TESS3: fast inference of spatial population structure and genome scans for selection

Geography and landscape are important determinants of genetic variation in natural populations, and several ancestry estimation methods have been proposed to investigate population structure using genetic and geographic data simultaneously. Those approaches are often based on computer‐intensive stochastic simulations and do not scale with the dimensions of the data sets generated by high‐throughput sequencing technologies. There is a growing demand for faster algorithms able to analyse genomewide patterns of population genetic variation in their geographic context. In this study, we present TESS3 , a major update of the spatial ancestry estimation program TESS . By combining matrix factorization and spatial statistical methods, TESS3 provides estimates of ancestry coefficients with accuracy comparable to TESS and with run‐times much faster than the Bayesian version. In addition, the TESS3 program can be used to perform genome scans for selection, and separate adaptive from nonadaptive genetic variation using ancestral allele frequency differentiation tests. The main features of TESS3 are illustrated using simulated data and analysing genomic data from European lines of the plant species Arabidopsis thaliana.     

Identifying coevolving loci using interspecific genetic correlations

Evaluating the importance of coevolution for a wide range of evolutionary questions, such as the role parasites play in the evolution of sexual reproduction, requires that we understand the genetic basis of coevolutionary interactions. Despite its importance, little progress has been made identifying the genetic basis of coevolution, largely because we lack tools designed specifically for this purpose. Instead, coevolutionary studies are often forced to re‐purpose single species techniques. Here, we propose a novel approach for identifying the genes mediating locally adapted coevolutionary interactions that relies on spatial correlations between genetic marker frequencies in the interacting species. Using individual‐based multi‐locus simulations, we quantify the performance of our approach across a range of coevolutionary genetic models. Our results show that when one species is strongly locally adapted to the other and a sufficient number of populations can be sampled, our approach accurately identifies functionally coupled host and parasite genes. Although not a panacea, the approach we outline here could help to focus the search for coevolving genes in a wide variety of well‐studied systems for which substantial local adaptation has been demonstrated.     

A program to compare genetic differentiation statistics across loci using resampling of individuals and loci

Comparisons of genetic differentiation across populations based on different loci can provide insight into the evolutionary patterns acting on various regions of genomes. Here, we develop a program to statistically compare population genetic differentiation statistics (F(ST) or G'(ST) ) calculated from different loci. The program employs a routine that resamples either or both of individuals and loci and calculates a bootstrap confidence interval in the statistics. Resampling individuals is important when fewer than 25 individuals are sampled per population and when confidence intervals are required for individual loci. Resampling loci provides confidence intervals for sets of loci, such as a set presumed to be neutral, but can be anticonservative if fewer than 20 loci are analysed. We demonstrate the program using previously published data on the genetic differentiation at a major histocompatibility complex locus and at microsatellite loci across 10 populations of the guppy (Poecilia reticulata).     

Detecting strong positive selection in the genome

New statistical tests have been developed in the past decade that enable us to infer evidence of recent strong positive selection from genome-wide data on single-nucleotide polymorphism and to localize the targets of selection in the genome. Based on these tests, past demographic events that led to distortions of the site-frequency spectrum of variation can be distinguished from selection, in particular if linkage disequilibrium is taken into account. These methods have been successfully applied to species from which complete sequence information and polymorphism data are available, including Drosophila melanogaster, humans, and several plant species. To make full use of the available data, however, the tests that were primarily designed for panmictic populations need to be extended to spatially structured populations.     

Interpreting population differentiation in terms of drift and selection

Summary Many empirical studies demonstrate some degree of genetic differentiation among populations of the same species. Understanding the relative importance of the processes causing this genetic differentiation has proven to be a difficult task. In particular, population differentiation can be influenced primarily by selection, genetic drift, and migration. We review the effect of drift and migration on patterns of genetic variation, with special reference to the conditions necessary for population differentiation. Conceptually, selection may be implicated in cases of population differentiation if the effect of drift and migration can be shown to be insufficient to cause the observed patterns. We examine some of the pitfalls of this approach when used with allozyme data, and revise a previous conclusion concerning the relative importance of selection in poulations of scale insects.     

