On the limits of fitting complex models of population history to f-statistics

Admixture graph models provide a powerful intellectual framework for describing the relationships among populations that allows not only branching of populations from a common ancestor but also mixture events. An admixture graph (abbreviated below as AG), as fit in the widely used software packages ADMIXTOOLS (Patterson et al., 2012) and TreeMix (Pickrell and Pritchard, 2012; Molloy et al., 2021), is a directed acyclic bifurcating graph with two types of edges: those representing genetic drift, and those representing gene flow. Each admixture event is represented as a confluence of two gene flow edges. Nodes of such a graph represent unsampled intermediate populations, and terminal nodes (leaves) represent sampled present-day or ancient groups (see a mathematical definition in Soraggi and Wiuf, 2019). An attractive feature of AGs is that they can summarize important features of population history without requiring specification of all parameters such as population sizes, split times, mixture times, and distinguishing between sudden splits or drawn-out separations. All these parameters describe important features of demographic history and are fit by many methods for fitting demographic models (Gutenkunst et al., 2009; Gronau et al., 2011; Schiffels et al., 2016; Flegontov et al., 2019; Kamm et al., 2020, Rogers, 2019; Hubisz et al., 2020). However, the fact that it is possible to factor this difficult problem by first inferring important aspects of the topology (AGs fitted to allele frequency correlation statistics), and then fitting additional demographic parameters to data such as site frequency spectra, simplifies demographic inference (Patterson et al., 2012; Pickrell and Pritchard, 2012; Lipson et al., 2013; Leppälä et al., 2017; Lipson et al., 2020b; Molloy et al., 2021; Yan et al., 2021). AGs thus serve both as conceptual frameworks that allow us to think about the relationships of populations deep in time, and as mathematical models we can fit to genetic data.

AGs are fitted to f-statistics (Reich et al., 2009; Patterson et al., 2012; Peter, 2016; Soraggi and Wiuf, 2019). For convenience, below we use a concise definition of f-statistics by Lipson, 2020a: ‘The most general definition is that of the f₄-statistic f₄(A, B; C, D), which measures the average correlation in allele frequency differences between (1) populations A and B and (2) populations C and D that is, (p_A – p_B)∗ (p_C – p_D), for allele frequencies p, typically averaged over many biallelic single-nucleotide polymorphisms. This f₄-statistic is the same as the D-statistic up to a normalization factor.’ The other f-statistics (f₂ and f₃) can be defined as special cases of f₄-statistics: f₂(A, B) = f₄(A, B; A, B) and f₃(A; B, C) = f₄(A, B; A, C). f₄-Statistics can be written as linear combinations of f₃- or f₂-statistics, and f₃-statistics can be written as linear combinations of f₄- and f₂-statistics. f₂-, f₃-, and f₄-statistics have straightforward interpretations in terms of drift edges along the tree, see Figure 2 in Patterson et al., 2012 and Appendix 1—figure 1b. A challenge for fitting AG models is that they are often not uniquely constrained by the data, with many providing equally good fits to the f₂-, f₃-, and f₄-statistics used to constrain them within the limits of statistical resolution. Previously published methods for finding fitting AGs (mainly qpGraph, Patterson et al., 2012 and TreeMix, Pickrell and Pritchard, 2012; Molloy et al., 2021) were not well equipped to handle the large range of equally well-fitting models for three reasons: (1) They did not reliably provide information on whether there is a uniquely fitting parsimonious model or alternatively whether there are many models that fit equally well to the limits of statistical resolution, (2) they did not provide formal goodness-of-fit tests, and related to this, (3) they did not provide tests for whether the difference between the fits of any two models is statistically significant. As a consequence, and as we demonstrate in what follows, many published AG models have been interpreted as providing more confidence than is merited about the extent to which genetic data allows us to disentangle ancestral relationships.

To appreciate these problems, we first need to consider two main approaches that were utilized to study demographic history with AGs.

The first approach is to identify AGs automatically, either without human intervention or with guidance. It is possible in theory to exhaustively test all possible graphs for a given set of populations and pre-specified number of admixture events, as implemented, for example, in the admixturegraph R package (Leppälä et al., 2017). An exhaustive approach can provide a complete view of the range of models that are consistent with the data for a specified level of parsimony (total number of admixture events allowed in the graph), which is not biased by the algorithm used to explore the space of possible AGs. However, this approach is limited to small graphs (typically up to six groups, two admixture events) due to the rapid increase in the number of possible AGs as the number of populations and admixture events grows. As we show in our discussion of case studies, the simple models explored with an exhaustive approach can lead to misleading conclusions about population history because not including additional populations can blind users to additional mixture events that occurred (and whose existence is revealed by examining data from additional populations). Furthermore, models with additional admixture events that are qualitatively different to the best-fitting parsimonious graph and that capture the true history, will sometimes be completely missed when constraining the number of gene flows. Alternatively, the programs TreeMix (Pickrell and Pritchard, 2012; Molloy et al., 2021), MixMapper (Lipson et al., 2013), miqoGraph (Yan et al., 2021), and AdmixtureBayes (Nielsen et al., 2023 preprint) all address the problem of how to rapidly explore the vast space of AGs relating a set of populations by applying algorithmic ideas or heuristics; all of these methods speed up model search by orders of magnitude.

The second approach to fitting AGs is to manually build them up by grafting additional populations onto simpler smaller graphs that fit the data. This approach involves stepwise addition of populations in an order that is chosen based on the best judgment of the user, and for each newly added population involves adding admixture events or tweaks in the graph until a fit is obtained; the user then moves on to adding the next population (see Reich et al., 2009; Reich et al., 2011; Reich et al., 2012; Lazaridis et al., 2014; Seguin-Orlando et al., 2014; Fu et al., 2016; Skoglund et al., 2016; Yang et al., 2017; McColl et al., 2018; Moreno-Mayar et al., 2018; Tambets et al., 2018; van de Loosdrecht et al., 2018; Flegontov et al., 2019; Sikora et al., 2019; Wang et al., 2019; Lipson et al., 2020b; Shinde et al., 2019; Yang et al., 2020; Hajdinjak et al., 2021; Wang et al., 2021; Bergström et al., 2022 for examples). The program qpGraph in the ADMIXTOOLS package (Patterson et al., 2012) has been the most common computational method used for testing fits of individual AGs. Most AGs in the literature have been constructed manually in this way, often acknowledging the existence of alternative models by presenting plausible models side-by-side, and this approach has been the basis for many claims about population history (Lazaridis et al., 2014; Yang et al., 2017; Posth et al., 2018; Sikora et al., 2019; Shinde et al., 2019; Bergström et al., 2020; Lipson et al., 2020b; Hajdinjak et al., 2021; Wang et al., 2021; Bergström et al., 2022). A strength of this approach is that it takes advantage of human judgment and outside knowledge about what graphs best fit the history of the human or animal populations being analyzed. This external information is powerful as it can incorporate non-genetic evidence such as geographic plausibility and temporal ordering of populations or linguistic similarity, or other genetic data such as estimates of population split times, or shared Y chromosomes, or rejection of proposed scenarios based on joint analysis of much larger numbers of populations than can reasonably be analyzed within a single AG. Thus, while manual approaches explore many orders of magnitude fewer topologies than automatic approaches often do, they still may provide inferences about population history that are more useful than those provided by automatic approaches. These methods’ strength is also their weakness: by relying on intuition, following a manual approach has the potential to validate the biases users have as to what types of histories are most plausible (these may be the only types of histories that will be carefully explored). This can blind users to surprises: to profoundly different topologies that may correspond more closely to the true history, and we discuss examples of this in the Results section.

In this study, we introduce a new method, findGraphs, that belongs to the first class of algorithms (those for automated AG topology inference). Algorithmic innovations and speedups in findGraphs enable us to explore a much larger proportion of plausible AG space than many other methods reported to date. The findGraphs method combines the advantages of automated and manual topology exploration by allowing users to encode various sources of information as constraints on the space of AGs, which is then explored automatically. However, the main innovations in findGraphs are not computational, but instead conceptual. Instead of finding one or a few AGs fitting the data well, we use findGraphs for exploring AG spaces and assessing if any reliable information on population history can be extracted from a given AG space (defined by a population set and parsimony constraints) in the first place.

Regardless of the approach used to search through the space of possibly fitting AGs, a challenge in the effort to find a uniquely well-fitting AG (or group of topologically similar AGs) is that it has been difficult to quantify the absolute goodness of fit of a model to date. We have not been entirely successful with this and are not aware of other work that has been successful. It is also difficult to assess the relative fits of multiple models, especially if they differ in complexity. Performance gains relative to the original implementation of qpGraph allow us to address this problem by obtaining bootstrap confidence intervals and p-values for estimated parameters of single models, as well as for the difference in fit quality of two models (see Appendix 1, Sections 1.B.3 and 2.E). In combination with the approach to automating the search of well-fitting AGs, this leads to a situation where we are able to find and test a large number of models, many of which fit equally well despite often having very different topological features. Published approaches to comparing the fits of AG models based on Akaike information criterion (AIC) or Bayesian information criterion (BIC), see Flegontov et al., 2019; Shinde et al., 2019 have the problem that it is often not clear what the effective number of degrees of freedom is in the two models being compared since in the case of AGs it depends not only on the number of graph edges, but also on graph topology.

The methods for automated graph topology inference and model comparison relying on bootstrap resampling are implemented in ADMIXTOOLS 2, a comprehensive platform for learning about population history from f-statistics. It is built to provide a stand-alone workspace for research in this area and is implemented as an R package. For all computations, ADMIXTOOLS 2 exhibits large speedups relative to previously published platforms for f-statistic analysis (e.g., popstats and ADMIXTOOLS version 6.0 which we call ‘Classic ADMIXTOOLS’ in what follows to distinguish it from updated ADMIXTOOLS version 7.0.2 which implements some of the speedup ideas also implemented in ADMIXTOOLS 2). This is achieved by deploying a series of algorithmic improvements, most notably storage of precomputed f-statistics in random access memory, which avoids having to rely on reading in extremely large genotype matrices to perform most computations. In addition to the new algorithmic ideas allowing efficient searching through the space of AGs and comparing the fits of two AGs, ADMIXTOOLS 2 also provides a solution to the question of which parameters of an AG are identifiable in the limit of infinite data. Methodological details are presented in Appendix 1, and below we focus on documenting problems of AG inference on simulated data and revisiting AGs from the literature to understand the extent to which methodological challenges with AG fitting biased previous studies.

Topological diversity of well-fitting models and effects of parsimony constraints on simulated data

First, we explored the performance of the findGraphs method for automated topology inference on simulated AGs of random topology, focusing on the following questions: (1) among findGraphs results, how common are AGs fitting nominally or significantly better than the true one but different topologically; (2) what is the degree of topological diversity among these models fitting the data better than the true one? For this purpose, we simulated AGs of four complexity classes using msprime v.1.1.1: eight or nine non-outgroup populations, and four or five admixture events. Only simulations where pairwise F_ST for groups were in the range characteristic for anatomically modern and archaic humans were selected for further analysis, resulting in 20 random topologies per complexity class, each including a distant outgroup that facilitates automated exploration of the topology space.

We ran findGraphs on each simulated dataset starting from random graphs and pre-specifying the true number of admixture events (n), or n − 1, or n + 1 events. For each of these graph complexity levels, we performed 100 independent findGraphs runs and recorded 5 AGs from each run having the best log-likelihood (LL) scores. Topologically redundant AGs were discarded, and for the remaining AGs we calculated worst f₄-statistic residuals (WR) and tested if the newly found models fit significantly better than the true model, using the bootstrap model comparison method developed in this study (see Appendix 1, Sections 1.B.3 and 2.E). In Figure 1a–c, we show the following statistics for each simulated AG, summarized across simulated complexity classes and parsimony levels allowed at the stage of topology exploration: fraction of topologies found with findGraphs that fit better than the true AG (according to LL score), or that fit significantly better than the true AG, or those with plausible absolute fits (WR < 3 SE). It is clear that for the great majority of simulated datasets, even a shallow exploration of the topology space with findGraph (100 independent runs) uncovers AGs that fit nominally better than the true topology (Figure 1a and d) and are topologically diverse (see Figure 1e for examples). When allowing for n admixture events, at least one AG fitting significantly better than the true one was found for 60% of simulated datasets. When n + 1 admixture events were allowed, this grew to 100% (all 80 datasets). It should be noted that some admixture events are indistinguishable with f-statistics; for instance, successive gene flows between two lineages, with no other edges branching off between the gene flows. If such gene flows were included in the random topologies we simulated, AGs with n events were overly complex for representing the true history. Thus, if we are dealing with random histories, choosing an optimal complexity class for topology search is not straightforward.

