Genome-wide Allele Frequency Research Articles

OptM: estimating the optimal number of migration edges on population trees using Treemix.

The software Treemix has become extensively used to estimate the number of migration events, or edges (m), on population trees from genome-wide allele frequency data. However, the appropriate number of edges to include remains unclear. Here, I show that an optimal value of m can be inferred from the second-order rate of change in likelihood (Δm) across incremental values of m. Repurposed from its original use to estimate the number of population clusters in the software Structure (ΔK), I show using simulated populations that Δm performs equally as well as current recommendations for Treemix. A demonstration of an empirical dataset from domestic dogs indicates that this method may be preferable in large, complex population histories and can prioritize migration events for subsequent investigation. The method has been implemented in a freely available R package called “OptM” and as a web application (https://rfitak.shinyapps.io/OptM/) to interface directly with the output files of Treemix.

Intercontinental genomic parallelism in multiple three-spined stickleback adaptive radiations.

Parallelism, the evolution of similar traits in populations diversifying in similar conditions, provides strong evidence of adaptation by natural selection. Many studies of parallelism focus on comparisons of different ecotypes or contrasting environments, defined a priori, which could upwardly bias the apparent prevalence of parallelism. Here, we estimated genomic parallelism associated with components of environmental and phenotypic variation at an intercontinental scale across four freshwater adaptive radiations (Alaska, British Columbia, Iceland, Scotland) of the three-spined stickleback (Gasterosteus aculeatus). We combined large-scale biological sampling and phenotyping with RAD-sequencing data from 73 freshwater lake populations and four marine ones (1,380 fish) to associate genome-wide allele frequencies with continuous distributions of environmental and phenotypic variation. Our three main findings demonstrate: 1) quantitative variation in phenotypes and environments can predict genomic parallelism; 2) genomic parallelism at the early stages of adaptive radiations, even at large geographic scales, is founded on standing variation; and 3) similar environments are a better predictor of genome-wide parallelism than similar phenotypes. Overall, this study validates the importance and predictive power of major phenotypic and environmental factors likely to influence the emergence of common patterns of genomic divergence, providing a clearer picture than analyses of dichotomous phenotypes and environments.

Speciation Associated with Shifts in Migratory Behavior in an Avian Radiation.

Migratory animals move up to thousands of kilometers every year [1]. Losses of migration (i.e., migratory drop-offs) occur when individuals of a migratory species stop migrating and establish founder sedentary populations, a phenomenon documented in birds [2-5] and butterflies [6]. In theory, losses-and also gains-of migration might promote speciation if sedentary and migratory populations become reproductively isolated [7-9]. Because migratory and sedentary strategies involve alternative physiological, behavioral, and morphological traits [10-13], divergence along multiple axes of organismal function is expected to accompany switches in migratory behavior, potentially accelerating speciation. We present evidence of speciation driven by a migratory drop-off in the fork-tailed flycatcher (Tyrannus savana) resulting in reproductive isolation likely driven by changes in breeding schedules (allochronic speciation [13-15]) and geographic isolation of breeding grounds (allopatric speciation [16]). Phylogenetic analyses across New World flycatchers (Tyrannidae) showed that an association between speciation and drop-offs is also observable at a macroevolutionary scale. Loss of migration was significantly more frequent than its gain, and speciation rates of migratory and partially migratory lineages (i.e., species having both migratory and sedentary populations) exceeded those of sedentary lineages. Models of trait evolution indicated that partial migration is an intermediate step between migratory and sedentary states in this family. Given that partial migration is widespread across migratory animals (e.g., of all migratory birds, ca. 51% are partially migratory [5]), speciation via switches in migratory behavior might be an important yet overlooked mechanism of animal diversification.

Genetic mapping of fitness determinants across the malaria parasitePlasmodium falciparum life cycle.

