Search

Accurately estimating biological variables of interest, such as parameters of demographic models, is a key problem in evolutionary genetics. Likelihood-based and likelihood-free methods both typically use only limited genetic information, such as carefully chosen summary statistics. Deep convolutional neural networks (CNNs) trained on genotype matrices can incorporate a great deal more information, and have been shown to have high accuracy for inferring parameters such as recombination rates and population sizes, when evaluated using simulations. However these methods are typically framed as regression or classification problems, and it is not straightforward to ensure that the training data adequately model the empirical data on which they are subsequently applied. It has recently been shown that generative adversarial networks (GANs) can be used to iteratively tune parameter values until simulations match a given target dataset. Here, we investigate an adversarial architecture for discriminator-based inference, which iteratively improves the sampling distribution for training the discriminator CNN via Monte Carlo density estimation. We show that this method produces parameter estimates with excellent agreement to simulated data. We developeddinf, a modular Python package for discriminator-based inference that incorporates this method, and is available fromhttps://github.com/RacimoLab/dinf/.

The role of natural selection in shaping biological diversity is an area of intense interest in modern biology. To date, studies of positive selection have primarily relied on genomic datasets from contemporary populations, which are susceptible to confounding factors associated with complex and often unknown aspects of population history. In particular, admixture between diverged populations can distort or hide prior selection events in modern genomes, though this process is not explicitly accounted for in most selection studies despite its apparent ubiquity in humans and other species. Through analyses of ancient and modern human genomes, we show that previously reported Holocene-era admixture has masked more than 50 historic hard sweeps in modern European genomes. Our results imply that this canonical mode of selection has probably been underappreciated in the evolutionary history of humans and suggest that our current understanding of the tempo and mode of selection in natural populations may be inaccurate. Ancient human genomes uncover historical hard sweeps Hard sweeps were common in ancient West Eurasian populations A historical hard sweep in the MHC-III region Admixture can obscure historical hard sweeps Evolutionary scenarios underlying sweep signal dilution The mutational basis of the Eurasian hard sweeps Discussion Methods

Understanding the demographic history of populations is a key goal in population genetics, and with improving methods and data, ever more complex models are being proposed and tested. Demographic models of current interest typically consist of a set of discrete populations, their sizes and growth rates, and continuous and pulse migrations between those populations over a number of epochs, which can require dozens of parameters to fully describe. There is currently no standard format to define such models, significantly hampering progress in the field. In particular, the important task of translating the model descriptions in published work into input suitable for population genetic simulators is labor intensive and error prone. We propose the Demes data model and file format, built on widely used technologies, to alleviate these issues. Demes provides a well-defined and unambiguous model of populations and their properties that is straightforward to implement in software, and a text file format that is designed for simplicity and clarity. We provide thoroughly tested implementations of Demes parsers in multiple languages including Python and C, and showcase initial support in several simulators and inference methods. An introduction to the file format and a detailed specification are available at:https://popsim-consortium.github.io/demes-spec-docs/.

Tropical islands are renowned as natural laboratories for evolutionary study. Lineage radiations across tropical archipelagos are ideal systems for investigating how colonization, speciation, and extinction processes shape biodiversity patterns. The expansion of the island thrush across the Indo-Pacific represents one of the largest yet most perplexing island radiations of any songbird species. The island thrush exhibits a complex mosaic of pronounced plumage variation across its range, and is arguably the world’s most polytypic bird. It is a sedentary species largely restricted to mountain forests, yet it has colonized a vast island region spanning a quarter of the globe. We conducted comprehensive sampling of island thrush populations and obtained genome-wide SNP data, which we used to reconstruct its phylogeny, population structure, gene flow, and demographic history. The island thrush evolved from migratory Palearctic ancestors and radiated explosively across the Indo-Pacific during the Pleistocene, with numerous instances of gene flow between populations. Its bewildering plumage variation masks a biogeographically intuitive stepping stone colonization path from the Philippines through the Greater Sundas, Wallacea and New Guinea to Polynesia. The island thrush’s success in colonizing Indo-Pacific mountains can be understood in light of its ancestral mobility and adaptation to cool climates; however, shifts in elevational range, degree of plumage variation and apparent dispersal rates in the eastern part of its range raise further intriguing questions about its biology.

Stochastic simulation is a key tool in population genetics, since the models involved are often analytically intractable and simulation is usually the only way of obtaining ground-truth data to evaluate inferences. Because of this, a large number of specialized simulation programs have been developed, each filling a particular niche, but with largely overlapping functionality and a substantial duplication of effort. Here, we introduce msprime version 1.0, which efficiently implements ancestry and mutation simulations based on the succinct tree sequence data structure and the tskit library. We summarize msprime’s many features, and show that its performance is excellent, often many times faster and more memory efficient than specialized alternatives. These high-performance features have been thoroughly tested and validated, and built using a collaborative, open source development model, which reduces duplication of effort and promotes software quality via community engagement.

