Lack of Low Frequency Variants Masks Patterns of Non-Neutral Evolution following Domestication

Céline H Frère,Ian D Godwin,Edward K Gilding,Peter J Prentis,Alan Cruickshank,Agnieszka M Mudge,Gro V Amdam

doi:10.1371/journal.pone.0023041

Céline H Frère, Ian D Godwin + Show 5 more

Open Access

PDF Available

https://doi.org/10.1371/journal.pone.0023041

Copy DOI

Export

Save

Cite

Abstract
Highlights/Summary
Full-Text PDF
Similar Papers

Abstract

Listen

Detecting artificial selection in the genome of domesticated species can not only shed light on human history but can also be beneficial to future breeding strategies. Evidence for selection has been documented in domesticated species including maize and rice, but few studies have to date detected signals of artificial selection in the Sorghum bicolor genome. Based on evidence that domesticated S. bicolor and its wild relatives show significant differences in endosperm structure and quality, we sequenced three candidate seed storage protein (kafirin) loci and three candidate starch biosynthesis loci to test whether these genes show non-neutral evolution resulting from the domestication process. We found strong evidence of non-neutral selection at the starch synthase IIa gene, while both starch branching enzyme I and the beta kafirin gene showed weaker evidence of non-neutral selection. We argue that the power to detect consistent signals of non-neutral selection in our dataset is confounded by the absence of low frequency variants at four of the six candidate genes. A future challenge in the detection of positive selection associated with domestication in sorghum is to develop models that can accommodate for skewed frequency spectrums.

Highlights

Since the release of the seminal book ‘On the origin of species’ [1], there has been much interest in identifying the evolutionary processes underlying crop and animal domestication
Top-down approaches rely on QTL and LD mapping to isolate candidate genes associated with traits of interest and use molecular population genetic methods to test for selection
Whitt et al [11] found evidence of positive selection at three starch pathway loci and Wright et al [46] estimated 2–4% of maize genome had been the target of recent artificial selection

Summary

Introduction

Since the release of the seminal book ‘On the origin of species’ [1], there has been much interest in identifying the evolutionary processes underlying crop and animal domestication. [2,3] but can benefit future breeding strategies [4]. This is evident when genes can be causally linked to adaptive phenotypes of interest in domesticated species (e.g. quality, pest resistance, drought tolerance, etc.). Top-down approaches rely on QTL and LD mapping to isolate candidate genes associated with traits of interest and use molecular population genetic methods to test for selection. The bottom-up approach relies on molecular population genetic methods to identify regions or candidate genes under selection before aiming to link it back to a phenotypic trait. Evidence for selection has been greatly documented in maize and rice (e.g. [9,10,11]), but few studies have to date detected signals of selection in Sorghum bicolor [12,13]

Methods

Results

Conclusion