Determining the genetic basis of fitness is central to understanding evolution and transmission of microbial pathogens. In human malaria parasites (Plasmodium falciparum), most experimental work on fitness has focused on asexual blood stage parasites, because this stage can be easily cultured, although the transmission of malaria requires both female Anopheles mosquitoes and vertebrate hosts. We explore a powerful approach to identify the genetic determinants of parasite fitness across both invertebrate and vertebrate life-cycle stages of P. falciparum. This combines experimental genetic crosses using humanized mice, with selective whole genome amplification and pooled sequencing to determine genome-wide allele frequencies and identify genomic regions under selection across multiple lifecycle stages. We applied this approach to genetic crosses between artemisinin resistant (ART-R, kelch13-C580Y) and ART-sensitive (ART-S, kelch13-WT) parasites, recently isolated from Southeast Asian patients. Two striking results emerge: we observed (i) a strong genome-wide skew (>80%) towards alleles from the ART-R parent in the mosquito stage, that dropped to ~50% in the blood stage as selfed ART-R parasites were selected against; and (ii) repeatable allele specific skews in blood stage parasites with particularly strong selection (selection coefficient (s) ≤ 0.18/asexual cycle) against alleles from the ART-R parent at loci on chromosome 12 containing MRP2 and chromosome 14 containing ARPS10. This approach robustly identifies selected loci and has strong potential for identifying parasite genes that interact with the mosquito vector or compensatory loci involved in drug resistance.

PGG.Han: the Han Chinese genome database and analysis platform.

As the largest ethnic group in the world, the Han Chinese population is nonetheless underrepresented in global efforts to catalogue the genomic variability of natural populations. Here, we developed the PGG.Han, a population genome database to serve as the central repository for the genomic data of the Han Chinese Genome Initiative (Phase I). In its current version, the PGG.Han archives whole-genome sequences or high-density genome-wide single-nucleotide variants (SNVs) of 114 783 Han Chinese individuals (a.k.a. the Han100K), representing geographical sub-populations covering 33 of the 34 administrative divisions of China, as well as Singapore. The PGG.Han provides: (i) an interactive interface for visualization of the fine-scale genetic structure of the Han Chinese population; (ii) genome-wide allele frequencies of hierarchical sub-populations; (iii) ancestry inference for individual samples and controlling population stratification based on nested ancestry informative markers (AIMs) panels; (iv) population-structure-aware shared control data for genotype-phenotype association studies (e.g. GWASs) and (v) a Han-Chinese-specific reference panel for genotype imputation. Computational tools are implemented into the PGG.Han, and an online user-friendly interface is provided for data analysis and results visualization. The PGG.Han database is freely accessible via http://www.pgghan.org or https://www.hanchinesegenomes.org.

GBS-derived SNP catalogue unveiled wide genetic variability and geographical relationships of Italian olive cultivars

Information on the distribution of genetic variation is essential to preserve olive germplasm from erosion and to recover alleles lost through selective breeding. In addition, knowledge on population structure and genotype–phenotype associations is crucial to support modern olive breeding programs that must respond to new environmental conditions imposed by climate change and novel biotic/abiotic stressors. To further our understanding of genetic variation in the olive, we performed genotype-by-sequencing on a panel of 94 Italian olive cultivars. A reference-based and a reference-independent SNP calling pipeline generated 22,088 and 8,088 high-quality SNPs, respectively. Both datasets were used to model population structure via parametric and non parametric clustering. Although the two pipelines yielded a 3-fold difference in the number of SNPs, both described wide genetic variability among our study panel and allowed individuals to be grouped based on fruit weight and the geographical area of cultivation. Multidimensional scaling analysis on identity-by-state allele-sharing values as well as inference of population mixtures from genome-wide allele frequency data corroborated the clustering pattern we observed. These findings allowed us to formulate hypotheses about geographical relationships of Italian olive cultivars and to confirm known and uncover novel cases of synonymy.

Male Infertility Is Responsible for Nearly Half of the Extinction Observed in the Mouse Collaborative Cross.

The goal of the Collaborative Cross (CC) project was to generate and distribute over 1000 independent mouse recombinant inbred strains derived from eight inbred founders. With inbreeding nearly complete, we estimated the extinction rate among CC lines at a remarkable 95%, which is substantially higher than in the derivation of other mouse recombinant inbred populations. Here, we report genome-wide allele frequencies in 347 extinct CC lines. Contrary to expectations, autosomes had equal allelic contributions from the eight founders, but chromosome X had significantly lower allelic contributions from the two inbred founders with underrepresented subspecific origins (PWK/PhJ and CAST/EiJ). By comparing extinct CC lines to living CC strains, we conclude that a complex genetic architecture is driving extinction, and selection pressures are different on the autosomes and chromosome X Male infertility played a large role in extinction as 47% of extinct lines had males that were infertile. Males from extinct lines had high variability in reproductive organ size, low sperm counts, low sperm motility, and a high rate of vacuolization of seminiferous tubules. We performed QTL mapping and identified nine genomic regions associated with male fertility and reproductive phenotypes. Many of the allelic effects in the QTL were driven by the two founders with underrepresented subspecific origins, including a QTL on chromosome X for infertility that was driven by the PWK/PhJ haplotype. We also performed the first example of cross validation using complementary CC resources to verify the effect of sperm curvilinear velocity from the PWK/PhJ haplotype on chromosome 2 in an independent population across multiple generations. While selection typically constrains the examination of reproductive traits toward the more fertile alleles, the CC extinct lines provided a unique opportunity to study the genetic architecture of fertility in a widely genetically variable population. We hypothesize that incompatibilities between alleles with different subspecific origins is a key driver of infertility. These results help clarify the factors that drove strain extinction in the CC, reveal the genetic regions associated with poor fertility in the CC, and serve as a resource to further study mammalian infertility.