Figure 1

Download asset Open asset

These results on simulated data raise concerns about the extent to which fitting AG topologies provide reliable information about population history. Even for histories including eight or nine groups, an outgroup, and four or five pulse-like admixture events, perfect diploid data, and groups as differentiated as Neanderthals and anatomically modern humans—a complexity class that is simpler than many models fitted to real genetic data in published papers—models fitting the data as well as or better than the true one are common, and their topological diversity is in most cases so high that it precludes consensus inference of topology by analysis of multiple topologies. As we demonstrate on another set of simulated data in Appendix 1 (Section 1.B.1), the probability of finding a ‘wrong’ model that fits better than the true one grows with increasing graph complexity, and that effect is reproduced with both findGraphs and TreeMix (Appendix 1—figure 2). We expect this problem to be even more acute when researchers are dealing with realistic complex histories. However, geneticists often rely on external constraints on AG topologies (such as temporal plausibility of a topology, results from qpAdm modelling, f₄-statistics, PCA, ADMIXTURE, geographical, archaeological, and linguistic considerations) that were not used for filtering the results of our topology searches on simulated data. Thus, it is possible in principle that published AG models are more robust than our results on simulated data, and we explore this issue in depth in the next section and in Appendix 2.

Revisiting published AGs

We studied AGs from eight publications (Lazaridis et al., 2014; Shinde et al., 2019; Sikora et al., 2019; Bergström et al., 2020; Lipson et al., 2020b; Hajdinjak et al., 2021Librado et al., 2021; Wang et al., 2021) with the goal of comparing published models to models identified by our algorithm for automatically inferring optimal (best-fitting) AGs (Table 1). In all but one study, qpGraph or its automated reimplementation (admixturegraph, Leppälä et al., 2017) was used for fitting topologies to genetic data, while Librado et al., 2021 relied on the automated OrientAGraph method (Molloy et al., 2021). The main question we were interested in is whether we can find alternative models which (1) fit as well as, or better than the published graph, (2) differ in important ways from the published graph, and (3) cannot immediately be rejected based on other evidence such as temporal plausibility. The studies were selected according to the criterion that an AG model inferred in the study is used as primary evidence for at least one statement about population history in the main text of the study. In other words, the AG method was used in the original studies to support new conclusions about population history, and not simply to show that there is a model that exists that does not contradict results of other genetic analyses, an approach that is a valid use of AGs and has been taken in some studies (e.g., Seguin-Orlando et al., 2014; Narasimhan et al., 2019; Wang et al., 2019). There are many published studies that could have been included in our re-evaluation exercise as they meet our key criterion (e.g., Yang et al., 2017; McColl et al., 2018; Posth et al., 2018; Flegontov et al., 2019; Carlhoff et al., 2021, Kutanan et al., 2021; Bergström et al., 2022; Lipson et al., 2022; Vallini et al., 2022). However, critical re-evaluation of each published graph is an intensive process, and the sample of studies we revisited is diverse enough to identify some general patterns.

Table 1

Publication	Figure in the original publication	Groups (populations)	Admixture events	SNPs used	Publ. model: worst residual, SE	Distinct alternative topologies found	Significantly better fitting topologies, %	Non-significantly better fitting topologies, %	Non-significantly worse fitting topologies, %	Significantly worse fitting topologies, %
Bergström et al., 2020	1e	7	3	312,282	2.1	221	0.5	2.3	16.7	80.5
Lazaridis et al., 2014	3	7	4	642,247	2.2	306	1.0	12.1	80.7	6.2
Shinde et al., 2019	3	8*	3	2,49,009	2.6	143	0.0	2.8	3.5	93.7
Librado et al., 2021	3b	10*	3	1,767,419	23.9	324	6.8	15.7	24.1	53.4
	Ext 5d		4		14.1	535	0.0	0.0	4.5	95.5
	Ext 5e		5		6.9	784	0.0	0.3	28.4	71.3
Hajdinjak et al., 2021	2d	12	8	263,698	4.8	1988	15.7	55.7	6.6	22.0
Lipson et al., 2020b	Ext 4	12	11†	211,738	2.3	2000	0.0	11.9	77.1	10.4
Wang et al., 2021	Ext 6	12	8	203,753	3.8	1778	12.6	84.3	3.1	0.0
Sikora et al., 2019	3f (left)	13	6†	344,903	3.8	894	0.3	17.1	34.6	48.0
	3f (right)	14	6†	613,509	4.2	2785	0.1	0.9	9.8	89.2

1. Publication: Last name of the first author and year of the relevant publication.

2. Figure in the original publication: Figure number in the original paper where the AG is presented.

3. Groups (populations): The number of populations in each graph.

4. Admixture events: The number of admixture events in each graph.

5. SNPs used: The number of single-nucleotide polymorphisms (SNPs; with no missing data at the group level) used for fitting the AGs. For all case studies, we tested the original data (SNPs, population composition, and the published graph topology) and obtained model fits very similar to the published ones. However, for the purpose of efficient topology search, we in some cases adjusted settings for f₃-statistic calculation, population composition, or graph complexity as noted in the footnotes, in Supplementary file 1, and discussed in the text.

6. Publ. model: Worst residual, SE: The worst f-statistic residual of the published graph fitted to the SNP set shown in the ‘SNPs used’ column, measured in standard errors (SE).

7. Distinct alternative topologies found: The number of distinct newly found topologies differing from the published one.

8. Significantly better fitting topologies, %: The percentage of distinct alternative topologies that fit significantly better than the published graph according to the bootstrap model comparison test (two-tailed empirical p-value <0.05). If the number of distinct topologies was very large, a representative sample of models (1/20 to 1/3 of models evenly distributed along the log-likelihood spectrum) was compared to the published one instead, and the percentages in this and following columns were calculated on this sample.

9. Non-significantly better fitting topologies, %: The percentage of distinct topologies that fit non-significantly (nominally) better than the published graph according to the bootstrap model comparison test (two-tailed empirical p-value ≥0.05).

10. Non-significantly worse fitting topologies, %: The percentage of distinct topologies that fit non-significantly (nominally) worse than the published graph according to the bootstrap model comparison test (two-tailed empirical p-value ≥0.05).

11. Significantly worse fitting topologies, %: The percentage of distinct topologies that fit significantly worse than the published graph according to the bootstrap model comparison test (two-tailed empirical p-value <0.05).

12. *

    The population composition was modified, see Supplementary file 1 and the text.

13. †

    Certain gene flows were removed from the published model for simplicity, see Supplementary file 1 and the text.

Here we present a high-level summary of these analyses. Discussion of individual case studies follows below, and for details see the exposition in Appendix 2.

For 19 out of 22 published graphs we examined, we were able to find at least one, but usually many, graphs of the same complexity (number of groups and admixture events), with an LL score that was nominally better than that of the published graph (see results for 11 selected graphs in Table 1 and full results for all 22 graphs in Supplementary file 1). The 22 graphs were drawn from the 8 publications as there were multiple final graphs presented in some of the publications (Shinde et al., 2019; Sikora et al., 2019; Librado et al., 2021), or we examined selected intermediates in the model construction process (Bergström et al., 2020; Lazaridis et al., 2014; Lipson et al., 2020b; Wang et al., 2021), or we introduced an outgroup not used in the original study (Hajdinjak et al., 2021; Sikora et al., 2019), or we tested additional graph complexity classes dropping ‘unnecessary’ admixture events (Lipson et al., 2020b; Sikora et al., 2019).

These alternative graphs often fit not significantly better than the published one after taking into account variability across single-nucleotide polymorphisms (SNPs) via bootstrapping. In the following cases, at least one model that fits significantly better than the published one according to our bootstrap model comparison method was found: the Bergström et al. and Lazaridis et al. 7-population graphs; the Librado et al., 2021 graph with 3 admixture events; the Hajdinjak et al. graphs with or without adding a chimpanzee outgroup; the Lipson et al., 2020b. intermediate graphs with 7 groups and 4 admixture events and with 10 groups and 8 admixture events; the Wang et al. 12-population graph; and the Sikora et al. graphs for West Eurasians and for East Eurasians with 10 or 6 admixture events (Supplementary file 1). In nearly all cases (except for the Lazaridis et al. six-population graph, Shinde et al. graph with eight populations and three admixture events, and the Librado et al., 2021. graph with four admixture events), we also identified many additional graphs that fit the data not significantly worse than the published ones. In every example, some of these graphs have topologies that are qualitatively different in important ways from those of the published graphs. Features such as which populations are admixed or unadmixed, direction of gene flow, or the order of split events, if not constrained a priori, are generally not the same between alternative fitting models for the same populations. This result agrees with the expectation from our exploration of simulated AGs (Figure 1). While some of these graphs can be rejected since their topologies appear highly unlikely because of non-genetic or unrelated genetic evidence, for all of the publications except one (Shinde et al., 2019), there are alternative equally-well-or-better-fitting graphs we identified and examined manually that differ in qualitatively important ways with regard to the implications about history, are temporally plausible (for instance, very ancient populations do not receive gene flows from sources closely related to much less ancient groups), and not obviously wrong based on other lines of evidence. These findings and the results on simulated AGs suggest that complex AG models, even with a very good fit to the data, often differ in important ways from true population histories.

The previous statements are valid if the original parsimony constraints are applied, that is, if the graph complexity (the number of admixture events) is not altered. Below in selected case studies (Shinde et al., Librado et al., 2021.) we also explore the effect of relaxing the parsimony constraint. Table 1 and Figure 2 summarize these results for one or a few graphs from each publication, while Supplementary file 1 contains the full results for all studied graphs and setups. Table 2 summarizes our assessment of inferences in the original publications that were supported by the published graphs.