Stochastic simulation is a key tool in population genetics, since the models involved are often analytically intractable and simulation is usually the only way of obtaining ground-truth data to evaluate inferences. Because of this necessity, a large number of specialised simulation programs have been developed, each filling a particular niche, but with largely overlapping functionality and a substantial duplication of effort. Here, we introducemsprimeversion 1.0, which efficiently implements ancestry and mutation simulations based on the succinct tree sequence data structure andtskitlibrary. We summarisemsprime’s many features, and show that its performance is excellent, often many times faster and more memory efficient than specialised alternatives. These high-performance features have been thoroughly tested and validated, and built using a collaborative, open source development model, which reduces duplication of effort and promotes software quality via community engagement.

A recent study of mammoth subfossil remains has demonstrated the potential of using relatively low-coverage high-throughput DNA sequencing to genetically sex specimens, revealing a strong male-biased sex ratio [P. Pečnerová et al., Curr. Biol. 27, 3505–3510.e3 (2017)]. Similar patterns were predicted for steppe bison, based on their analogous female herd-based structure. We genetically sexed subfossil remains of 186 Holarctic bison (Bison spp.), and also 91 brown bears (Ursus arctos), which are not female herd-based, and found that ∼75% of both groups were male, very close to the ratio observed in mammoths (72%). This large deviation from a 1:1 ratio was unexpected, but we found no evidence for sex differences with respect to DNA preservation, sample age, material type, or overall spatial distribution. We further examined ratios of male and female specimens from 4 large museum mammal collections and found a strong male bias, observable in almost all mammalian orders. We suggest that, in mammals at least, 1) wider male geographic ranges can lead to considerably increased chances of detection in fossil studies, and 2) sexual dimorphic behavior or appearance can facilitate a considerable sex bias in fossil and modern collections, on a previously unacknowledged scale. This finding has major implications for a wide range of studies of fossil and museum material.

SummaryTwo genera and multiple species of short-faced bear from the Americas went extinct during or toward the end of the Pleistocene, and all belonged to the endemic New World subfamily Tremarctinae [1-7]. Two of these species were giants, growing in excess of 1,000 kg [6, 8, 9], but it remains uncertain how these extinct bears were related to the sole surviving short-faced bear: the spectacled bear (Tremarctos ornatus). Ancient mitochondrial DNA has recently suggested phylogenetic relationships among these lineages that conflict with interpretations based on morphology [1, 10-12]. However, widespread hybridisation and incomplete lineage sorting among extant bears mean that the mitochondrial phylogeny frequently does not reflect the true species tree [13, 14]. Here we present ancient nuclear genome sequences from representatives of the two extinct short-faced bear genera, Arctotherium and Arctodus. Our new data support a third hypothesis for the relationships among short-faced bears, which conflicts with existing mitochondrial and morphological data. Based on genome-wide D-statistics, we suggest that the extant spectacled bear derives substantial ancestry from Pleistocene hybridisation with an extinct short-faced bear lineage, resulting in a discordant phylogenetic signal between the mitochondrion and portions of the nuclear genome.

Two genera and multiple species of short-faced bear from the Americas went extinct during or toward the end of the Pleistocene, and all belonged to the endemic New World subfamily Tremarctinae. Two of these species were giants, growing in excess of 1,000 kg, but it remains uncertain how these extinct bears were related to the sole surviving short-faced bear: the spectacled bear (Tremarctos ornatus). Ancient mitochondrial DNA has recently suggested phylogenetic relationships among these lineages that conflict with interpretations based on morphology. However, widespread hybridisation and incomplete lineage sorting among extant bears mean that the mitochondrial phylogeny frequently does not reflect the true species tree. Here we present ancient nuclear genome sequences from representatives of the two extinct short-faced bear genera, Arctotherium and Arctodus. Our new data support a third hypothesis for the relationships among short-faced bears, which conflicts with existing mitochondrial and morphological data. Based on genome-wide D-statistics, we suggest that the extant spectacled bear derives substantial ancestry from Pleistocene hybridisation with an extinct short-faced bear lineage, resulting in a discordant phylogenetic signal between the mitochondrion and portions of the nuclear genome.