Abstract 3099: Longitudinal genomic and transcriptomic analysis of mantle cell lymphoma in a targeted combination trial of a selective CDK4/6 inhibitor

Abstract We used whole exome (WES) and whole transcriptome sequencing (WTS) to study progression of Mantle Cell Lymphoma (MCL) in the context of a clinical trial of targeted combination therapy of a CDK4 inhibitor PD0332991 (palbociclib) and a proteasome inhibitor bortezomib. Our longitudinal study design coupled with an integrative analysis approach enabled detailed molecular characterization of each subject's disease. This includes multiple time points for three therapy Responders (Rs) and two Non-Responders (NRs); a primary goal was to understand genetic reasons for differential response. To that end, we identified copy number variants (CNVs) and single nucleotide variants (SNVs) specific to NRs that are supported by both WES and WTS data and implicate pathways relevant to mechanisms of drug action. CNVs were identified using multiple orthogonal methods that use read depth and genome-wide allele frequencies. These revealed a 29Mb chr9 hemizygous deletion unique to NRs. This region includes 15 interferon genes of the type 1 interferon signaling pathway (enrichment: p<10-9), which is involved in B cell differentiation and proliferation, and T cell and natural killer cell activation. Interferon signaling is implicated in bortezomib function and loss of interferon genes may contribute to resistance in non-responders. Additional key genes in deleted regions are JAK2, RPS6, and CDKN2A, a CDK2 inhibitor, loss of which may antagonize palbociclib-induced G1 arrest NRs also shared a 24Mb chr3 duplication resulting in gain of PIK3CA and PIK3CB. Amplification of these genes is common in cancers, including MCL (classical analysis) and DLBCL, and often associated with poor prognosis. But, this is the only longitudinal study of PI3K CNV in a targeted therapy. In context of PD+bortezomib treatment, higher levels of PI3K may affect sensitivity to proteasome inhibition activated autophagy since bortezomib's inhibitory activity involves suppression of the PI3K/AKT/mTOR pathway. PIK3CA is amplified in a responder as well, suggesting additional determinants for clinical response. Other CNVs differentiating Rs and NRs are a 24Mb chr8 deletion found only in Rs and a 4Mb chr17 deletion found only in NRs, which includes p53 in a subject who also has a p53 mutation, resulting in 100% expression of the mutant allele. The described CNVs result in detectable shifts in expression of a large proportion of affected genes. Coupled with SNV data, where we identified 21 SNVs shared by NRs and absent in Rs, these results offer molecular clues to mechanisms of resistance in NRs, which we are currently pursuing with functional studies. Overall, we performed a thorough genomic and transcriptomic analysis of a longitudinal investigation of MCL patients during a targeted combination therapy phase I clinical trial and identified potential biomarkers of response spanning CNVs, SNVs, and differentially expressed genes. Citation Format: Priyanka Vijay, Pedro Blecua, Maurizio Di Liberto, David Chiron, Xiangao Huang, Olivier Elemento, Peter Martin, John P. Leonard, Christopher E. Mason, Selina Chen-Kiang. Longitudinal genomic and transcriptomic analysis of mantle cell lymphoma in a targeted combination trial of a selective CDK4/6 inhibitor. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 3099. doi:10.1158/1538-7445.AM2015-3099

Identifying signatures of sexual selection using genomewide selection components analysis.