Figure 2

Download asset Open asset

Table 2

Study	Groups/admixture events	Features of the published model supported by all temporally plausible alternative models generated by findGraphs that we scrutinized	Features of the published model lacking universal support among temporally plausible alternative models generated by findGraphs
Bergström et al., 2020	7/3*	Early divergence of domesticated dog lineages (prior to the date of the Karelian dog, 10,900 ya).	Siberian (Baikal), American, and East Mediterranean dog lineages are unadmixed, and the West European (Germany Early Neolithic), East European (Karelia), and dogs of Southeast Asian origin (New Guinea singing dog) are admixed.
Lazaridis et al., 2014	7/4	Present-day Europeans represent a mixture of three ancestral sources related to the following groups: Mal’ta (MA1), West European hunter–gatherers, and early European farmers.	N/A
Shinde et al., 2019	8/3*	(1) Iranian farmer-related ancestry in the Indus Periphery group is not derived from the Hajji Firuz Neolithic or Tepe Hissar Chalcolithic groups. (2)There is Asian-related ancestry in the Indus Periphery group.	N/A
	8/4	(2) There is Asian ancestry in the Indus Periphery group.	(1) Iranian farmer-related ancestry in the Indus Periphery group is not derived from the Hajji Firuz Neolithic or Tepe Hissar Chalcolithic groups.
Librado et al., 2021	10/8 or 9	(2) DOM2 and C-PONT are sister groups (they form a clade). (4) There is gene flow from a deep-branching ghost group to the NEO-ANA group.	(1) NEO-ANA-related admixture is absent in the DOM2 group. (3) There is no gene flow connecting the CWC group and the cluster associated with Yamnaya horses and horses of the later Sintashta culture whose ancestry is maximized in the Western Steppe (DOM2, C-PONT, TURG). (5) Tarpan is a mixture of a CWC-related and a DOM2-related lineage.
Hajdinjak et al., 2021	12/8	(3) The Vestonice16 lineage is a mixture of a Sunghir-related and a BK1653-related lineage.	(1) There are gene flows from the lineage found in the ~45,000- to 43,000-year-old Bacho Kiro Initial Upper Paleolithic (IUP)-associated lineage to the Ust’-Ishim, Tianyuan, and GoyetQ116-1 lineages. (2) The ~35,000-year-old Bacho Kiro Cave individual BK1653 belonged to a population that was related, but not identical, to that of the GoyetQ116-1 individual.
Lipson et al., 2020b	12/11	N/A	(1) A lineage maximized in present-day West African groups (Lemande, Mende, and Yoruba) also contributed some ancestry to the ancient Shum Laka individual and to present-day Biaka and Mbuti. (2) Another ancestry component in Shum Laka is a deep-branching lineage maximized in the rainforest hunter–gatherers Biaka and Mbuti. (3) ‘Super-archaic’ ancestry (i.e., diverging at the modern human/Neanderthal split point or deeper) contributed to Biaka, Mbuti, Shum Laka, Lemande, Mende, and Yoruba. (4) A ghost modern human lineage (or lineages) contributed to Agaw, Mota, Biaka, Mbuti, Shum Laka, Lemande, Mende, and Yoruba.
Wang et al., 2021	12/8	N/A	Admixture from a source related to Andamanese hunter–gatherers is almost universal in East Asians, occurring in the Jomon, Tibetan, Upper Yellow River Late Neolithic, West Liao River Late Neolithic, Taiwan Iron Age, and China Island Early Neolithic (Liangdao) groups.
Sikora et al., 2019 ‘West’	13/6	N/A	The Mal’ta (MA1_ANE) lineage received gene flow from the Caucasus hunter–gatherer (CaucasusHG_LP or CHG) lineage.
Sikora et al., 2019 ‘East’	14/6	(2) European-related ancestry in the Kolyma, USR1, and Clovis lineages is closer to Mal’ta than to Yana.	(1)The Mal’ta (MA1_ANE) and Yana (Yana_UP) lineages received gene flow from a common East Asian-associated source diverging before the ones contributing to the Devil’s Cave (DevilsCave_N), Kolyma (Kolyma_M), USR1 (Alaska_LP), and Clovis (Clovis_LP) lineages. (3)The Devil’s Cave lineage received no European-related gene flows, and Kolyma has less European-related ancestry than ancient Americans (USR1 and Clovis).

To identify alternative models, we ran many iterations of findGraphs for each set of input populations, constraints, and the number of admixture events being fit to the data, and we selected the best-fitting graph in each iteration, that is, a graph with the lowest LL score. Each iteration was initiated from a random graph. The algorithm is non-deterministic so that in each iteration it takes a different trajectory through graph space, possibly terminating in a different final best graph. The number of admixture events in the initial random graphs and in the output graphs was always kept equal to that of the published graph. For each example, we counted how many distinct topologies were found with significantly or non-significantly better or worse LL scores than that of the published graph (Table 1, Supplementary file 1). To obtain a formally correct comparison of model fit, the published graph and each alternative model were fitted to resampled replicates of the dataset and the resulting LL score distributions were compared (see Appendix 1, Sections 1.B.3 and 2.E). As shown in Figure 2, for four of the eight publications we re-analyzed, the LL score of the published graph run on the full data is better than almost all the bootstrap replicates on the same data (it falls below the fifth percentile), which is a sign of overfitting, and underscores the importance of applying bootstrap to assess the robustness of fitted models and conclusions drawn from them.

The fraction of graphs with scores better than the score of the published graph should not be overinterpreted, as it is influenced by the findGraphs algorithm, which does not guarantee ergodic sampling from the space of well-fitting AGs. In particular, it is possible that despite findGraph’s strategies for efficiently identifying classes of well-fitting AGs (see Appendix 1, Sections 1.B.1 and 2.C), it has a bias toward missing particular classes of graph topologies. However, even one alternative graph which is not significantly worse-fitting than the published graph suggests that we are not able to identify a single best-fitting model. Many of these alternatives, despite providing a good fit to the data, appear unlikely, for example, because they suggest that Paleolithic-era humans are mixed between different lineages closely related to present-day humans. We were mainly interested in alternative models which are also plausible, and so we constrained the space of allowed topologies in findGraphs to those we considered plausible a priori, in cases where this was necessary for reducing the search space size. Constraints were either integrated into the topology search itself, or were applied to outcomes of unconstrained searches, as detailed below.

Below we summarize our key findings and the methodological implications from our re-analysis of the eight published datasets. For more detailed discussions see Appendix 2.

1. Bergström et al., 2020

The AG for ancient and present-day dogs in Figure 1e of Bergström et al., 2020 includes an outgroup, six other groups and three admixture events (Figure 3a, Figure 3—source data 1). A best-fitting newly found graph fits the data nominally better than the published one (two-tailed empirical p-value = 0.332), and it bears a closer resemblance to the human population history (Figure 3—source data 2). In this new seven-population model (Figure 3a), both American and Siberian dog lineages represent a mixture between groups related to the Asian and East European dog lineages, and robust genetic results suggest that in the time horizon investigated in the original publication (after ca. 10,900 years ago) nearly all Siberian (Jeong et al., 2019; Sikora et al., 2019) and all American (Raghavan et al., 2014; Raghavan et al., 2015; Moreno-Mayar et al., 2018) human populations were admixed between groups most closely related to Europeans and East Asians. According to this model, East Mediterranean dogs are modeled as a mixture of a basal branch (splitting deeper than the divergence of the Asian and European dogs) and West European dogs, again in agreement with current models of genetic history of West Asian human populations who are modeled as a mixture of ‘basal Eurasians’ and West European hunter–gatherers (Lazaridis et al., 2016; Lipson et al., 2017). Although greater congruence with human history increases the plausibility of findGraph’s newly identified model relative to the published model, to make unbiased comparisons between the history of the two species, model selection should be done strictly independently for each species, and so the genetic data alone does not favor one model more than another.

Figure 3

Download asset Open asset

Figure 3—source data 1

The published (Bergström et al., 2020) and alternative admixture graphs for dogs found with findGraphs.

Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The published model, (b) alternative models fitting nominally better than the published one, sorted by the fit score, and (c) alternative models fitting nominally or significantly (the graph framed in blue) better than the published one, sorted by the fit score.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data1-v2.pdf

Download elife-85492-fig3-data1-v2.pdf

Figure 3—source data 2

Alternative admixture graphs for humans found with findGraphs for the dataset from Bergström et al., 2020.

Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a, b) Best-fitting models for humans sorted by the fit score.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data2-v2.pdf

Download elife-85492-fig3-data2-v2.pdf

Figure 3—source data 3

The published admixture graph from Lazaridis et al., 2014 and alternative graphs found with findGraphs (seven populations, four admixture events).

Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The re-fitted published graph and (b) 10 examples of graphs inferred by findGraphs (arranged according to LL score) and fitting significantly (the graph framed in blue) or nominally better than the published model.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data3-v2.pdf

Download elife-85492-fig3-data3-v2.pdf

Figure 3—source data 4

The published admixture graph from Shinde et al., 2019 and alternative graphs found with findGraphs (8 pops., 3 adm. events) relying on the original set of SNPs and group composition, and original (incorrect) algorithm for calculating f-statistics.

Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The published model; the original set of SNPs and individuals, and the original algorithm for calculating f₃-statistics was used (470,389 variable sites with no missing data at the group level available). The following claim in Shinde et al., 2019 relies on the admixture graph: Primary ancestry in the Indus Periphery group forms the deepest branch in the Iranian Neolithic clade composed of the Indus Periphery, Ganj Dareh Neolithic, Hajji Firuz Neolithic, and Tepe Hissar Chalcolithic groups. (b-f) Selected alternative models fitting significantly better (graphs framed in blue), nominally better (graphs without frames), or not significantly worse (graphs framed in red) than the published one.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data4-v2.pdf

Download elife-85492-fig3-data4-v2.pdf

Figure 3—source data 5

The published admixture graph from Shinde et al., 2019 and alternative graphs found with findGraphs (eight populations, three admixture events) for the modified group composition and using the updated algorithm for calculating f-statistics.

The graphs were also re-fitted on the original set of single-nucleotide polymorphisms (SNPs)/individuals and using the original algorithm for calculating f-statistics. Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The published model with three admixture events. The following claim in Shinde et al., 2019 relies on the admixture graph: Primary ancestry in the Indus Periphery group forms the deepest branch in the Iranian Neolithic clade composed of the Indus Periphery, Ganj Dareh Neolithic, Hajji Firuz Neolithic, and Tepe Hissar Chalcolithic groups. (b) Alternative models fitting nominally better than the published one and confirming all of its important topological details and (c) alternative models fitting not significantly worse than the published one and differing from it in important ways.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data5-v2.pdf

Download elife-85492-fig3-data5-v2.pdf

Figure 3—source data 6

Alternative graphs allowing for an additional admixture event found with findGraphs for the dataset from Shinde et al., 2019: 8 populations, 4 admixture events, the modified group composition, and the updated algorithm for calculating f-statistics.

The graphs were also re-fitted on the original set of single-nucleotide polymorphisms (SNPs)/individuals and using the original algorithm for calculating f-statistics. Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The highest-ranking model with four admixture events that confirms all important features of the published model with three admixture events. The following claim in Shinde et al., 2019 relies on the admixture graph: Primary ancestry in the Indus Periphery group forms the deepest branch in the Iranian Neolithic clade composed of the Indus Periphery, Ganj Dareh Neolithic, Hajji Firuz Neolithic, and Tepe Hissar Chalcolithic groups. (b) Alternative models fitting not significantly worse than the highest-ranking one and contradicting the historical interpretation of the admixture graph results by Shinde et al., 2019.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data6-v2.pdf

Download elife-85492-fig3-data6-v2.pdf

Figure 3—source data 7

The published admixture graphs from Librado et al., 2021 and alternative graphs found with findGraphs (10 populations, 3–5 admixture events) for the modified group composition and using the updated algorithm for calculating f-statistics.

The graphs were also re-fitted on the original set of single-nucleotide polymorphisms (SNPs)/individuals and using the original algorithm for calculating f-statistics. Selected alternative graphs found with findGraphs when more admixture events were allowed (from 6 to 9) are also shown. Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The published model, three admixture events. The following claims in Librado et al., 2021 rely on the admixture graph: (1) NEO-ANA-related admixture is absent in DOM2; (2) DOM2 and C-PONT are sister groups; (3) there is no gene flow connecting the CWC group and the cluster associated with Yamnaya horses and horses of the later Sintashta culture whose ancestry is maximized in the Western Steppe (DOM2, C-PONT, and TURG); (4) there was a gene flow from a deep-branching ghost group to NEO-ANA; (5) Tarpan is a mixture of a CWC-related and a DOM2-related lineage. (b) An alternative model with three admixture events fitting significantly better than the published one, (c) the published model, four admixture events, (d) an alternative model with four admixture events fitting not significantly worse than the published one, (e) the published model, five admixture events, (f) an alternative model with five admixture events fitting not significantly worse than the published one, (g) selected models with six admixture events, (h,i) selected models with seven admixture events (all plausible models with WR < 5 SE), (j-l) selected models with eight admixture events (all plausible models with WR < 4 SE), (m-r) selected models with nine admixture events (all plausible models with WR < 4 SE).

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data7-v2.pdf

Download elife-85492-fig3-data7-v2.pdf

Figure 3—source data 8

Published admixture graph from Hajdinjak et al., 2021 and alternative graphs found with findGraphs (12 populations, 8 admixture events).

Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a, b) The published model and its simplified representation. The following claims in Hajdinjak et al., 2021 rely on the admixture graph: (1) gene flow from the Bacho Kiro lineage to Ust’-Ishim, Tianyuan, and GoyetQ116-1; (2) the BK1653 individual belonged to a population that was related, but not identical, to that of the GoyetQ116-1 individual; (3) Vestonice is a mixture of a Sunghir-related and a BK1653-related lineage. (c-e) Selected alternative models fitting significantly better than the published one.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig3-data8-v2.pdf

Download elife-85492-fig3-data8-v2.pdf

To explain why the original paper on the population history of dogs missed the model that findGraphs identified, we observe that the Bergström et al., 2020 AG search was exhaustive under the parsimony constraint (no more than two admixture events for six populations, with the seventh group added at a later stage without an exhaustive topology search), and thus missed the potentially true topology including three admixture events for these six populations. This case study also illustrates that even in a relatively low complexity context (seven groups and three admixture events) applying manual approaches for finding optimal models is risky. When any new group such as an Early Neolithic dog from Germany is added to the model, it may introduce crucial new constraints into the system, and re-exploring the whole graph space in an automated way is necessary to avoid missing the true model. In contrast, mapping a newly added group on a simple skeleton graph (even when that skeleton is a uniquely best-fitting model like in Bergström et al., 2020) may yield a topology that is at odds with the true history.

2. Lazaridis et al., 2014

The graph in Figure 3 (in Lazaridis et al., 2014) suggested that present-day Europeans are derived from at least three populations that are very much differentiated genetically: West European hunter–gatherers (WHG), early European farmers (EEF), and Siberian hunter–gatherers from the same lineage as that of the Mal’ta boy who lived about 24,000 years ago (MA1). For seven-population graphs with four admixture events, we found 40 out of 306 distinct graphs with a score better than that of the published graph (10 of those graphs are shown in Figure 3—source data 3). The best-fitting newly found model and two other models fit the data significantly better than the published model (Supplementary file 1), but their topology is qualitatively very similar to that of the published graph (Figure 3—source data 3). In the best-fitting newly found model, French and Karitiana share some drift to the exclusion of MA1, while in the published model the source of MA1-related ancestry in French is closer to MA1 than to Karitiana.

It is important to point out that not all of the 40 alternative graphs that fit nominally or significantly better than the published one are consistent with the conclusion that modern European populations are admixed between three different ancestral populations (Figure 3—source data 3). According to the fifth alternative graph in Figure 3—source data 3 that fits nominally better than the published model (p-value = 0.464), the present-day European population was formed by admixture of an MA1-related lineage and a European Neolithic-related lineage, with no WHG contribution. Of course, other lines of evidence make it clear that LBK Stuttgart is a mixture of Anatolian farmer- and WHG Loschbour-related ancestry (e.g., Lazaridis et al., 2016; Lipson et al., 2017), thus providing external information in favor of the Lazaridis et al., 2014 model. The use of such ancillary information in concert with graph exploration is important in order to obtain more confident inferences about population history taking advantage of AGs. We also note that a large group of newly found models (247 graphs) fits not significantly worse than the published one (Supplementary file 1), and those are topologically diverse. Thus, strictly speaking, the AG method on the given dataset cannot be used to prove that the published model is the only one fitting the data.

3. Shinde et al., 2019

The skeleton AG in the original study (Shinde et al., 2019) was constructed manually, and subsequently all possible branching orders (105) within the five-population Iranian farmer/herder-related clade were tested. The published model (Figure 3 in that study) included nine groups and three admixture events, but one group (Belt Cave Mesolithic) had a very high missing data rate. Following the approach of the published paper, we repeated findGraphs analysis both with and without the Belt Cave individual. Thus, we initially explored the following topology classes: 9 groups with 3 admixture events on ca. 19,000 polymorphic sites and 8 groups with 3 admixture events on ca. 470,000 sites (Figure 3—source data 4, Supplementary file 1). The finding that the predominant ancestry component of the Indus Periphery group was the most basal branch in the Iranian farmer clade was a prominent claim of the original study (Shinde et al., 2019). This finding if correct is important, since it implies that the Iranian-related ancestry in the Indus Valley Civilization genetic grouping (which is the same group as Indus Periphery or IP) split from the Iranian-related ancestry in the first Iranian plateau farmers before the date of the Hajji Firuz farmers, who at ~8000 years ago are among the earliest people living on the Iranian plateau known to have grown West Asian crops. The ancient DNA record combined with radiocarbon dating evidence suggests that beginning around the time of the Hajji Firuz farmers, both West Asian domesticated plants such as wheat and barley, and Anatolian farmer-related admixture, began spreading eastward across the Iranian plateau. If the Iranian-related ancestry in IP was spread eastward into the Indus Valley across the Iranian plateau as part of the same agriculturally associated expansion—perhaps brought by people speaking Indo-European languages as well as introducing West Asian crops—then we would expect to see at least some of the Iranian-related ancestry in IP being a clade with that in Hajji Firuz relative to Ganj Dareh.

The following groups were admixed by default in the graph models compared in the original study: Hajji Firuz Neolithic (labeled ‘Chalcolithic’ in that study but the dates are Neolithic) and Tepe Hissar Chalcolithic were considered as mixtures of an Anatolian farmer-related lineage and an Iranian farmer-related lineage. Indus Periphery was assumed to be a mixture of an Andamanese-related lineage representing ancient South Indians (ASI) and an Iranian farmer-related lineage. The original study differed from ours since these constraints were introduced manually, but we wanted our topology search to be automatic and to explore a wider range of parameter space.

To provide power to detect negative f₃-statistics useful for constraining the model search, we introduced several modifications to the original group composition (described in Appendix 2) and a new algorithm that makes it possible to compute negative f₃-statistics on pseudo-haploid data, but at a cost of removing sites with only one chromosome genotyped in any non-singleton population (see Appendix 1, Section 1.A). We repeated topology search with this set of f-statistics providing additional constraints, performing 4,000 runs of the findGraphs algorithm. The Mota ancient African individual was set as an outgroup and three admixture events were allowed in the eight-population graph. Among 4,000 resulting graphs (one from each findGraphs run), 144 were distinct topologically, and the published model was recovered in 13 runs of 4,000 (Supplementary file 1). Only four distinct topologies fitting nominally better than the published one were found, and those had LL scores almost identical to that of the published AG. These four alternative models (Figure 3—source data 5b) shared all topologically important features of the published model (Figure 3—source data 5). Five other topologies differed in important ways from the published one and emerged as fitting the data worse, but not significantly worse, than the published one (Figure 3—source data 5c).

But in fact, the AG analysis reported above may not be an adequate exploration of the problem. Although absolute fits of the best models found are good (WR = 2.5 SE), the parsimony constraint allowing only three admixture events precluded correct modeling of basal Eurasian ancestry shared by all West Asian groups (Lazaridis et al., 2016) or of the Indus Periphery group itself, for which a more complex 3-component admixture model was proposed elsewhere (Narasimhan et al., 2019). Concerned that this oversimplification could be causing our search to miss important classes of models, we explored qpAdm models for the Indus Periphery group, following the protocol by Narasimhan et al., 2019 (see the dataset composition in Supplementary file 4). Our qpAdm results (Supplementary files 5 and 6, Appendix 2) show that the parsimony assumption that was made when constructing the AG analysis in Shinde et al., 2019 is contradicted by f-statistic evidence since the simplest fitting qpAdm model for the IP group includes four ancestry sources, not two (Indus Periphery = Ganj Dareh Neolithic + Onge (ASI) + WSHG + Anatolia Neolithic), and indeed Narasimhan et al. themselves showed this when they presented a qpAdm model that was more complex (Ganj Dareh Neolithic + Onge (ASI) + WSHG) than the one used for constraining the AG model comparison (Ganj Dareh Neolithic + Onge (ASI)).

To explore how the parsimony constraint influences results, we allowed four admixture events in the eight-population graph (Supplementary file 1). Among 4,000 resulting graphs (one from each findGraphs run), 443 were distinct topologically, and 270 had WRs between 2 and 3 SE, that is, fitted the data well. In Figure 3—source data 6b, we show four graphs with four admixture events that model the Indus Periphery group as a mixture of three or four sources, with a significant fraction of its ancestry derived from the Hajji Firuz Neolithic or Tepe Hissar Chalcolithic lineages including both Iranian and Anatolian ancestries. The fits of these models are just slightly different (e.g., LL = 11.7 vs 9.3, both WRs = 2.4 SE) from that of the best-fitting model (Figure 3—source data 6a), and are similar to that of the simpler published graph. Besides these four illustrative graphs, dozens of topologies with very different models for the Indus Periphery group fit the data approximately equally well, suggesting that there is no useful signal in this type of AG analysis when the parsimony constraint is relaxed (this finding is similar to that in our re-analysis of the dog AG in Bergström et al., 2020, where relaxation of the parsimony constraint identified equally well-fitting AGs that were very different with regard to their inferences about population history). These results show that at least with regard to the AG analysis, a key historical conclusion of the study (that the predominant genetic component in the Indus Periphery lineage diverged from the Iranian clade prior to the date of the Ganj Dareh Neolithic group at ca. 10 kya and thus prior to the arrival of West Asian crops and Anatolian genetics in Iran) depends on the parsimony assumption, but the preference for three admixture events instead of four is hard to justify based on archaeological or other arguments.

Why did the Shinde et al., 2019 AG analysis find support for the IP Iranian-related lineage being the first to split, while our findGraphs analysis did not? Shinde et al., 2019 study sought to carry out a systematic exploration of the AG space in the same spirit as findGraphs—one of only a few papers in the literature where there has been an attempt to do so—and thus this qualitative difference in findings is notable. We hypothesize that the inconsistency reflects the fact that the deeply diverging WSHG-related ancestry (Narasimhan et al., 2019) present in the IP genetic grouping at a level of ca. 10% was not taken into account explicitly neither in the AG analysis nor in the admixture-corrected f₄-symmetry tests also reported in Shinde et al., 2019. The difference in qualitative conclusions may also reflect the fact that the Shinde et al. study was distinguishing between fitting models relying on an LL difference threshold of 4 units (based on the AIC). As discussed in Appendix 1, AIC is not applicable to AGs where the number of independent model parameters is topology-dependent even if the numbers of groups and admixture events are fixed, and models compared with AIC should have the same number of parameters. Thus, we believe that the analysis by Shinde et al. was over-optimistic (as compared to the bootstrap model comparison method we use) about being able to reject models that were in fact plausible using its AG fitting setup.

4.Librado et al., 2021

In contrast to the other studies revisited in our work, the AG published by Librado et al., 2021 was inferred automatically using OrientAGraph. Models with three (Figure 3b in that study) and zero to five (Ext. Data Figure 5a–d in that study) admixture events were shown. The dataset included 10 populations (9 horse populations and donkey as an outgroup) and was based on 7.4 million polymorphic transversion sites with no missing data at the group level. Unlike all the other AGs we re-evaluate in this study whose fits to the data were evaluated in the published studies using qpGraph, the topologies published in Librado et al., 2021 were not evaluated for statistical goodness-of-fit, and in fact fit the f-statistic data so poorly that even simple statistics show they cannot be correct (Figure 3b, Figure 3—source data 7a, c, e,, Supplementary file 1). In this case, the approach of using findGraphs to identify alternative topologies with the same number of admixture events that fit the data better is meaningless, as both the published models and the alternative models do not have enough degrees of freedom to accommodate the complexity present in the real data (Figure 3—source data 7). In particular, we found that WR of the published model with three admixture events is 23.9 SE (Figure 3—source data 7a).

For this reason, we moved to topology searches in more complex model spaces incorporating six to nine admixture events. Temporally plausible models with even a modest fit (WR between 3 and 4 SE) were encountered only among models with eight and nine admixture events (Figure 3—source data 7j–r). Librado et al., 2021 discussed five inferences relying fully or partially on their published AGs reported in that study (Table 2). The simplest temporally plausible and best-fitting (WR = 3.4 SE) model we found (eight admixture events, see Figure 3b and the first model in Figure 3—source data 7j) supports inferences 2 and 4, and is incompatible with inferences 1, 3, and 5 (Table 2). We consider this model to be plausible also from the geographical perspective (see Appendix 2 for an interpretation of this topology). We are not arguing here that this AG represents the true history; in fact, it is highly unlikely to be the truth, given how large the space of all possible admixture events is and how much admixture evidently occurred relating all these groups (which makes finding the true model extremely unlikely, see the results on simulated data in Figure 1 and Appendix 1—figure 2). However, our set of 16 temporally plausible and fitting (WR < 4 SE) models with eight or nine admixture events (Figure 3—source data 7j–r) is consistent with some features of the published graph being stable: the features (2) that DOM2 and C-PONT are sister groups, and (4) that there was a gene flow from a deep-branching ghost group to NEO-ANA (Table 2).

