Reconstructing the spatiotemporal patterns of admixture during the European Holocene using a novel genomic dating method
Abstract
Recent studies have shown that gene flow or admixture has been pervasive throughout human history. While several methods exist for dating admixture in contemporary populations, they are not suitable for sparse, low coverage data available from ancient specimens. To overcome this limitation, we developed DATES that leverages ancestry covariance patterns across the genome of a single individual to infer the timing of admixture. By performing simulations, we show that DATES provides reliable results under a range of demographic scenarios and outperforms available methods for ancient DNA applications. We apply DATES to ~1,100 ancient genomes to reconstruct gene flow events during the European Holocene. Present-day Europeans derive ancestry from three distinct groups, local Mesolithic hunter-gatherers, Anatolian farmers, and Yamnaya Steppe pastoralists. These ancestral groups were themselves admixed. By studying the formation of Anatolian farmers, we infer that the gene flow related to Iranian Neolithic farmers occurred no later than 9,600 BCE, predating agriculture in Anatolia. We estimate the early Steppe pastoralist groups genetically formed more than a millennium before the start of steppe pastoralism, providing new insights about the history of proto-Yamnaya cultures and the origin of Indo-European languages. Using ancient genomes across sixteen regions in Europe, we provide a detailed chronology of the Neolithization across Europe that occurred from ~6,400–4,300 BCE. This movement was followed by a rapid spread of steppe ancestry from ~3,200–2,500 BCE. Our analyses highlight the power of genomic dating methods to elucidate the legacy of human migrations, providing insights complementary to archaeological and linguistic evidence.
Significance
The European continent was subject to two major migrations during the Holocene: the movement of Near Eastern farmers during the Neolithic and the migration of Steppe pastoralists during the Bronze Age. To understand the timing and dynamics of these movements, we developed DATES that leverages ancestry covariance patterns across the genome of a single individual to infer the timing of admixture. Using ~1,100 ancient genomes spanning ~8,000–350 BCE, we reconstruct the chronology of the formation of the ancestral populations and the fine-scale details of the spread of Neolithic farming and Steppe pastoralist-related ancestry to Europe. Our analysis demonstrates the power of genomic dating methods to provide an independent and complementary timeline of population origins and movements using genetic data.
Article activity feed
-
Evaluation Summary:
This manuscript presents DATES, a method to infer the timing of admixture events using genetic data from present-day or ancient individuals. This is a robust method that is useful in the field of paleogenomics and outperforms existing methods. In this manuscript, DATES is applied to >1000 ancient human genomes to characterize major admixture events during the European Holocene. This work will be of interest to scholars in the fields of population genetics, paleogenomics, archeology, biological anthropology, and history.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
Chintalapati et al. present DATES, a method that leverages ancestry covariance patterns across the genome of a single individual to infer the timing of admixture events. Authors perform simulations of demographic models that mimic key events in European prehistory, as well as the characteristics of ancient DNA (aDNA) datasets (e.g., pseudo-haploid calls and a large proportion of missing data). Based on these simulations, the authors assess the performance of DATES across a range of admixture proportions, timeframes, the proportion of missing data, and different numbers of populations as surrogate parental groups. One of the major strengths of this manuscript is that it presents a very robust method that is broadly useful in paleogenomic studies and that fills an important gap in the field (previous LD- and …
Reviewer #1 (Public Review):
Chintalapati et al. present DATES, a method that leverages ancestry covariance patterns across the genome of a single individual to infer the timing of admixture events. Authors perform simulations of demographic models that mimic key events in European prehistory, as well as the characteristics of ancient DNA (aDNA) datasets (e.g., pseudo-haploid calls and a large proportion of missing data). Based on these simulations, the authors assess the performance of DATES across a range of admixture proportions, timeframes, the proportion of missing data, and different numbers of populations as surrogate parental groups. One of the major strengths of this manuscript is that it presents a very robust method that is broadly useful in paleogenomic studies and that fills an important gap in the field (previous LD- and haplotype-based methods do not perform as well with aDNA datasets).
The method is applied to >1000 ancient human genomes (publicly available) to characterize major admixture events during the European Holocene. The manuscript addresses long-standing questions in the field, for instance, the genetic formation of Anatolian farmers and Steppe pastoralists, the chronology of the neolithization of Europe, and the spread of Steppe ancestry across Europe, among others. The authors do a very good job in synthesizing known and new observations into a comprehensive and exciting story - this is particularly impressive given the breadth of events that this work covers.
The authors' claims are fully justified by their results. The manuscript is very easy to read, and the results are presented in a very clear way.
-
Reviewer #2 (Public Review):
Chintalapati et al. present an updated and refined version of the software DATES, for estimating the number of generations since admixture in either a single diploid individual or a pool of admixed samples. They show via extensive simulations that DATES performs well under many different individual conditions; including small samples sizes, missing data or pseudo-haploid genotype calling. The authors then systematically applied DATES to a large dataset of published ancient human SNP capture data, to investigate the timings of admixture between populations in Holocene West Eurasia.
Through simulations, the authors convincingly demonstrate that DATES outperforms comparable methods, and produces accurate inference under many challenging circumstances. However, the strength of the empirical analyses is …
Reviewer #2 (Public Review):
Chintalapati et al. present an updated and refined version of the software DATES, for estimating the number of generations since admixture in either a single diploid individual or a pool of admixed samples. They show via extensive simulations that DATES performs well under many different individual conditions; including small samples sizes, missing data or pseudo-haploid genotype calling. The authors then systematically applied DATES to a large dataset of published ancient human SNP capture data, to investigate the timings of admixture between populations in Holocene West Eurasia.
Through simulations, the authors convincingly demonstrate that DATES outperforms comparable methods, and produces accurate inference under many challenging circumstances. However, the strength of the empirical analyses is undermined by the lack of simulations which explicitly test the performance of DATES under the combination of conditions present in the empirical data. Many of the empirical analyses contain all of the problematic features that are only tested in isolation in the simulations-e.g., small sample sizes in both the target and reference populations, with non-contemporaneous sampling, variable data missingness, and pseudo-haploid genotypes. For example, in Figure 1b the authors show the effects of pseudo-haploid genotypes and data missingness for a target population of size n=10; however, this is substantially larger than the average target population used in the empirical analyses, and does not test the combined effects of small sample sizes and poor data quality in the reference populations. The authors also show that the performance of DATES is sensitive to low Fst between the admixing populations, but do not provide simulations calibrated to the levels of Fst between the empirical reference populations. These issues make it difficult to interpret the robustness of the empirical analyses, despite the large number of simulations presented.
Overall, the paper is well written, and the authors do a good job of presenting a complex series of results. The software DATES represents a significant improvement over comparable methods and has the potential to make substantial contributions to the study of admixture history in many species.
-