[Background] The evolutionary relationships of Felidae during their Early-Middle Miocene radiation is contentious. Although the early common ancestors have been subsumed under the grade-group Pseudaelurus, this group is thought to be paraphyletic, including the early ancestors of both modern cats and extinct sabretooths. [Methods] Here, we sequenced a draft nuclear genome of Smilodon populator, dated to 13,182 ± 90 cal BP, making this the oldest palaeogenome from South America to date, a region known to be problematic for ancient DNA preservation. We analysed this genome, together with genomes from other extinct and extant cats to investigate their phylogenetic relationships. [Results] We confirm a deep divergence (~20.65 Ma) within sabre-toothed cats. Through the analysis of both simulated and empirical data, we show a lack of gene flow between Smilodon and contemporary Felidae. [Conclusions] Given that some species traditionally assigned to Pseudaelurus originated in the Early Miocene ~20 Ma, this indicates that some species of Pseudaelurus may be younger than the lineages they purportedly gave rise to, further supporting the hypothesis that Pseudaelurus was paraphyletic., This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No [681396]), (project Extinction Genomics). This project has also received funding from the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no. FP7-PEOPLE-2011-IEF-298820. The work was also supported by the Independent Research Fund Denmark | Natural Sciences, Forskningsprojekt 1, grant no. 8021-00218B to EDL.

Studies in a variety of species have shown evidence for positively selected variants introduced into one population via introgression from another, distantly related population—a process known as adaptive introgression. However, there are few explicit frameworks for jointly modelling introgression and positive selection, in order to detect these variants using genomic sequence data. Here, we develop an approach based on convolutional neural networks (CNNs). CNNs do not require the specification of an analytical model of allele frequency dynamics, and have outperformed alternative methods for classification and parameter estimation tasks in various areas of population genetics. Thus, they are potentially well suited to the identification of adaptive introgression. Using simulations, we trained CNNs on genotype matrices derived from genomes sampled from the donor population, the recipient population and a related non-introgressed population, in order to distinguish regions of the genome evolving under adaptive introgression from those evolving neutrally or experiencing selective sweeps. Our CNN architecture exhibits 95% accuracy on simulated data, even when the genomes are unphased, and accuracy decreases only moderately in the presence of heterosis. As a proof of concept, we applied our trained CNNs to human genomic datasets—both phased and unphased—to detect candidates for adaptive introgression that shaped our evolutionary history.

Dire wolves are considered to be one of the most common and widespread large carnivores in Pleistocene America1, yet relatively little is known about their evolution or extinction. Here, to reconstruct the evolutionary history of dire wolves, we sequenced five genomes from sub-fossil remains dating from 13,000 to more than 50,000 years ago. Our results indicate that although they were similar morphologically to the extant grey wolf, dire wolves were a highly divergent lineage that split from living canids around 5.7 million years ago. In contrast to numerous examples of hybridization across Canidae2,3, there is no evidence for gene flow between dire wolves and either North American grey wolves or coyotes. This suggests that dire wolves evolved in isolation from the Pleistocene ancestors of these species. Our results also support an early New World origin of dire wolves, while the ancestors of grey wolves, coyotes and dholes evolved in Eurasia and colonized North America only relatively recently. © 2021, The Author(s), under exclusive licence to Springer Nature Limited. Acknowledgements We thank the staff at the Carnegie Museum of Natural History, Cincinnati Museum Center, Danish Zoological Museum, Harrison Zoological Museum, Harvard Museum of Comparative Zoology, Idaho Museum of Natural History, Institute of Archaeology (Russian Academy of Sciences), Institute of Systematics and Animal Ecology (Russian Academy of Sciences), Institute of Zoology (Chinese Academy of Sciences), Instituto de Conservação da Natureza e das Florestas, Kansas Museum of Natural History, La Brea Tar Pits and Museum, Ludwig Maximilian University, McClung Museum, Museum of the Institute of Plant and Animal Ecology (Russian Academy of Sciences), Museum national d’Histoire naturelle, National Museums Scotland, Natural History Museum London, Naturalis Biodiversity Center, Naturhistorisches Museum Bern, Smithsonian National Museum of Natural History, Swedish Naturhistoriska Riksmuseet, SYLVATROP, US Bureau of Reclamation, University of California Museum of Paleontology, University of Texas at El Paso, University of Washington Burke Museum and the Zoological Institute (Russian Academy of Sciences; state assignment no. АААА-А19-119032590102-7) for access to specimens in their care; T. Barnosky, S. Bray, A. Farrell, R. Fischer, A. Harris, J. Harris, A. Henrici, P. Holroyd, R. MacPhee, T. Martin, A. Philpot, J. Saunders, J. Southon, G. Storrs, G. Takeuchi, X. Wang and C. Widga for assistance; and L. DeSantis for comments. A.M. used computational and storage services associated with the Hoffman2 Shared Cluster provided by UCLA Institute for Digital Research and Education’s Research Technology Group. DireGWC was sequenced using the Vincent J. Coates Genomics Sequencing Laboratory at UC Berkeley, supported by NIH S10 OD018174 Instrumentation Grant. We acknowledge the assistance of the Danish National High-Throughput Sequencing Centre, BGI-Europe, the Garvan Institute of Medical Research and the Australian Cancer Research Foundation (ACRF) Cancer Genomics Facility for assistance in Illumina and BGIseq500 data generation. A.R.P. was supported by a Marie Curie COFUND Junior Research Fellowship (Durham University). A.M. was supported by an NSF grant (award number: 1457106) and the QCB Collaboratory Postdoctoral Fellowship (UCLA). L.A.F.F., J.H., A.H.-B. and G.L. were supported by either European Research Council grant (ERC-2013-StG-337574-UNDEAD and ERC-2019-StG-853272-PALAEOFARM) and/or Natural Environmental Research Council grants (NE/K005243/1 and NE/K003259/1). K.S. was supported by a grant from Barrett, the Honors College at Arizona State University. A.T.O. was supported by the Strategic Initiative Funds, Office of the President, Arizona State University to the Institute of Human Origins DNA and Human Origins at Arizona State University project. L.A.F.F. was supported by a Junior Research Fellowship (Wolfson College, University of Oxford) and L.A.F.F. and A. Carmagnini were supported by the Wellcome Trust (210119/Z/18/Z). S.G. was supported by Carlsbergfondet grant CF14–0995 and Marie Skłodowska-Curie Actions grant 655732-WhereWolf. M.T.P.G. was supported by ERC Consolidator grant 681396-Extinction Genomics. B.S. and J.K. were supported by IMLS MG-30-17-0045-17 and NSF DEB-1754451. A.H.-B. was supported by the Leverhulme Trust (ECF-2017-315). A. Cooper, K.J.M. and H.H. were supported by the Australian Research Council. A.T.S. and G.G. were supported by Australian Government Research Training Program Scholarships. A.T.L. was supported by the Peter Buck Postdoctoral Fellowship from the Smithsonian Institution’s National Museum of Natural History. Y.V.K. was supported by the by State Assignment of the Sobolev Institute of Geology and Mineralogy.