Equally important is our finding that there are plausible models that are inconsistent with other inferences in Librado et al., 2021. (Table 2). For example, 13 of these 16 models are inconsistent with the suggestion that there was no gene flow connecting the CWC group and the cluster maximized in the Western steppe (DOM2, C-PONT, and TURG) (Figure 3—source data 7j–r). In the eight-admixture-event best-fitting AG (Figure 3b, the first model in Figure 3—source data 7j), CWC actually derives appreciable ancestry from the early domestic horse lineage (DOM2) associated with the Sintashta culture to the exclusion of the more distant Yamnaya-associated TURG and C_PONT horses. This scenario presents a parallel to the one observed in humans, with individuals associated with the CWC receiving admixture from Steppe pastoralists albeit in different proportions: ~75% for humans, versus ~20% in horses. These models specifying a substantial Steppe horse contribution to CWC horses are inconsistent with the inference in Librado et al., 2021. that ‘Our results reject the commonly held association between horseback riding and the massive expansion of Yamnaya steppe pastoralists into Europe around 3000 BC’. We are not aware of other lines of evidence in the paper (apart from the fitted AG) that support the claim of no Yamnaya horse impact on CWC horses.

Another example of a feature of the published graph that turned out to be unstable is the model for the Tarpan horse. Only 8 of 16 temporally plausible and fitting models (Figure 3—source data 7j–r) support the conclusion by Librado et al. that the Tarpan is a mixture of a DOM2-related and a CWC-related lineage. The other eight models suggest that Tarpan is a mixture of a deep lineage and a DOM2-related lineage (Figure 3b, the first model in Figure 3—source data 7j), echoing a hypothesis that Tarpan may be a hybrid with Przewalski-related horses not represented in the AG (Librado et al., 2021). Again, we are not arguing here that our alternative model is right—indeed we are nearly certain it is wrong in important aspects—but we are merely pointing out that the complexity of the AG space means that qualitatively quite different conclusions are compatible with the statistics fitted in the published paper.

5. Hajdinjak et al., 2021

The AG inferred by Hajdinjak et al. was constructed manually and incorporated 11 groups and 8 admixture events (Figure 2d in the original study). Most (71.4%) models found with findGraphs fit nominally better, and 15.7% fit significantly better than the published model (Table 1, Supplementary file 1, Figure 2), which has a poor absolute fit on this set of sites and groups (WR = 4.8 SE, Figure 3c, Figure 3—source data 8). The statistics described above and the fact that LL scores on all sites lie outside of the LL distribution on resampled datasets (Figure 2) suggest that models in this complexity class are overfitted, but the published topology emerged as fitting relatively poorly. Overfitting arises naturally during manual graph construction as performed in many studies (not only in Hajdinjak et al., 2021, but also in Fu et al., 2016; Skoglund et al., 2016; Yang et al., 2017; Posth et al., 2018; McColl et al., 2018; Moreno-Mayar et al., 2018; Tambets et al., 2018; van de Loosdrecht et al., 2018; Flegontov et al., 2019; Sikora et al., 2019; Wang et al., 2019; Lipson et al., 2020b; Shinde et al., 2019; Yang et al., 2020; Wang et al., 2021). The graph grew one group at a time, and each newly added group was mapped on to the pre-existing skeleton graph as unadmixed or as a two-way mixture. Another requirement was that all intermediate graphs have good absolute fits (WR below 3 or 4 SE). When the model-building process is constrained in a particular path and fits of all intermediates are required to be good, unnecessary admixture events are often added along the way, and the resulting graph belongs to a complexity class in which models are overfitted.

Hajdinjak et al., 2021’s published graph had three notable features that were interpreted by the authors and used to support some conclusions of the study (Table 2), with the following feature considered the most important: there are gene flows from the lineage found in the ~45,000- to 43,000-year-old Bacho Kiro Initial Upper Paleolithic (IUP) individuals to the Ust’-Ishim, Tianyuan, and GoyetQ116-1 lineages. We identified 1,421 topologies fitting nominally or significantly better than the published model and moved on to inspect 50 best-fitting topologies for temporal plausibility (all of them fitting significantly better than the published model). All non-African individuals included in the model are Upper Paleolithic and their dates are not drastically different in relative terms, from ca. 45 to 30 kya (1,000 years before present). Nevertheless, we considered most gene flows from later- to earlier-attested lineages as temporally implausible (for instance, GoyetQ116-1 (~35 kya) → Ust’-Ishim (~44 kya), Sunghir III (34.5 kya) → Tianyuan (40 kya), etc.) since they imply great antiquity of the later-attested lineages and of all lineages derived from them at least partially.

Of the 50 topologies inspected, 32 were considered temporally plausible. Of those topologies, none supported the finding of gene flows from the Bacho Kiro IUP lineage specifically into all three of the Ust’-Ishim, Tianyuan, and GoyetQ116-1 lineages. A total of 17 topologies supported features 2 and 3 but were inconsistent with feature 1; and 14 topologies supported feature 3 only (Table 2). Best-fitting representatives of each of these topology classes are shown along with the published model in Figure 3—source data 8. Considering topological diversity among models that are temporally plausible, conform to current knowledge about relationships between modern and archaic humans, and fit significantly better than the published model, we conclude that feature 3 is probably robust but other details of the fitted AG in Hajdinjak et al.—for example, gene flows to the Ust’-Ishim, Tianyuan, and Goyet Q116-1 lineages from sources sharing drift exclusively with the Upper Paleolithic Bacho Kiro lineage—should not be interpreted as providing meaningful inferences about population history of Upper Paleolithic modern humans.

A central finding of Hajdinjak et al. is that the Bacho Kiro IUP group shares more alleles with present-day East Asians than with Upper Paleolithic Holocene Europeans despite coming from Europe. Specifically, the study documents significantly positive statistics of the form D(an Asian group, Kostenki14; Bacho Kiro IUP, Mbuti). Hajdinjak et al.’s interpretation of this observation is that ‘there was at least some continuity between the earliest modern humans in Europe [Bacho Kiro IUP] and later people in Eurasia [East Asians]’. However, a significant D- or f₄-statistic can have multiple explanations. The statistic f₄ (Tianyuan, Kostenki14; Bacho Kiro IUP, Mbuti) is fitted equally well by the published 12-population AG (Z-score for the difference between the observed and fitted statistics = 0.64) and by, for example, the AG in Figure 3c (Z-score = 0.94). Under the latter model that fits the data significantly better than the published model (p-value = 0.02), the Bacho Kiro IUP and Tianyuan branches are not connected by a gene flow and do not receive gene flows from a third common source, but the common ancestor of Ust’-Ishim and all European Paleolithic lineages receives an 8% gene flow from a divergent modern human lineage splitting deeper than Bacho Kiro IUP and Tianyuan (Figure 3c, Figure 3—source data 8c). This scenario or some version of it seems archaeologically and geographically plausible and is not disproven by any other line of genetic or non-genetic evidence of which we are aware. It could correspond to a scenario where a primary modern human expansion out of West Asia contributed serially to the major lineages leading to Bacho Kiro, then later East Asians, then Ust’-Ishim, and finally the primary ancestry in later European hunter–gatherers. This has a very different interpretation from the scenario of distinctive shared ancestry between the earliest modern humans in Europe such as Bacho Kiro IUP and later people in East Asia—to the exclusion of later European hunter–gatherers—that is suggested by the Hajdinjak et al. published graph.

We are not claiming that this specific alternative model is correct—indeed, it is almost certainly not the correct one given the topological complexity of the set of all AGs consistent with the data—but the existence of it and many other models that fit the data makes it clear that we do not yet have a unique historical explanation for the excess sharing of alleles that has been documented between some Upper Paleolithic European groups (Bacho Kiro IUP, Hajdinjak et al., 2021, and GoyetQ116-1, Yang et al., 2017 and Hajdinjak et al., 2021) and all East Asians.

6. Lipson et al., 2020b

The AG in the original study was constructed manually (Extended Data Figure 4 in that study) and is very complex (12 groups and 12 admixture events): it exists in a space of ~10⁴⁴ topologies of this complexity. We note that one admixture event was added by Lipson et al., 2020b to account for potential modern DNA contamination in ancient Shum Laka individuals, and removing it caused a negligible difference in the fit of the published model (Supplementary file 1). Thus, to decrease the complexity of the graph search space, we considered graphs with 12 groups and 11 admixture events. Among 2,000 newly found models, 11.9% fit nominally (but not significantly) better than the published model (Table 1, Supplementary file 1, Figure 2), and absolute fits of 36.7% of novel models are good (WR <3 SE). These metrics, along with the fact that LL scores on all sites lie outside of the LL distribution on resampled datasets (Figure 2), suggest that models in this complexity class, including the published model, are overfitted. Of the AGs we re-evaluate in this study, the graph from Lipson et al., 2020b shares with the graphs from Hajdinjak et al., 2021; Sikora et al., 2019; Wang et al., 2021, evidence of being overfitted (Figure 2).

Figure 4

Download asset Open asset

Figure 4—source data 1

The published admixture graph from Lipson et al., 2020b and alternative graphs found with findGraphs (12 populations, 11 admixture events) using the updated algorithm for calculating f-statistics.

The graphs were also re-fitted using the original algorithm for calculating f-statistics. Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a, b) The published model and its simplified representation. Below we list selected claims by Lipson et al., 2020b relying on the admixture graph: (1) A lineage maximized in present-day West African groups (Lemande, Mende, and Yoruba) also contributed some ancestry to the ancient Shum Laka individual, and present-day Biaka and Mbuti; (2) another ancestry component in Shum Laka is a deep-branching lineage maximized in rainforest hunter–gatherers Biaka and Mbuti; (3) ‘super-archaic’ ancestry (i.e., diverging at the modern human/Neanderthal split point or deeper) contributed to Biaka, Shum Laka, Mbuti, Lemande, Mende, and Yoruba; (4) a ghost modern human lineage (or lineages) contributed to Agaw, Mota, Biaka, Shum Laka, Mbuti, Lemande, Mende, and Yoruba. (c-q) Selected alternative models fitting nominally better than the published one.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig4-data1-v2.pdf

Download elife-85492-fig4-data1-v2.pdf

Figure 4—source data 2

The published admixture graph from Wang et al., 2021 and alternative graphs found with findGraphs (12 populations, 8 admixture events) using the updated algorithm for calculating f-statistics.

The graphs were also re-fitted using the original algorithm for calculating f-statistics. Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a, b) The published model and its simplified representation. The following claim in Wang et al., 2021 relies on the admixture graph: There is Onge-related admixture in the Jomon (Japan_HG_Jomon), Tibetan (Nepal_Chokhopani_SG), Upper Yellow River Late Neolithic (China_Upper_YR_LN), West Liao River Late Neolithic (China_WLR_LN), Taiwan Iron Age (Taiwan_IA), and China Island Early Neolithic (Liangdao, China_Island_EN) lineages. (c-l) Alternative models fitting significantly better than the published one.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig4-data2-v2.pdf

Download elife-85492-fig4-data2-v2.pdf

Figure 4—source data 3

The simplified published admixture graph for West Eurasian groups from Sikora et al., 2019 and alternative graphs found with findGraphs (13 populations, 6 admixture events).

Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The published model, with chimpanzee added as an outgroup and four low-level gene flows from the Neanderthal lineage dropped. The following claim in Sikora et al., 2019 relies on the admixture graph: The Mal'ta (MA1_ANE) lineage receives a gene flow from the Caucasus hunter-gatherer (CaucasusHG_LP) lineage. (b-h) Models fitting nominally better than the published one and not supporting the claim, (i) models fitting nominally better than the published one and supporting the claim.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig4-data3-v2.pdf

Download elife-85492-fig4-data3-v2.pdf

Figure 4—source data 4

The simplified published admixture graph for East Eurasian groups from Sikora et al., 2019 and alternative graphs found with findGraphs (14 populations, 6 admixture events).

Model parameters (graph edges) that were inferred to be unidentifiable are plotted in red. (a) The published model, with chimpanzee added as an outgroup and four low-level gene flows from the Neanderthal lineage dropped. (b) The published model simplified by dropping unidentifiable edges. The following claims in Sikora et al., 2019 rely on the admixture graph: (1) The Mal’ta (MA1_ANE) and Yana (Yana_UP) lineages receive a gene flow from an Asian source diverging before the Devil’s Cave (DevilsCave_N), Kolyma (Kolyma_M), USR1 (Alaska_LP), and Clovis (Clovis_LP) lineages; (2) European ancestry in the Kolyma, USR1, and Clovis lineages is closer to Mal’ta than to Yana; (3) The Devil’s Cave lineage receives no European-related gene flows, and Kolyma has less European-related ancestry than ancient Americans (USR1 and Clovis). (c) An alternative model fitting nominally better than the published one and supporting all three claims, (d-g) alternative models fitting not significantly worse than the published model.

https://cdn.elifesciences.org/articles/85492/elife-85492-fig4-data4-v2.pdf

Download elife-85492-fig4-data4-v2.pdf

Below we discuss four prominent features of the AG published in the original study (that were interpreted by the authors and used to support some conclusions of the study) and the extent to which these features consistently replicate across the large number of fitting 12-population graphs with 11 admixture events (Table 2). High topological diversity is observed among temporally plausible newly found AGs (see an example in Figure 4a and further topologies in Figure 4—source data 1). Considering extreme cases, two AGs completely lacked support for three features of the published graph (Figure 4a, Figure 4—source data 1c), and one graph supported all four features of the published graph fully (Figure 4—source data 1q, the second model). There are some graphs where defining two distinct ancestral lineages maximized in West Africans and in Mbuti and Biaka (features 1 and 2, Table 2) is essentially impossible since all or nearly all Africans are modeled as a mixture of at least two deep lineages (see alternative graph no. 4 shown in Figure 4—source data 1d, the second model). In some graphs there is no single lineage specific to rainforest hunter–gatherers (Biaka, Mbuti, and Shum Laka) since the primary ancestries in these groups form independent deep branches in the African graph (see Figure 4a and graph no. 16 shown in Figure 4—source data 1j, the second model). The ghost modern and super-archaic gene flows to Africans also had no universal support in the set of alternative graphs we examined (see, for example, Figure 4a and Figure 4—source data 1c).

Considering the high degree of topological diversity among models that are temporally plausible, conform to known findings about relationships between modern and archaic humans, and fit nominally better than the published model, we conclude that none of the four AG features from the original study are consistently supported by our re-analysis (Table 2). This situation may be attributed to (1) overfitting and/or to (2) the lack of information in the dataset (in the combination of groups and SNP sites) and/or to (3) inherent limitations of f-statistics, when distinct topologies predict identical f-statistics. Our results highlight the mystery around the highly distinctive genetic ancestry of the Shum Laka individuals themselves, who represent the newly reported data in the Lipson et al., 2020b study, and represent a highly important set of genetic datapoints that was not available prior to the study. The ancestral relationships of these four individuals to rainforest hunter–gatherers, and to the primary lineage in present-day West Africans, remains an open question, one whose resolution promises meaningful new insights into modern human population history.

7. Wang et al., 2021

The AG inferred by Wang et al., 2021 was constructed manually, and the final graph (Extended Data Figure 6 in the original study) included 12 groups and 8 admixture events. We applied several constraints on the graph space exploration process all of which were shared with the Wang et al. graphs (Supplementary file 1). An important feature of the published graphs in Wang et al., 2021 that was remarked upon in the study is admixture from a source related to Andamanese hunter–gatherers that is almost universal in East Asians (Table 2). For example, the abstract states ‘Hunter-gatherers from Japan, the Amur River Basin, and people of Neolithic and Iron Age Taiwan and the Tibetan Plateau are linked by a deeply splitting lineage that probably reflects a coastal migration during the Late Pleistocene epoch.’ We performed 2,000 findGraphs iterations and obtained 1,778 distinct topologies satisfying all the constraints, nearly all of them (1,724) fitting nominally better than the published model, and 12.6% fitting significantly better (Table 1, Supplementary file 1). The models were ranked by LL scores, and 56 highest-ranking topologies, all of them fitting significantly better than the published one, were assessed for temporal plausibility, and 20 topologies were considered temporally plausible (all of them are shown in Figure 4—source data 2). According to these topologies, 0–2 East Asian groups had a fraction of their ancestry derived from a source specifically related to Onge, and 19 topologies included gene flows from the European (Loschbour)-related branch to all 8 East Asian groups (Figure 4—source data 2). The inferred topological relationships among East Asians are variable in this group of AGs, and we decided to apply further constraints that guided model ranking and elimination by Wang et al., based on considerations from archaeological evidence, Y chromosome haplogroup divergence patterns, and population split time estimation (see Appendix 2 for details). Applying these three additional constraints, we identified two models (among the 56 subjected to manual inspection) that satisfied all of them. The highestranking of those models is shown in Figure 4b and Figure 4—source data 2c (the second model), and it includes a 13% (deeply) European-related gene flow to the common ancestor of all East Asians, and gene flows from the Onge-related branch to just two East Asian groups: Nepal Chokhopani and China WLR LN. This model fits the data significantly better than the published model (p-value = 0.028). We do not claim that this is the correct model (indeed we are almost certain that it is not given the high topological diversity of fitting models), but it is not obviously wrong and differs in qualitatively important ways from the published one.

The Wang et al., 2021 AG provides an illuminating example that helps us to understand the value added by AG construction. The AG construction process in Wang et al. followed a philosophy of not relying entirely on the allele frequency correlation data (not treating the genetic data as independent to explore how much new insight could come from genetic data alone). Instead, the study integrated other lines of genetic evidence as well as linguistic and archaeological insights explicitly into the AG construction process, with the goal of identifying models consistent with multiple lines of evidence. The fact that after this procedure a fitting graph was obtained is not of great interest, as it is essentially always possible to obtain a fit to allele frequency correlation data when enough admixture events are added. The important question is whether any of the emergent features of the graph that were not applied as constraints in the construction process—for example the evidence of ubiquitous Andamanese-related gene flow throughout East Asia suggesting a coastal route expansion that admixed with an interior route expansion proxied by Tianyuan—were stably inferred. Our analysis does not come to this finding consistently among well-fitting and plausible AGs. We conclude that this important feature of the published graph is not supported by f-statistic analysis alone (Table 2), and indeed we are not aware of a single feature of the Wang et al., 2021 AG that is stably inferred beyond the constraints applied to build it.

8. Sikora et al., 2019

Two AGs inferred by Sikora et al., 2019 were constructed manually based on an SNP set derived from whole-genome shotgun data and incorporated 12 or 13 groups and 10 admixture events (Extended Data Figure 3f in the original study). One graph was focused on West Eurasians, and the other one on East Eurasians, and both included a Neanderthal, a Denisovan, and an African group (Dinka). Although the chimpanzee outgroup was not included in the original graphs, we added it as it drastically constrains the topology search space. In contrast to most other published graphs discussed above, gene flows in the graphs inferred by Sikora et al. do not have equal standing: four low-level gene flows (0–1%) connect the Neanderthal lineage to Upper Paleolithic lineages. We repeated each topology search under two alternative settings: either keeping the number of admixture events at 10 to match the published graphs, or at 6 to match simplified versions of the published graphs lacking these low-level Neanderthal gene flows. We performed that modification to simplify the search space and to alleviate the overfitting problem which becomes severe if 10 gene flows across the graph are allowed (Supplementary file 1).

In the case of the "Western" graphs with 6 admixture events, 1,000 topology search iterations were performed, 894 distinct topologies were found, 4 models fit significantly better, and 151 models fit nominally better than the published one (Table 1, Supplementary file 1). We inspected those 155 topologies and identified 29 topologies (Figure 4—source data 3) that are temporally plausible. Sikora et al. came to the following striking conclusion relying on the "Western" AG (Table 2): the Mal’ta (MA1_ANE) lineage received a gene flow from the Caucasus hunter–gatherer (CHG) lineage. However, in our findGraphs exploration this direction of gene flow (CHG → Mal’ta) was supported by two of the 29 topologies, and the opposite gene flow direction (from the Mal’ta and East European hunter–gatherer lineages to CHG) was supported by the remaining 27 plausible topologies (Figure 4—source data 3). The highest-ranking plausible topology (Figure 4c) has a fit that is not significantly different from that of the simplified published model with six admixture events (p-value = 0.392). We note that the gene flow direction contradicting the graph by Sikora et al. was supported by published qpAdm analyses (Lazaridis et al., 2016; Narasimhan et al., 2019), and qpAdm is not affected by the same model degeneracy issues that are the focus of this study. Considering the topological diversity among models that are temporally plausible, conform to robust findings about relationships between modern and archaic humans, and fit nominally better than the published model, we conclude that the direction of the Mal’ta-CHG gene flow cannot be resolved by AG analysis (Table 2).

Some important conclusions based on the "Eastern" graph also do not replicate across all plausible AGs (Table 2). In the case of the "Eastern" graphs with 6 admixture events, 4,446 topology search iterations were performed, and 2,785 distinct topologies were found. Only 3 topologies fit significantly and 13 nominally better than the published one, and 9.8% of topologies fit not significantly worse than the published one (Table 1, Supplementary file 1). Of the AGs belonging to these groups, we inspected 116 best-fitting ones and identified 97 AGs that are temporally plausible. The Sikora et al. "Eastern" AG had three distinctive features that were used to support some conclusions of the study (Table 2). Only feature 2 was universally supported by all the 97 plausible alternative models fitting significantly better, nominally better, or not significantly worse than the simplified published model, while feature 3 was supported by 83 of 97 plausible models, and feature 1 was supported by 28 of 97 plausible models (Table 2). We plotted 14 plausible graphs as examples of topologies supporting all three features, two features, or one feature of the published graph (Figure 4—source data 4). We note that all the "Eastern" graphs discussed here, both the published and alternative ones, have relatively poor absolute fits with WR above 4 or 5 SE. Increasing the number of gene flows to 10 allowed us to reach much better absolute fits (with WR as low as 2.42 SE), but that resulted in high topological diversity (on a par with some other case studies discussed above).

The new software presented in this manuscript (the ADMIXTOOLS 2 R package) is freely available at https://github.com/uqrmaie1/admixtools (copy archived at Maier et al., 2022), along with a detailed manual at https://uqrmaie1.github.io/admixtools/. The ancient human genome newly reported in this manuscript (Supplementary file 2) is freely available at the European Nucleotide Archive in the form of an alignment of reads to the hg19 human reference genome (project accession number PRJEB58199). Published software packages re-used in this manuscript are available at: https://bitbucket.org/nygcresearch/treemix/src/master/ (TreeMix, Pickrell and Pritchard, 2012) and at https://github.com/DReichLab/AdmixTools (David Reich Lab, 2023, ADMIXTOOLS, Patterson et al., 2012). Published archaeogenetic datasets re-analyzed in this manuscript were kindly shared by the corresponding authors of the following publications upon our requests: Bergström et al., 2020; Lazaridis et al., 2014; Librado et al., 2021; Lipson et al., 2020b; Shinde et al., 2019; Sikora et al., 2019; Wang et al., 2021; Hajdinjak et al., 2021. Various statistics for these re-used datasets are summarized in Supplementary file 1.