Sexual selection must affect the genome for it to have an evolutionary impact, yet signatures of selection remain elusive. Here we use an individual-based model to investigate the utility of genome-wide selection components analysis, which compares allele frequencies of individuals at different life history stages within a single population to detect selection without requiring a priori knowledge of traits under selection. We modeled a diploid, sexually reproducing population and introduced strong mate choice on a quantitative trait to simulate sexual selection. Genome-wide allele frequencies in adults and offspring were compared using weighted FST values. The average number of outlier peaks (i.e., those with significantly large FST values) with a quantitative trait locus in close proximity (“real” peaks) represented correct diagnoses of loci under selection, whereas peaks above the FST significance threshold without a quantitative trait locus reflected spurious peaks. We found that, even with moderate sample sizes, signatures of strong sexual selection were detectable, but larger sample sizes improved detection rates. The model was better able to detect selection with more neutral markers, and when quantitative trait loci and neutral markers were distributed across multiple chromosomes. Although environmental variation decreased detection rates, the identification of real peaks nevertheless remained feasible. We also found that detection rates can be improved by sampling multiple populations experiencing similar selection regimes. In short, genome-wide selection components analysis is a challenging but feasible approach for the identification of regions of the genome under selection.

The impact of library preparation protocols on the consistency of allele frequency estimates in Pool-Seq data.

Sequencing pools of individuals (Pool‐Seq) is a cost‐effective method to determine genome‐wide allele frequency estimates. Given the importance of meta‐analyses combining data sets, we determined the influence of different genomic library preparation protocols on the consistency of allele frequency estimates. We found that typically no more than 1% of the variation in allele frequency estimates could be attributed to differences in library preparation. Also read length had only a minor effect on the consistency of allele frequency estimates. By far, the most pronounced influence could be attributed to sequence coverage. Increasing the coverage from 30‐ to 50‐fold improved the consistency of allele frequency estimates by at least 27%. We conclude that Pool‐Seq data can be easily combined across different library preparation methods, but sufficient sequence coverage is key to reliable results.

A new method for studying population genetics of cyst nematodes based on Pool-Seq and genomewide allele frequency analysis.

Cyst nematodes are important agricultural pests responsible for billions of dollars of losses each year. Plant resistance is the most effective management tool, but it requires a close monitoring of population genetics. Current technologies for pathotyping and genotyping cyst nematodes are time-consuming, expensive and imprecise. In this study, we capitalized on the reproduction mode of cyst nematodes to develop a simple population genetic analysis pipeline based on genotyping-by-sequencing and Pool-Seq. This method yielded thousands of SNPs and allowed us to study the relationships between populations of different origins or pathotypes. Validation of the method on well-characterized populations also demonstrated that it was a powerful and accurate tool for population genetics. The genomewide allele frequencies of 23 populations of golden nematode, from nine countries and representing the five known pathotypes, were compared. A clear separation of the pathotypes and fine genetic relationships between and among global populations were obtained using this method. In addition to being powerful, this tool has proven to be very time- and cost-efficient and could be applied to other cyst nematode species.

Divergent, coexisting Pseudomonas aeruginosa lineages in chronic cystic fibrosis lung infections.

Pseudomonas aeruginosa, the predominant cause of chronic airway infections of patients with cystic fibrosis, exhibits extensive phenotypic diversity among isolates within and between sputum samples, but little is known about the underlying genetic diversity. To characterize the population genetic structure of transmissible P. aeruginosa Liverpool Epidemic Strain in chronic infections of nine patients with cystic fibrosis, and infer evolutionary processes associated with adaptation to the cystic fibrosis lung. We performed whole-genome sequencing of P. aeruginosa isolates and pooled populations and used comparative analyses of genome sequences including phylogenetic reconstructions and resolution of population structure from genome-wide allele frequencies. Genome sequences were obtained for 360 isolates from nine patients. Phylogenetic reconstruction of the ancestry of 40 individually sequenced isolates from one patient sputum sample revealed the coexistence of two genetically diverged, recombining lineages exchanging potentially adaptive mutations. Analysis of population samples for eight additional patients indicated coexisting lineages in six cases. Reconstruction of the ancestry of individually sequenced isolates from all patients indicated smaller genetic distances between than within patients in most cases. Our population-level analysis demonstrates that coexistence of distinct lineages of P. aeruginosa Liverpool Epidemic Strain within individuals is common. In several cases, coexisting lineages may have been present in the infecting inoculum or assembled through multiple transmissions. Divergent lineages can share mutations via homologous recombination, potentially aiding adaptation to the airway during chronic infection. The genetic diversity of this transmissible strain within infections, revealed by high-resolution genomics, has implications for patient segregation and therapeutic strategies.

SHOREmap v3.0: fast and accurate identification of causal mutations from forward genetic screens.

