle finding is that GxS interactions contribute little to phenotypic variation in addition to that attributable to main effect

Summary

Introduction

Genome-wide association studies (GWAS) typically seek only main effects of genetic variation on phenotypes. While this methodology has succeeded in identifying quantitative trait loci (QTL), there are two reasons for being interested in interaction of genetics with other factors, such as sex and aspects of the organisms’ environment. Lander and colleagues have recently argued that a significant portion of ‘missing heritability’ in human GWAS is not due to the failure to detect sequence variants that contribute to phenotypic variation, but is hidden within unacknowledged interactions [1]. Identifying QTL involved in interactions might be important for understanding specific mechanisms, such as the biology of sex differences. It is possible that sex effects are manifest in a subset of the main effect QTL, but it is possible that they represent a completely different set of loci whose biological function is restricted to the sex specific features of the phenotype

Methods

Results

Conclusion