The Tasmanian tiger or thylacine (Thylacinus cynocephalus) was the largest carnivorous Australian marsupial to survive into the modern era. Despite last sharing a common ancestor with the eutherian canids ~160 million years ago, their phenotypic resemblance is considered the most striking example of convergent evolution in mammals. The last known thylacine died in captivity in 1936 and many aspects of the evolutionary history of this unique marsupial apex predator remain unknown. Here we have sequenced the genome of a preserved thylacine pouch young specimen to clarify the phylogenetic position of the thylacine within the carnivorous marsupials, reconstruct its historical demography and examine the genetic basis of its convergence with canids. Retroposon insertion patterns placed the thylacine as the basal lineage in Dasyuromorphia and suggest incomplete lineage sorting in early dasyuromorphs. Demographic analysis indicated a long-term decline in genetic diversity starting well before the arrival of humans in Australia. In spite of their extraordinary phenotypic convergence, comparative genomic analyses demonstrated that amino acid homoplasies between the thylacine and canids are largely consistent with neutral evolution. Furthermore, the genes and pathways targeted by positive selection differ markedly between these species. Together, these findings support models of adaptive convergence driven primarily by cis-regulatory evolution.

The explosion in population genomic data demands ever more complex modes of analysis, and increasingly these analyses depend on sophisticated simulations. Recent advances in population genetic simulation have made it possible to simulate large and complex models, but specifying such models for a particular simulation engine remains a difficult and error-prone task. Computational genetics researchers currently re-implement simulation models independently, leading to inconsistency and duplication of effort. This situation presents a major barrier to empirical researchers seeking to use simulations for power analyses of upcoming studies or sanity checks on existing genomic data. Population genetics, as a field, also lacks standard benchmarks by which new tools for inference might be measured. Here we describe a new resource,stdpopsim, that attempts to rectify this situation.Stdpopsimis a community-driven open source project, which provides easy access to a growing catalog of published simulation models from a range of organisms and supports multiple simulation engine backends. This resource is available as a well-documented python library with a simple command-line interface. We share some examples demonstrating howstdpopsimcan be used to systematically compare demographic inference methods, and we encourage a broader community of developers to contribute to this growing resource.

The Pairwise Sequentially Markov Coalescent (PSMC), and its extension PSMC′, model past population sizes from a single diploid genome. Both models have been widely applied, even to organisms with scaffold-level genome reference assemblies of limited contiguity. However it is unclear how PSMC and PSMC′ perform on short scaffolds. We evaluated psmc and msmc, implementations of the PSMC and PSMC′ models respectively, on simulated genomes with low contiguity, and compared results to those from fully contiguous data. Simulations with scaffolds from 100 Mb to 10 kb revealed that psmc maintains high accuracy down to lengths of 100 kb, while msmc is accurate down to 1 Mb. The discrepancy is not due to differing models, but stems from an implementation detail of msmc—homozygous tracts at the ends of scaffolds are discarded, making msmc unreliable for low contiguity genomes. We recommend excluding data that are aligned to shorter scaffolds when undertaking demographic inference.