Whole-genome resequencing of pools of recombinant mutant genomes allows direct linking of phenotypic traits to causal mutations. Such analysis, called mapping-by-sequencing, combines classical genetic mapping and next-generation sequencing by relying on selection-induced patterns within genome-wide allele frequency (AF) in pooled genomes. Mapping-by-sequencing can be performed with computational tools such as SHOREmap. Previous versions of SHOREmap, however, did not implement standardized analyses, but were specifically designed for particular experimental settings. Here, we introduce the usage of a novel and advanced implementation of SHOREmap (version 3.0), including several new features like file readers for commonly used file formats, SNP marker selection, and a stable calculation of mapping intervals. SHOREmap can be downloaded at shoremap.org.

Experimental evidence for ecological selection on genome variation in the wild

Understanding natural selection's effect on genetic variation is a major goal in biology, but the genome-scale consequences of contemporary selection are not well known. In a release and recapture field experiment we transplanted stick insects to native and novel host plants and directly measured allele frequency changes within a generation at 186 576 genetic loci. We observed substantial, genome-wide allele frequency changes during the experiment, most of which could be attributed to random mortality (genetic drift). However, we also documented that selection affected multiple genetic loci distributed across the genome, particularly in transplants to the novel host. Host-associated selection affecting the genome acted on both a known colour-pattern trait as well as other (unmeasured) phenotypes. We also found evidence that selection associated with elevation affected genome variation, although our experiment was not designed to test this. Our results illustrate how genomic data can identify previously underappreciated ecological sources and phenotypic targets of selection.

Dynamics of Adaptive Alleles in Divergently Selected Body Weight Lines of Chickens

By studying genomic changes over time in populations subjected to strong artificial directional selection, we can gain insights to the dynamics of beneficial alleles originating from the founder population or emerging as novel mutations undergoing ongoing selection. The Virginia lines are a chicken resource population generated by long-term bi-directional, single-trait selection for juvenile body weight. We studied genome-wide allele frequency changes from generation 40 to 53 using genome-wide genotypes from directional and relaxed selection lines. Overall, there were small changes in allele frequencies at individual loci over the studied time period; but, on average, the changes were greater in lines with larger phenotypic changes. This is consistent with previous findings that much of the response to selection over the first 40 years of selection was attributable to utilization of standing genetic variation at many loci in the genome, indicating a mostly polygenic architecture for body weight. Over the course of the selection experiment, the largest phenotypic response to selection was observed in the high-weight selected line, and in this line we detected a single locus where the allele frequency changed rapidly during a late stage of the experiment. This locus likely contains a novel, beneficial mutation that appeared between generations 40 and 45 and was driven to fixation in 5 to 10 generations. This result illustrates the dependence of continued long-term selection response on standing genetic variation at many loci as well as strong, novel, beneficial mutations.

Inferring Population Histories Using Genome-Wide Allele Frequency Data

The recent development of high-throughput genotyping technologies has revolutionized the collection of data in a wide range of both model and nonmodel species. These data generally contain huge amounts of information about the demographic history of populations. In this study, we introduce a new method to estimate divergence times on a diffusion time scale from large single-nucleotide polymorphism (SNP) data sets, conditionally on a population history that is represented as a tree. We further assume that all the observed polymorphisms originate from the most ancestral (root) population; that is, we neglect mutations that occur after the split of the most ancestral population. This method relies on a hierarchical Bayesian model, based on Kimura's time-dependent diffusion approximation of genetic drift. We implemented a Metropolis-Hastings within Gibbs sampler to estimate the posterior distribution of the parameters of interest in this model, which we refer to as the Kimura model. Evaluating the Kimura model on simulated population histories, we found that it provides accurate estimates of divergence time. Assessing model fit using the deviance information criterion (DIC) proved efficient for retrieving the correct tree topology among a set of competing histories. We show that this procedure is robust to low-to-moderate gene flow, as well as to ascertainment bias, providing that the most distantly related populations are represented in the discovery panel. As an illustrative example, we finally analyzed published human data consisting in genotypes for 452,198 SNPs from individuals belonging to four populations worldwide. Our results suggest that the Kimura model may be helpful to characterize the demographic history of differentiated populations, using genome-wide allele frequency data.

Inference of population splits and mixtures from genome-wide allele frequency data.