1. 1. Maier R
   2. Flegontov P
   3. Flegontova O
   4. Işıldak U
   5. Changmai P
   6. Reich D

   (2022) European Nucleotide Archive

   ID PRJEB58199. On the limits of fitting complex models of population history to f-statistics.

   https://www.ebi.ac.uk/ena/browser/view/PRJEB58199

1. 1. Baumdicker F
   2. Bisschop G
   3. Goldstein D
   4. Gower G
   5. Ragsdale AP
   6. Tsambos G
   7. Zhu S
   8. Eldon B
   9. Ellerman EC
   10. Galloway JG
   11. Gladstein AL
   12. Gorjanc G
   13. Guo B
   14. Jeffery B
   15. Kretzschumar WW
   16. Lohse K
   17. Matschiner M
   18. Nelson D
   19. Pope NS
   20. Quinto-Cortés CD
   21. Rodrigues MF
   22. Saunack K
   23. Sellinger T
   24. Thornton K
   25. van Kemenade H
   26. Wohns AW
   27. Wong Y
   28. Gravel S
   29. Kern AD
   30. Koskela J
   31. Ralph PL
   32. Kelleher J

   (2022) Efficient ancestry and mutation simulation with msprime 1.0

   Genetics 220:iyab229.

   https://doi.org/10.1093/genetics/iyab229

   • PubMed
   • Google Scholar
2. 1. Bellwood P

   (2011) The checkered prehistory of rice movement southwards as a domesticated cereal—from the yangzi to the equator

   Rice 4:93–103.

   https://doi.org/10.1007/s12284-011-9068-9

   • Google Scholar
3. 1. Bergström A
   2. Frantz L
   3. Schmidt R
   4. Ersmark E
   5. Lebrasseur O
   6. Girdland-Flink L
   7. Lin AT
   8. Storå J
   9. Sjögren KG
   10. Anthony D
   11. Antipina E
   12. Amiri S
   13. Bar-Oz G
   14. Bazaliiskii VI
   15. Bulatović J
   16. Brown D
   17. Carmagnini A
   18. Davy T
   19. Fedorov S
   20. Fiore I
   21. Fulton D
   22. Germonpré M
   23. Haile J
   24. Irving-Pease EK
   25. Jamieson A
   26. Janssens L
   27. Kirillova I
   28. Horwitz LK
   29. Kuzmanovic-Cvetković J
   30. Kuzmin Y
   31. Losey RJ
   32. Dizdar DL
   33. Mashkour M
   34. Novak M
   35. Onar V
   36. Orton D
   37. Pasarić M
   38. Radivojević M
   39. Rajković D
   40. Roberts B
   41. Ryan H
   42. Sablin M
   43. Shidlovskiy F
   44. Stojanović I
   45. Tagliacozzo A
   46. Trantalidou K
   47. Ullén I
   48. Villaluenga A
   49. Wapnish P
   50. Dobney K
   51. Götherström A
   52. Linderholm A
   53. Dalén L
   54. Pinhasi R
   55. Larson G
   56. Skoglund P

   (2020) Origins and genetic legacy of prehistoric dogs

   Science 370:557–564.

   https://doi.org/10.1126/science.aba9572

   • PubMed
   • Google Scholar
4. 1. Bergström A
   2. Stanton DWG
   3. Taron UH
   4. Frantz L
   5. Sinding M-HS
   6. Ersmark E
   7. Pfrengle S
   8. Cassatt-Johnstone M
   9. Lebrasseur O
   10. Girdland-Flink L
   11. Fernandes DM
   12. Ollivier M
   13. Speidel L
   14. Gopalakrishnan S
   15. Westbury MV
   16. Ramos-Madrigal J
   17. Feuerborn TR
   18. Reiter E
   19. Gretzinger J
   20. Münzel SC
   21. Swali P
   22. Conard NJ
   23. Carøe C
   24. Haile J
   25. Linderholm A
   26. Androsov S
   27. Barnes I
   28. Baumann C
   29. Benecke N
   30. Bocherens H
   31. Brace S
   32. Carden RF
   33. Drucker DG
   34. Fedorov S
   35. Gasparik M
   36. Germonpré M
   37. Grigoriev S
   38. Groves P
   39. Hertwig ST
   40. Ivanova VV
   41. Janssens L
   42. Jennings RP
   43. Kasparov AK
   44. Kirillova IV
   45. Kurmaniyazov I
   46. Kuzmin YV
   47. Kosintsev PA
   48. Lázničková-Galetová M
   49. Leduc C
   50. Nikolskiy P
   51. Nussbaumer M
   52. O’Drisceoil C
   53. Orlando L
   54. Outram A
   55. Pavlova EY
   56. Perri AR
   57. Pilot M
   58. Pitulko VV
   59. Plotnikov VV
   60. Protopopov AV
   61. Rehazek A
   62. Sablin M
   63. Seguin-Orlando A
   64. Storå J
   65. Verjux C
   66. Zaibert VF
   67. Zazula G
   68. Crombé P
   69. Hansen AJ
   70. Willerslev E
   71. Leonard JA
   72. Götherström A
   73. Pinhasi R
   74. Schuenemann VJ
   75. Hofreiter M
   76. Gilbert MTP
   77. Shapiro B
   78. Larson G
   79. Krause J
   80. Dalén L
   81. Skoglund P

   (2022) Grey wolf genomic history reveals a dual ancestry of dogs

   Nature 607:313–320.

   https://doi.org/10.1038/s41586-022-04824-9

   • Google Scholar
5. 1. Boos DD

   (2003) Introduction to the bootstrap world

   Statistical Science 18:168–174.

   https://doi.org/10.1214/ss/1063994971

   • Google Scholar
6. 1. Border R
   2. Johnson EC
   3. Evans LM
   4. Smolen A
   5. Berley N
   6. Sullivan PF
   7. Keller MC

   (2019) No support for historical candidate gene or candidate gene-by-interaction hypotheses for major depression across multiple large samples

   The American Journal of Psychiatry 176:376–387.

   https://doi.org/10.1176/appi.ajp.2018.18070881

   • PubMed
   • Google Scholar
7. 1. Campbell MC
   2. Tishkoff SA

   (2008) African genetic diversity: implications for human demographic history, modern human origins, and complex disease mapping

   Annual Review of Genomics and Human Genetics 9:403–433.

   https://doi.org/10.1146/annurev.genom.9.081307.164258

   • Google Scholar
8. 1. Carlhoff S
   2. Duli A
   3. Nägele K
   4. Nur M
   5. Skov L
   6. Sumantri I
   7. Oktaviana AA
   8. Hakim B
   9. Burhan B
   10. Syahdar FA
   11. McGahan DP
   12. Bulbeck D
   13. Perston YL
   14. Newman K
   15. Saiful AM
   16. Ririmasse M
   17. Chia S
   18. Pulubuhu DAT
   19. Jeong C
   20. Peter BM
   21. Prüfer K
   22. Powell A
   23. Krause J
   24. Posth C
   25. Brumm A

   (2021) Genome of a middle Holocene Hunter-Gatherer from Wallacea

   Nature 596:543–547.

   https://doi.org/10.1038/s41586-021-03823-6

   • PubMed
   • Google Scholar
9. 1. Chen FH
   2. Dong GH
   3. Zhang DJ
   4. Liu XY
   5. Jia X
   6. An CB
   7. Ma MM
   8. Xie YW
   9. Barton L
   10. Ren XY
   11. Zhao ZJ
   12. Wu XH
   13. Jones MK

   (2015) Agriculture facilitated permanent human occupation of the Tibetan Plateau after 3600 B.P

   Science 347:248–250.

   https://doi.org/10.1126/science.1259172

   • PubMed
   • Google Scholar
10. 1. Collins AL
    2. Kim Y
    3. Sklar P
    4. O’Donovan MC
    5. Sullivan PF
    6. International Schizophrenia Consortium

    (2012) Hypothesis-driven candidate genes for schizophrenia compared to genome-wide association results

    Psychological Medicine 42:607–616.

    https://doi.org/10.1017/S0033291711001607

    • PubMed
    • Google Scholar
11. Software

    1. David Reich Lab

    (2023) Admixtools, version ed1d857

    GitHub.

    https://github.com/DReichLab/AdmixTools
12. 1. Duncan LE
    2. Ostacher M
    3. Ballon J

    (2019) How genome-wide association studies (GWAS) made traditional candidate gene studies obsolete

    Neuropsychopharmacology 44:1518–1523.

    https://doi.org/10.1038/s41386-019-0389-5

    • PubMed
    • Google Scholar
13. 1. Excoffier L
    2. Dupanloup I
    3. Huerta-Sánchez E
    4. Sousa VC
    5. Foll M

    (2013) Robust demographic inference from genomic and SNP data

    PLOS Genetics 9:e1003905.

    https://doi.org/10.1371/journal.pgen.1003905

    • PubMed
    • Google Scholar
14. 1. Flegontov P
    2. Altınışık NE
    3. Changmai P
    4. Rohland N
    5. Mallick S
    6. Adamski N
    7. Bolnick DA
    8. Broomandkhoshbacht N
    9. Candilio F
    10. Culleton BJ
    11. Flegontova O
    12. Friesen TM
    13. Jeong C
    14. Harper TK
    15. Keating D
    16. Kennett DJ
    17. Kim AM
    18. Lamnidis TC
    19. Lawson AM
    20. Olalde I
    21. Oppenheimer J
    22. Potter BA
    23. Raff J
    24. Sattler RA
    25. Skoglund P
    26. Stewardson K
    27. Vajda EJ
    28. Vasilyev S
    29. Veselovskaya E
    30. Hayes MG
    31. O’Rourke DH
    32. Krause J
    33. Pinhasi R
    34. Reich D
    35. Schiffels S

    (2019) Palaeo-eskimo genetic ancestry and the peopling of chukotka and North America

    Nature 570:236–240.

    https://doi.org/10.1038/s41586-019-1251-y

    • PubMed
    • Google Scholar
15. 1. Fu Q
    2. Posth C
    3. Hajdinjak M
    4. Petr M
    5. Mallick S
    6. Fernandes D
    7. Furtwängler A
    8. Haak W
    9. Meyer M
    10. Mittnik A
    11. Nickel B
    12. Peltzer A
    13. Rohland N
    14. Slon V
    15. Talamo S
    16. Lazaridis I
    17. Lipson M
    18. Mathieson I
    19. Schiffels S
    20. Skoglund P
    21. Derevianko AP
    22. Drozdov N
    23. Slavinsky V
    24. Tsybankov A
    25. Cremonesi RG
    26. Mallegni F
    27. Gély B
    28. Vacca E
    29. Morales MRG
    30. Straus LG
    31. Neugebauer-Maresch C
    32. Teschler-Nicola M
    33. Constantin S
    34. Moldovan OT
    35. Benazzi S
    36. Peresani M
    37. Coppola D
    38. Lari M
    39. Ricci S
    40. Ronchitelli A
    41. Valentin F
    42. Thevenet C
    43. Wehrberger K
    44. Grigorescu D
    45. Rougier H
    46. Crevecoeur I
    47. Flas D
    48. Semal P
    49. Mannino MA
    50. Cupillard C
    51. Bocherens H
    52. Conard NJ
    53. Harvati K
    54. Moiseyev V
    55. Drucker DG
    56. Svoboda J
    57. Richards MP
    58. Caramelli D
    59. Pinhasi R
    60. Kelso J
    61. Patterson N
    62. Krause J
    63. Pääbo S
    64. Reich D

    (2016) The genetic history of ice age Europe

    Nature 534:200–205.

    https://doi.org/10.1038/nature17993

    • PubMed
    • Google Scholar
16. 1. Gray RD
    2. Jordan FM

    (2000) Language trees support the express-train sequence of austronesian expansion

    Nature 405:1052–1055.

