Elevated Linkage Disequilibrium and Signatures of Soft Sweeps Are Common in Drosophila melanogaster

Abstract

The extent to which selection and demography impact patterns of genetic diversity in natural populations of Drosophila melanogaster is yet to be fully understood. We previously observed that linkage disequilibrium (LD) at scales of 10 kb in the Drosophila Genetic Reference Panel (DGRP), consisting of 145 inbred strains from Raleigh, North Carolina, measured both between pairs of sites and as haplotype homozygosity, is elevated above neutral demographic expectations. We also demonstrated that signatures of strong and recent soft sweeps are abundant. However, the extent to which these patterns are specific to this derived and admixed population is unknown. It is also unclear whether these patterns are a consequence of the extensive inbreeding performed to generate the DGRP data. Here we analyze LD statistics in a sample of .100 fully-sequenced strains from Zambia; an ancestral population to the Raleigh population that has experienced little to no admixture and was generated by sequencing haploid embryos rather than inbred strains. We find an elevation in long-range LD and haplotype homozygosity compared to neutral expectations in the Zambian sample, thus showing the elevation in LD is not specific to the DGRP data set. This elevation in LD and haplotype structure remains even after controlling for possible confounders including genomic inversions, admixture, population substructure, close relatedness of individual strains, and recombination rate variation. Furthermore, signatures of partial soft sweeps similar to those found in the DGRP as well as partial hard sweeps are common in Zambia. These results suggest that while the selective forces and sources of adaptive mutations may differ in Zambia and Raleigh, elevated long-range LD and signatures of soft sweeps are generic in D. melanogaster.KEYWORDS Drosophila melanogaster; demography; linkage disequilibrium; haplotype homozygosity; selection D ISENTANGLING the effects of demography and selection on patterns of genomic variation remains a central challenge in evolutionary biology. Until recently, inference of demography and selection primarily relied on short genomic fragments sampled from a limited number of individuals (e.g., Pritchard et al. 2000;Molina et al. 2001;Li and Stephan 2006;Thornton and Andolfatto 2006;Gutenkunst et al. 2009;Duchen et al. 2013). While these studies provided important insights into the evolutionary forces acting on populations, they were ultimately limited by both sample size and the physical scale in which patterns of polymorphism and linkage could be investigated. The recent availability of wholegenome sequences from multiple individuals with populations in a variety of species (e.g., Abecasis et al. 2010;Cao et al. 2011;Mackay et al. 2012) is enabling us to examine long-range patterns of linkage disequilibrium (LD) (10 kb), measured either as correlations between pairs of sites or as haplotype homozygosity. LD offers powerful insights into selective and demographic processes shaping genetic variation in a natural populatio...

Molecular Population Genetics

“…These mutations represent most of the genetic variation in the region depauperated by the selective sweep, and their initial frequency is low, so that a region harboring a positively selected variant will also harbor an excess of rare derived alleles. Furthermore, if an allele influenced by recent positive selection increases in frequency faster than local recombination reduces the range of LD between the allele and linked markers, then the region will also show unusually long-range LD (Nielsen 2005; Franssen et al 2015; Garud et al 2015; Garud and Petrov 2016). As a whole, natural selection leaves signatures in the genome that can be used to identify the regions that have been selected, including:

A reduction in the genetic diversity.

A skew toward rare derived alleles.

An increase in the LD.

…”

Soft sweeps and beyond: understanding the patterns and probabilities of selection footprints under rapid adaptation

“…For adaptations with a weak trade-off, also single-locus soft sweeps can contribute, in particular if mutation targets are polygenic (Θ g > Θ l ). Recent genome scans by Garud & Petrov (2016) and Sheehan & Song (2016) for a Zambian D. melanogaster population conform to this expectation. Both find a mix of hard and soft sweep signals for regions with the strongest evidence for recent selection.…”

“…Garud et al . () and Garud & Petrov () report much higher rates for soft sweeps for D. melanogaster from North Carolina. According to their test, all top 50 signals are consistent with soft sweeps rather than hard sweeps.…”

“…It seems that problems can be avoided if scans pre‐select loci with the strongest signals (Garud et al . ; Garud & Petrov , for haplotype homozygosity; N. Garud pers. comm.)…”

Estimating the parameters of background selection and selective sweeps in Drosophila in the presence of gene conversion

“…If soft sweeps are prevalent in Drosophila, as has recently been argued (49), then the same pattern of bias as from a subdivided population would arise (50,51). (Note, however, that gene conversion of a favored mutation onto an ancestral haplotype could generate the appearance of a soft sweep.)…”