Controlling false discoveries in genome scans for selection

Population differentiation (PD) and ecological association (EA) tests have recently emerged as prominent statistical methods to investigate signatures of local adaptation using population genomic data. Based on statistical models, these genomewide testing procedures have attracted considerable attention as tools to identify loci potentially targeted by natural selection. An important issue with PD and EA tests is that incorrect model specification can generate large numbers of false‐positive associations. Spurious association may indeed arise when shared demographic history, patterns of isolation by distance, cryptic relatedness or genetic background are ignored. Recent works on PD and EA tests have widely focused on improvements of test corrections for those confounding effects. Despite significant algorithmic improvements, there is still a number of open questions on how to check that false discoveries are under control and implement test corrections, or how to combine statistical tests from multiple genome scan methods. This tutorial study provides a detailed answer to these questions. It clarifies the relationships between traditional methods based on allele frequency differentiation and EA methods and provides a unified framework for their underlying statistical tests. We demonstrate how techniques developed in the area of genomewide association studies, such as inflation factors and linear mixed models, benefit genome scan methods and provide guidelines for good practice while conducting statistical tests in landscape and population genomic applications. Finally, we highlight how the combination of several well‐calibrated statistical tests can increase the power to reject neutrality, improving our ability to infer patterns of local adaptation in large population genomic data sets.     

Conservation genomics of anadromous Atlantic salmon across its North American range: outlier loci identify the same patterns of population structure as neutral loci

Anadromous Atlantic salmon (Salmo salar) is a species of major conservation and management concern in North America, where population abundance has been declining over the past 30 years. Effective conservation actions require the delineation of conservation units to appropriately reflect the spatial scale of intraspecific variation and local adaptation. Towards this goal, we used the most comprehensive genetic and genomic database for Atlantic salmon to date, covering the entire North American range of the species. The database included microsatellite data from 9142 individuals from 149 sampling locations and data from a medium‐density SNP array providing genotypes for >3000 SNPs for 50 sampling locations. We used neutral and putatively selected loci to integrate adaptive information in the definition of conservation units. Bayesian clustering with the microsatellite data set and with neutral SNPs identified regional groupings largely consistent with previously published regional assessments. The use of outlier SNPs did not result in major differences in the regional groupings, suggesting that neutral markers can reflect the geographic scale of local adaptation despite not being under selection. We also performed assignment tests to compare power obtained from microsatellites, neutral SNPs and outlier SNPs. Using SNP data substantially improved power compared to microsatellites, and an assignment success of 97% to the population of origin and of 100% to the region of origin was achieved when all SNP loci were used. Using outlier SNPs only resulted in minor improvements to assignment success to the population of origin but improved regional assignment. We discuss the implications of these new genetic resources for the conservation and management of Atlantic salmon in North America.     

Temporal variation in genetic diversity and effective population size of Mediterranean and subalpine Arabidopsis thaliana populations

Currently, there exists a limited knowledge on the extent of temporal variation in population genetic parameters of natural populations. Here, we study the extent of temporal variation in population genetics by genotyping 151 genome-wide SNP markers polymorphic in 466 individuals collected from nine populations of the annual plant Arabidopsis thaliana during 4 years. Populations are located along an altitudinal climatic gradient from Mediterranean to subalpine environments in NE Spain, which has been shown to influence key demographic attributes and life cycle adaptations. Genetically, A. thaliana populations were more variable across space than over time. Common multilocus genotypes were detected several years in the same population, whereas low-frequency multilocus genotypes appeared only 1 year. High-elevation populations were genetically poorer and more variable over time than low-elevation populations, which might be caused by a higher overall demographic instability at higher altitudes. Estimated effective population sizes were low but also showed a significant decreasing trend with increasing altitude, suggesting a deeper impact of genetic drift at high-elevation populations. In comparison with single-year samplings, repeated genotyping over time captured substantially higher amount of genetic variation contained in A. thaliana populations. Furthermore, repeated genotyping of populations provided novel information on the genetic properties of A. thaliana populations and allowed hypothesizing on their underlying mechanisms. Therefore, including temporal genotyping programmes into traditional population genetic studies can significantly increase our understanding of the dynamics of natural populations.     

