Estimating genomic diversity and population differentiation – an empirical comparison of microsatellite and SNP variation in Arabidopsis halleri

Abstract

BackgroundMicrosatellite markers are widely used for estimating genetic diversity within and differentiation among populations. However, it has rarely been tested whether such estimates are useful proxies for genome-wide patterns of variation and differentiation. Here, we compared microsatellite variation with genome-wide single nucleotide polymorphisms (SNPs) to assess and quantify potential marker-specific biases and derive recommendations for future studies. Overall, we genotyped 180 Arabidopsis halleri individuals from nine populations using 20 microsatellite markers. Twelve of these markers were originally developed for Arabidopsis thaliana (cross-species markers) and eight for A. halleri (species-specific markers). We further characterized 2 million SNPs across the genome with a pooled whole-genome re-sequencing approach (Pool-Seq).ResultsOur analyses revealed that estimates of genetic diversity and differentiation derived from cross-species and species-specific microsatellites differed substantially and that expected microsatellite heterozygosity (SSR-H e) was not significantly correlated with genome-wide SNP diversity estimates (SNP-H e and θ Watterson) in A. halleri. Instead, microsatellite allelic richness (A r) was a better proxy for genome-wide SNP diversity. Estimates of genetic differentiation among populations (F ST) based on both marker types were correlated, but microsatellite-based estimates were significantly larger than those from SNPs. Possible causes include the limited number of microsatellite markers used, marker ascertainment bias, as well as the high variance in microsatellite-derived estimates. In contrast, genome-wide SNP data provided unbiased estimates of genetic diversity independent of whether genome- or only exome-wide SNPs were used. Further, we inferred that a few thousand random SNPs are sufficient to reliably estimate genome-wide diversity and to distinguish among populations differing in genetic variation.ConclusionsWe recommend that future analyses of genetic diversity within and differentiation among populations use randomly selected high-throughput sequencing-based SNP data to draw conclusions on genome-wide diversity patterns. In species comparable to A. halleri, a few thousand SNPs are sufficient to achieve this goal.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-016-3459-7) contains supplementary material, which is available to authorized users.

Natural selection and neutral evolutionary processes contribute to genetic divergence in leaf traits across a precipitation gradient in the tropical oak Quercus oleoides

“…Alternatively, single‐nucleotide polymorphisms (SNPs) might be more adequate for Q ST ‐ F ST comparisons because of their presumably lower mutation rates (Leinonen et al., ). Studies using both microsatellites and SNPs for genotyping the same set of populations have shown lower heterozygosity and higher population differentiation for SNPs (e.g., DeFaveri, Viitaniemi, Leder, & Merilä, ; Hauser, Baird, Hilborn, Seeb, & Seeb, ; Narum et al., ; Rajora, Eckert, & Zinck, ; Rodríguez‐Quilón et al., ), although the opposite has been also true (e.g., Fischer et al., ; Gärke et al., ; Sork et al., ). In other words, we did not find a general pattern regarding the F ST obtained from microsatellites vs. SNPs in the literature but, in any case, the population differentiation obtained from one or another type of marker tends to be similar.…”

Whole‐genome sequencing approaches for conservation biology: Advantages, limitations and practical recommendations

“…With the 50 most informative loci, they created a SNP panel for the genetic identification and monitoring of native and introduced bees. Fischer et al, () compared genetic diversity and population differentiation estimates from Pool‐seq and microsatellite data of 9 wild populations of the plant A. halleri in south‐eastern Switzerland and northern Italy. They found no concordance of expected heterozygosity ( H e ) estimates between marker types, and microsatellite allelic richness was a better descriptor of genomewide diversity than H e .…”

Urbanization reduces genetic connectivity in bobcats (Lynx rufus) at both intra– and interpopulation spatial scales

“…Next–generation sequencing technologies have greatly enhanced our ability to accurately estimate neutral genomic variation compared to microsatellites (Fischer et al, ; Helyar et al, ; Santure et al, ). Coupling genomic data with rigorous landscape genetic approaches that incorporate the individual (as opposed to the population) as the statistical unit provides considerable power for identifying genetic variation at fine spatial scales to quantify functional connectivity (Cushman & Landguth, ; Holderegger & Wagner, ; Manel, Schwartz, Luikart, & Taberlet, ).…”