The evolution of genetic architectures underlying quantitative traits

TL;DR

This paper presents a population-genetic model explaining how the genetic architecture of quantitative traits evolves under different selection pressures, unifying diverse empirical observations from human traits and yeast gene expression.

Contribution

It introduces a simple theoretical framework predicting how selection influences the number and effect sizes of loci underlying traits.

Findings

Traits under moderate selection involve many loci with variable effects.

Strong or weak selection lead to fewer loci controlling traits.

The model explains empirical patterns across species and traits.

Abstract

In the classic view introduced by R. A. Fisher, a quantitative trait is encoded by many loci with small, additive effects. Recent advances in QTL mapping have begun to elucidate the genetic architectures underlying vast numbers of phenotypes across diverse taxa, producing observations that sometimes contrast with Fisher's blueprint. Despite these considerable empirical efforts to map the genetic determinants of traits, it remains poorly understood how the genetic architecture of a trait should evolve, or how it depends on the selection pressures on the trait. Here we develop a simple, population-genetic model for the evolution of genetic architectures. Our model predicts that traits under moderate selection should be encoded by many loci with highly variable effects, whereas traits under either weak or strong selection should be encoded by relatively few loci. We compare these theoretical predictions to qualitative trends in the genetics of human traits, and to systematic data on the genetics of gene expression levels in yeast. Our analysis provides an evolutionary explanation for broad empirical patterns in the genetic basis of traits, and it introduces a single framework that unifies the diversity of observed genetic architectures, ranging from Mendelian to Fisherian.