Genetic diversity and differentiation of guanaco populations from Argentina inferred from microsatellite data

Genotype data from 14 microsatellite markers were used to assess the genetic diversity and differentiation of four guanaco populations from Argentine Patagonia. These animals were recently captured in the wild and maintained in semi-captivity for fibre production. Considerable genetic diversity in these populations was suggested by the finding of a total of 162 alleles, an average mean number of alleles per locus ranging from 6.50 to 8.19, and H(e) values ranging from 0.66 to 0.74. Assessment of population differentiation showed moderate but significant values of F(ST)=0.071 (P=0.000) and R(ST)=0.083 (P=0.000). An amova test showed that the genetic variation among populations was 5.6% while within populations it was 94.4%. A number of 6.6 migrants per generation may support these results. Unambiguous individual assignment to original populations was obtained for the Pilcaniyeu, Las Heras and La Esperanza populations. The erroneous assignment of 18.75% Rio Mayo individuals to the Las Heras population can be explained by the low genetic differentiation found between these two populations. Thirty-nine of 56 loci per population combinations were in Hardy--Weinberg disequilibrium because of guanaco heterozygote deficiency, which may be explained by population subdivision. The high level of genetic diversity of the guanacos analysed here indicates that the Patagonian guanaco constitutes an important genetic resource for conservation or economic utilization programmes.     

Whole‐genome sequencing of two North American Drosophila melanogaster populations reveals genetic differentiation and positive selection

The prevailing demographic model for Drosophila melanogaster suggests that the colonization of North America occurred very recently from a subset of European flies that rapidly expanded across the continent. This model implies a sudden population growth and range expansion consistent with very low or no population subdivision. As flies adapt to new environments, local adaptation events may be expected. To describe demographic and selective events during North American colonization, we have generated a data set of 35 individual whole‐genome sequences from inbred lines of D. melanogaster from a west coast US population (Winters, California, USA) and compared them with a public genome data set from Raleigh (Raleigh, North Carolina, USA). We analysed nuclear and mitochondrial genomes and described levels of variation and divergence within and between these two North American D. melanogaster populations. Both populations exhibit negative values of Tajima's D across the genome, a common signature of demographic expansion. We also detected a low but significant level of genome‐wide differentiation between the two populations, as well as multiple allele surfing events, which can be the result of gene drift in local subpopulations on the edge of an expansion wave. In contrast to this genome‐wide pattern, we uncovered a 50‐kilobase segment in chromosome arm 3L that showed all the hallmarks of a soft selective sweep in both populations. A comparison of allele frequencies within this divergent region among six populations from three continents allowed us to cluster these populations in two differentiated groups, providing evidence for the action of natural selection on a global scale.     

Identifying Signatures of Selection in Genetic Time Series

Both genetic drift and natural selection cause the frequencies of alleles in a population to vary over time. Discriminating between these two evolutionary forces, based on a time series of samples from a population, remains an outstanding problem with increasing relevance to modern data sets. Even in the idealized situation when the sampled locus is independent of all other loci, this problem is difficult to solve, especially when the size of the population from which the samples are drawn is unknown. A standard χ²-based likelihood-ratio test was previously proposed to address this problem. Here we show that the χ²-test of selection substantially underestimates the probability of type I error, leading to more false positives than indicated by its P-value, especially at stringent P-values. We introduce two methods to correct this bias. The empirical likelihood-ratio test (ELRT) rejects neutrality when the likelihood-ratio statistic falls in the tail of the empirical distribution obtained under the most likely neutral population size. The frequency increment test (FIT) rejects neutrality if the distribution of normalized allele-frequency increments exhibits a mean that deviates significantly from zero. We characterize the statistical power of these two tests for selection, and we apply them to three experimental data sets. We demonstrate that both ELRT and FIT have power to detect selection in practical parameter regimes, such as those encountered in microbial evolution experiments. Our analysis applies to a single diallelic locus, assumed independent of all other loci, which is most relevant to full-genome selection scans in sexual organisms, and also to evolution experiments in asexual organisms as long as clonal interference is weak. Different techniques will be required to detect selection in time series of cosegregating linked loci.     

Genetic differentiation between Apis cerana cerana populations from Damen Island and adjacent mainland in China

Geographic isolation interrupted gene flow between populations leading to population differentiation during the long evolutionary period. In this paper, 33 colonies from Damen Island and 100 colonies from adjacent mainland populations, Juxi and Chixi, were analyzed with both mitochondrial tRNA^leu-COII sequences and five microsatellite loci. The results showed that Apis cerana cerana population from Damen Island significantly differentiated from its adjacent mainland populations. In addition, Damen Island population showed a lower level of genetic diversity in terms of the number of mitochondrial haplotypes while both island and mainland populations showed a low level of genetic diversity with mutilocus analysis. The divergent small island population A.c. cerana might probably have suffered inbreeding and genetic drift as well as limited gene flow across the strait. Our data provides useful information for management and preservation for the Damen Island population.     