Many aspects of the historical relationships between populations in a species are reflected in genetic data. Inferring these relationships from genetic data, however, remains a challenging task. In this paper, we present a statistical model for inferring the patterns of population splits and mixtures in multiple populations. In our model, the sampled populations in a species are related to their common ancestor through a graph of ancestral populations. Using genome-wide allele frequency data and a Gaussian approximation to genetic drift, we infer the structure of this graph. We applied this method to a set of 55 human populations and a set of 82 dog breeds and wild canids. In both species, we show that a simple bifurcating tree does not fully describe the data; in contrast, we infer many migration events. While some of the migration events that we find have been detected previously, many have not. For example, in the human data, we infer that Cambodians trace approximately 16% of their ancestry to a population ancestral to other extant East Asian populations. In the dog data, we infer that both the boxer and basenji trace a considerable fraction of their ancestry (9% and 25%, respectively) to wolves subsequent to domestication and that East Asian toy breeds (the Shih Tzu and the Pekingese) result from admixture between modern toy breeds and “ancient” Asian breeds. Software implementing the model described here, called TreeMix, is available at http://treemix.googlecode.com.

What paths do advantageous alleles take during short-term evolutionary change?

Combining experimental evolution with whole-genome resequencing is a promising new strategy for investigating the dynamics of evolutionary change. Published studies that have resequenced laboratory-selected populations of sexual organisms have typically focused on populations sampled at the end of an evolution experiment. These studies have attempted to associate particular alleles with phenotypic change and attempted to distinguish between different theoretical models of adaptation. However, neither the population used to initiate the experiment nor multiple time points sampled during the evolutionary trajectory are generally available for examination. In this issue of Molecular Ecology, Orozco-terWengel et al. (2012) take a significant step forward by estimating genome-wide allele frequencies at the start, 15 generations into and at the end of a 37-generation Drosophila experimental evolution study. The authors identify regions of the genome that have responded to laboratory selection and describe the temporal dynamics of allele frequency change. They identify two common trajectories for putatively adaptive alleles: alleles either gradually increase in frequency throughout the entire 37 generations or alleles plateau at a new frequency by generation 15. The identification of complex trajectories of alleles under selection contributes to a growing body of literature suggesting that simple models of adaptation, whereby beneficial alleles arise and increase in frequency unimpeded until they become fixed, may not adequately describe short-term response to selection.

Empirical Validation of Pooled Whole Genome Population Re-Sequencing in Drosophila melanogaster

The sequencing of pooled non-barcoded individuals is an inexpensive and efficient means of assessing genome-wide population allele frequencies, yet its accuracy has not been thoroughly tested. We assessed the accuracy of this approach on whole, complex eukaryotic genomes by resequencing pools of largely isogenic, individually sequenced Drosophila melanogaster strains. We called SNPs in the pooled data and estimated false positive and false negative rates using the SNPs called in individual strain as a reference. We also estimated allele frequency of the SNPs using “pooled” data and compared them with “true” frequencies taken from the estimates in the individual strains. We demonstrate that pooled sequencing provides a faithful estimate of population allele frequency with the error well approximated by binomial sampling, and is a reliable means of novel SNP discovery with low false positive rates. However, a sufficient number of strains should be used in the pooling because variation in the amount of DNA derived from individual strains is a substantial source of noise when the number of pooled strains is low. Our results and analysis confirm that pooled sequencing is a very powerful and cost-effective technique for assessing of patterns of sequence variation in populations on genome-wide scales, and is applicable to any dataset where sequencing individuals or individual cells is impossible, difficult, time consuming, or expensive.

Inference of population splits and mixtures from genome-wide allele frequency data

AbstractMany aspects of the historical relationships between populations in a species are reflected in genetic data. Inferring these relationships from genetic data, however, remains a challenging task. In this paper, we present a statistical model for inferring the patterns of population splits and mixtures in multiple populations. In this model, the sampled populations in a species are related to their common ancestor through a graph of ancestral populations. Using genome-wide allele frequency data and a Gaussian approximation to genetic drift, we infer the structure of this graph. We applied this method to a set of 55 human populations and a set of 82 dog breeds and wild canids. In both species, we show that a simple bifurcating tree does not fully describe the data; in contrast, we infer many migration events. While some of the migration events that we find have been detected previously, many have not. For example, in the human data we infer that Cambodians trace approximately 16% of their ancestry to a population ancestral to other extant East Asian populations. In the dog data, we infer that both the boxer and basenji trace a considerable fraction of their ancestry (9% and 25%, respectively) to wolves subsequent to domestication, and that East Asian toy breeds (the Shih Tzu and the Pekingese) result from admixture between modern toy breeds and ``ancient" Asian breeds. Software implementing the model described here, called TreeMix, is available at "http://treemix.googlecode.com":http://treemix.googlecode.com.