Calculations of population differentiation based on GST and D: forget GST but not all of statistics!

G _ST‐values and its relatives (F_ST) belong to the most used parameters to define genetic differences between populations. Originally, they were developed for allozymes with very low number of alleles. Using highly polymorphic microsatellite markers it was often puzzling that G_ST‐values were very low but statistically significant. In their papers, Jost (2008) and Hedrick (2005) explained that G_ST‐values do not show genetic differentiation, and Jost suggested calculating D‐values instead. Theoretical mathematical considerations are often difficult to follow; therefore, we chose an applied approach comparing two artificial populations with different number of alleles at equal frequencies and known genetic divergence. Our results show that even for more than one allele per population G_ST‐values do not calculate population differentiation correctly; in contrast, D‐values do reflect the genetic differentiation indicating that data based on G_ST‐values need to be re‐evaluated. In our approach, statistical evaluations remained similar. We provide information about the impact of different sample sizes on D‐values in relation to number of alleles and genetic divergence.     

Genomic signatures reveal selection of characteristics within and between Meishan pig populations

The Chinese Meishan pig breed is well known for its high prolificacy. Moreover, this breed can be divided into three types based on their body size: big Meishan, middle Meishan (MMS) and small Meishan (SMS) pigs. Few studies have reported on the genetic signatures of Meishan pigs, particularly on a genome‐wide scale. Exploring for genetic signatures could be quite valuable for revealing the genetic architecture of phenotypic variation. Thus, we performed research in two parts based on the genome reducing and sequencing data of 143 Meishan pigs (74 MMS pigs, 69 SMS pigs). First, we detected the selection signatures among all Meishan pigs studied using the relative extended haplotype homozygosity test. Second, we detected the selection signatures between MMS and SMS pigs using the cross‐population extended haplotype homozygosity and F_ST methods. A total of 111 398 SNPs were identified from the sequenced genomes. In the population analysis, the most significant genes were associated with the mental development (RGMA), reproduction (HDAC4, FOXL2) and lipid metabolism (ACACB). From the cross‐population analysis, we detected genes related to body weight (SPDEF, PACSIN1) in both methods. We suggest that rs341373351, located within the PACSIN1 gene, might be the causal variant. This study may have achieved consistency between selection signatures and characteristics within and between Meishan pig populations. These findings can provide insight into investigating the molecular background of high prolificacy and body size in pig.     

Microsatellite analysis of population structure and genetic differentiation within and between populations of the root vole, Microtus oeconomus in the Netherlands

Eight microsatellite markers for the root vole (Microtus oeconomus) were developed to assess the amount of genetic variation for nine Dutch root vole populations from four different regions, and to evaluate the degree of differentiation and isolation. All eight microsatellite loci were found to be highly variable with observed heterozygosity values ranging from 0.61 to 0.82. These values are similar to those observed for more distant populations from Norway, Finland and Germany. Therefore, the populations seem not particularly depauperate of genetic variation at the microsatellite level. Genetically, the Dutch populations were found to have diverged considerably. Pairwise comparisons of all populations studied revealed FST values significantly greater than zero for most comparisons. However, the magnitude of these values considerably depends on the compared population pair. The level of differentiation between local populations within Dutch regions is generally significantly lower than the differentiation between Dutch regions. The level of differentiation between Dutch regions, however, is not significantly different from that between populations of larger geographical distance. This implies that the regional Dutch populations are both isolated from each other and from other European populations. The observation that even local populations show low but significant genetic differentiation may be indicative for progressive isolation of these populations.     

Unpredictable selection in a structured population leads to local genetic differentiation in evolved reaction norms

Unpredictability during development of the optimum phenotype under future selection leads to a compromise reaction norm with a slope that is shallower than the slope of the optimum reaction norm. Unpredictability of selection can lead to an evolved curved reaction norm when genetic variation for curvature is available even if the optimum reaction norm is linear. This requires asymmetry in the frequency distribution of the habitats of selection; at small population size, stochasticity in the number of individuals per selection habitat is sufficient to generate such asymmetry. Unpredictability of selection in structured populations leads to local genetic differentiation of reaction norms. The mean habitat of a subpopulation is defined as the subpopulation's focal habitat. The evolved mean reaction norm of each subpopulation is anchored at the optimum genotypic value in its focal habitat. Linear reaction norms are parallel if the conditional distribution of adults around the focal habitats is the same for each subpopulation. Adult migration and absence of zygote dispersal represents the ultimate structured population, each habitat playing the role of focal habitat. Absence of zygote dispersal requires that the flow of individuals through the habitats is used instead of the habitats’ frequencies in the prediction of the evolved reaction norm. Adult migration in absence of zygote dispersal leads to an evolved pattern of locally differentiated reaction norms with optimum genotypic value anchored in the focal habitat and, for linear reaction norms, parallel slopes.     

Identifying the minimum number of microsatellite loci needed to assess population genetic structure: A case study in fly culturing

Small, isolated populations are constantly threatened by loss of genetic diversity due to drift. Such situations are found, for instance, in laboratory culturing. In guarding against diversity loss, monitoring of potential changes in population structure is paramount; this monitoring is most often achieved using microsatellite markers, which can be costly in terms of time and money when many loci are scored in large numbers of individuals. Here, we present a case study reducing the number of microsatellites to the minimum necessary to correctly detect the population structure of two Drosophila nigrosparsa populations. The number of loci was gradually reduced from 11 to 1, using the Allelic Richness (AR) and Private Allelic Richness (PAR) as criteria for locus removal. The effect of each reduction step was evaluated by the number of genetic clusters detectable from the data and by the allocation of individuals to the clusters; in the latter, excluding ambiguous individuals was tested to reduce the rate of incorrect assignments. We demonstrate that more than 95% of the individuals can still be correctly assigned when using eight loci and that the major population structure is still visible when using two highly polymorphic loci. The differences between sorting the loci by AR and PAR were negligible. The method presented here will most efficiently reduce genotyping costs when small sets of loci (“core sets”) for long-time use in large-scale population screenings are compiled.     

Molecular and quantitative genetic differentiation in European populations of Silene latifolia (Caryophyllaceae)

BACKGROUND AND AIMS: Among-population differentiation in phenotypic traits and allelic variation is expected as a consequence of isolation, drift, founder effects and local selection. Therefore, investigating molecular and quantitative genetic divergence is a pre-requisite for studies of local adaptation in response to selection under variable environmental conditions. METHODS: Among- and within-population variation were investigated in six geographically separated European populations of the white campion, Silene latifolia, both for molecular variation at six newly developed microsatellite loci and for quantitative variation in morphological and life-history traits. To avoid confounding effects of the maternal environment, phenotypic traits were measured on greenhouse-reared F(1) offspring. Tests were made for clinal variation, and the correlations among molecular, geographic and phenotypic distances were compared with Mantel tests. KEY RESULTS: The six populations of Silene latifolia investigated showed significant molecular and quantitative genetic differentiation. Geographic and phenotypic distances were significantly associated. Age at first flowering increased significantly with latitude and exhibited a Q(st) value of 0.17 in females and 0.10 in males, consistent with adaptation to local environmental conditions. By contrast, no evidence of isolation-by-distance and no significant association between molecular and phenotypic distances were found. CONCLUSIONS: Significant molecular genetic divergence among populations of Silene latifolia, from the European native range is consistent with known limited seed and pollen flow distances, while significant quantitative genetic divergence among populations and clinal variation for age at first flowering suggest local adaptation.     

Parental selection,number of breeding populations,and size of each population in inbred development

Some breeders select inbreds from many F₂ or backcross breeding populations, each with relatively few progenies. Other breeders select inbreds from only a few breeding populations, each with many progenies. My objectives were to: (1) determine the relative importance of parental selection, number of breeding populations, and size of each population, and (2) find optimum combinations between number and size of breeding populations. I assumed that a breeder has resources to test a total of 2,000 recombinant inbreds for a quantitative trait that was controlled by 100 additive loci and had a heritability of 0.20, 0.60, or 1.0. The parental inbreds had an inherent pedigree structure due to advanced cycle breeding. The parental inbreds were ranked according to their mean performance, and breeding populations were made among all parents, the top 25% of parents, and the top 10% of parents. I found that the issue of number versus size of breeding populations was only secondary compared with the ability to identify, prior to making the crosses, the breeding populations with the highest mean performance. For a given level of effectiveness of parental selection, the selection response was largest when the maximum number of breeding populations was used. The effect of the number of breeding populations was minor, however, when selection was practiced among the parents or when heritability was less than 1.0. The results suggested that, in practice, large selection responses could be obtained with a wide range of combinations between number and size of breeding populations.Communicated by H.C. Becker