    https://doi.org/10.1038/35016575

    • PubMed
    • Google Scholar
17. 1. Green RE
    2. Krause J
    3. Briggs AW
    4. Maricic T
    5. Stenzel U
    6. Kircher M
    7. Patterson N
    8. Li H
    9. Zhai W
    10. Fritz MH-Y
    11. Hansen NF
    12. Durand EY
    13. Malaspinas A-S
    14. Jensen JD
    15. Marques-Bonet T
    16. Alkan C
    17. Prüfer K
    18. Meyer M
    19. Burbano HA
    20. Good JM
    21. Schultz R
    22. Aximu-Petri A
    23. Butthof A
    24. Höber B
    25. Höffner B
    26. Siegemund M
    27. Weihmann A
    28. Nusbaum C
    29. Lander ES
    30. Russ C
    31. Novod N
    32. Affourtit J
    33. Egholm M
    34. Verna C
    35. Rudan P
    36. Brajkovic D
    37. Kucan Ž
    38. Gušic I
    39. Doronichev VB
    40. Golovanova LV
    41. Lalueza-Fox C
    42. de la Rasilla M
    43. Fortea J
    44. Rosas A
    45. Schmitz RW
    46. Johnson PLF
    47. Eichler EE
    48. Falush D
    49. Birney E
    50. Mullikin JC
    51. Slatkin M
    52. Nielsen R
    53. Kelso J
    54. Lachmann M
    55. Reich D
    56. Pääbo S

    (2010) A draft sequence of the Neandertal genome

    Science 328:710–722.

    https://doi.org/10.1126/science.1188021

    • Google Scholar
18. 1. Gronau I
    2. Hubisz MJ
    3. Gulko B
    4. Danko CG
    5. Siepel A

    (2011) Bayesian inference of ancient human demography from individual genome sequences

    Nature Genetics 43:1031–1034.

    https://doi.org/10.1038/ng.937

    • PubMed
    • Google Scholar
19. 1. Gutenkunst RN
    2. Hernandez RD
    3. Williamson SH
    4. Bustamante CD

    (2009) Inferring the joint demographic history of multiple populations from multidimensional SNP frequency data

    PLOS Genetics 5:e1000695.

    https://doi.org/10.1371/journal.pgen.1000695

    • PubMed
    • Google Scholar
20. 1. Haak W
    2. Lazaridis I
    3. Patterson N
    4. Rohland N
    5. Mallick S
    6. Llamas B
    7. Brandt G
    8. Nordenfelt S
    9. Harney E
    10. Stewardson K
    11. Fu Q
    12. Mittnik A
    13. Bánffy E
    14. Economou C
    15. Francken M
    16. Friederich S
    17. Pena RG
    18. Hallgren F
    19. Khartanovich V
    20. Khokhlov A
    21. Kunst M
    22. Kuznetsov P
    23. Meller H
    24. Mochalov O
    25. Moiseyev V
    26. Nicklisch N
    27. Pichler SL
    28. Risch R
    29. Rojo Guerra MA
    30. Roth C
    31. Szécsényi-Nagy A
    32. Wahl J
    33. Meyer M
    34. Krause J
    35. Brown D
    36. Anthony D
    37. Cooper A
    38. Alt KW
    39. Reich D

    (2015) Massive migration from the steppe was a source for Indo-European languages in Europe

    Nature 522:207–211.

    https://doi.org/10.1038/nature14317

    • PubMed
    • Google Scholar
21. 1. Hajdinjak M
    2. Mafessoni F
    3. Skov L
    4. Vernot B
    5. Hübner A
    6. Fu Q
    7. Essel E
    8. Nagel S
    9. Nickel B
    10. Richter J
    11. Moldovan OT
    12. Constantin S
    13. Endarova E
    14. Zahariev N
    15. Spasov R
    16. Welker F
    17. Smith GM
    18. Sinet-Mathiot V
    19. Paskulin L
    20. Fewlass H
    21. Talamo S
    22. Rezek Z
    23. Sirakova S
    24. Sirakov N
    25. McPherron SP
    26. Tsanova T
    27. Hublin JJ
    28. Peter BM
    29. Meyer M
    30. Skoglund P
    31. Kelso J
    32. Pääbo S

    (2021) Initial upper Palaeolithic humans in Europe had recent neanderthal ancestry

    Nature 592:253–257.

    https://doi.org/10.1038/s41586-021-03335-3

    • Google Scholar
22. 1. Hammer MF
    2. Woerner AE
    3. Mendez FL
    4. Watkins JC
    5. Wall JD

    (2011) Genetic evidence for archaic admixture in Africa

    PNAS 108:15123–15128.

    https://doi.org/10.1073/pnas.1109300108

    • PubMed
    • Google Scholar
23. 1. Harney É
    2. Patterson N
    3. Reich D
    4. Wakeley J

    (2021) Assessing the performance of qpAdm: a statistical tool for studying population Admixture

    Genetics 217:iyaa045.

    https://doi.org/10.1093/genetics/iyaa045

    • PubMed
    • Google Scholar
24. 1. Hirschhorn JN
    2. Daly MJ

    (2005) Genome-Wide association studies for common diseases and complex traits

    Nature Reviews. Genetics 6:95–108.

    https://doi.org/10.1038/nrg1521

    • PubMed
    • Google Scholar
25. 1. Hubisz M
    2. Siepel A

    (2020) Inference of ancestral recombination graphs using argweaver

    Methods in Molecular Biology 2090:231–266.

    https://doi.org/10.1007/978-1-0716-0199-0_10

    • PubMed
    • Google Scholar
26. 1. Hubisz MJ
    2. Williams AL
    3. Siepel A

    (2020) Mapping gene flow between ancient hominins through demography-aware inference of the ancestral recombination graph

    PLOS Genetics 16:e1008895.

    https://doi.org/10.1371/journal.pgen.1008895

    • PubMed
    • Google Scholar
27. 1. Ioannidis JPA

    (2005) Why most published research findings are false

    PLOS Medicine 2:e124.

    https://doi.org/10.1371/journal.pmed.0020124

    • PubMed
    • Google Scholar
28. 1. Jeong C
    2. Balanovsky O
    3. Lukianova E
    4. Kahbatkyzy N
    5. Flegontov P
    6. Zaporozhchenko V
    7. Immel A
    8. Wang CC
    9. Ixan O
    10. Khussainova E
    11. Bekmanov B
    12. Zaibert V
    13. Lavryashina M
    14. Pocheshkhova E
    15. Yusupov Y
    16. Agdzhoyan A
    17. Koshel S
    18. Bukin A
    19. Nymadawa P
    20. Turdikulova S
    21. Dalimova D
    22. Churnosov M
    23. Skhalyakho R
    24. Daragan D
    25. Bogunov Y
    26. Bogunova A
    27. Shtrunov A
    28. Dubova N
    29. Zhabagin M
    30. Yepiskoposyan L
    31. Churakov V
    32. Pislegin N
    33. Damba L
    34. Saroyants L
    35. Dibirova K
    36. Atramentova L
    37. Utevska O
    38. Idrisov E
    39. Kamenshchikova E
    40. Evseeva I
    41. Metspalu M
    42. Outram AK
    43. Robbeets M
    44. Djansugurova L
    45. Balanovska E
    46. Schiffels S
    47. Haak W
    48. Reich D
    49. Krause J

    (2019) The genetic history of admixture across inner Eurasia

    Nature Ecology & Evolution 3:966–976.

    https://doi.org/10.1038/s41559-019-0878-2

    • PubMed
    • Google Scholar
29. 1. Kamm J
    2. Terhorst J
    3. Durbin R
    4. Song YS

    (2020) Efficiently inferring the demographic history of many populations with allele count data

    Journal of the American Statistical Association 115:1472–1487.

    https://doi.org/10.1080/01621459.2019.1635482

    • PubMed
    • Google Scholar
30. 1. Ko AM-S
    2. Chen C-Y
    3. Fu Q
    4. Delfin F
    5. Li M
    6. Chiu H-L
    7. Stoneking M
    8. Ko Y-C

    (2014) Early Austronesians: into and out of Taiwan

    American Journal of Human Genetics 94:426–436.

    https://doi.org/10.1016/j.ajhg.2014.02.003

    • PubMed
    • Google Scholar
31. 1. Kutanan W
    2. Liu D
    3. Kampuansai J
    4. Srikummool M
    5. Srithawong S
    6. Shoocongdej R
    7. Sangkhano S
    8. Ruangchai S
    9. Pittayaporn P
    10. Arias L
    11. Stoneking M

    (2021) Reconstructing the human genetic history of mainland Southeast Asia: insights from genome-wide data from Thailand and laos

    Molecular Biology and Evolution 38:3459–3477.

    https://doi.org/10.1093/molbev/msab124

    • PubMed
    • Google Scholar
32. 1. Lachance J
    2. Vernot B
    3. Elbers CC
    4. Ferwerda B
    5. Froment A
    6. Bodo JM
    7. Lema G
    8. Fu W
    9. Nyambo TB
    10. Rebbeck TR
    11. Zhang K
    12. Akey JM
    13. Tishkoff SA

    (2012) Evolutionary history and adaptation from high-coverage whole-genome sequences of diverse African hunter-gatherers

    Cell 150:457–469.

    https://doi.org/10.1016/j.cell.2012.07.009

    • PubMed
    • Google Scholar
33. 1. Lazaridis I
    2. Patterson N
    3. Mittnik A
    4. Renaud G
    5. Mallick S
    6. Kirsanow K
    7. Sudmant PH
    8. Schraiber JG
    9. Castellano S
    10. Lipson M
    11. Berger B
    12. Economou C
    13. Bollongino R
    14. Fu Q
    15. Bos KI
    16. Nordenfelt S
    17. Li H
    18. de Filippo C
    19. Prüfer K
    20. Sawyer S
    21. Posth C
    22. Haak W
    23. Hallgren F
    24. Fornander E
    25. Rohland N
    26. Delsate D
    27. Francken M
    28. Guinet JM
    29. Wahl J
    30. Ayodo G
    31. Babiker HA
    32. Bailliet G
    33. Balanovska E
    34. Balanovsky O
    35. Barrantes R
    36. Bedoya G
    37. Ben-Ami H
    38. Bene J
    39. Berrada F
    40. Bravi CM
    41. Brisighelli F
    42. Busby GBJ
    43. Cali F
    44. Churnosov M
    45. Cole DEC
    46. Corach D
    47. Damba L
    48. van Driem G
    49. Dryomov S
    50. Dugoujon JM
    51. Fedorova SA
    52. Gallego Romero I
    53. Gubina M
    54. Hammer M
    55. Henn BM
    56. Hervig T
    57. Hodoglugil U
    58. Jha AR
    59. Karachanak-Yankova S
    60. Khusainova R
    61. Khusnutdinova E
    62. Kittles R
    63. Kivisild T
    64. Klitz W
    65. Kučinskas V
    66. Kushniarevich A
    67. Laredj L
    68. Litvinov S
    69. Loukidis T
    70. Mahley RW
    71. Melegh B
    72. Metspalu E
    73. Molina J
    74. Mountain J
    75. Näkkäläjärvi K
    76. Nesheva D
    77. Nyambo T
    78. Osipova L
    79. Parik J
    80. Platonov F
    81. Posukh O
    82. Romano V
    83. Rothhammer F
    84. Rudan I
    85. Ruizbakiev R
    86. Sahakyan H
    87. Sajantila A
    88. Salas A
    89. Starikovskaya EB
    90. Tarekegn A
    91. Toncheva D
    92. Turdikulova S
    93. Uktveryte I
    94. Utevska O
    95. Vasquez R
    96. Villena M
    97. Voevoda M
    98. Winkler CA
    99. Yepiskoposyan L
    100. Zalloua P
    101. Zemunik T
    102. Cooper A
    103. Capelli C
    104. Thomas MG
    105. Ruiz-Linares A
    106. Tishkoff SA
    107. Singh L
    108. Thangaraj K
    109. Villems R
    110. Comas D
    111. Sukernik R
    112. Metspalu M
    113. Meyer M
    114. Eichler EE
    115. Burger J
    116. Slatkin M
    117. Pääbo S
    118. Kelso J
    119. Reich D
    120. Krause J

    (2014) Ancient human Genomes suggest three ancestral populations for present-day Europeans

    Nature 513:409–413.

    https://doi.org/10.1038/nature13673

    • PubMed
    • Google Scholar
34. 1. Lazaridis I
    2. Nadel D
    3. Rollefson G
    4. Merrett DC
    5. Rohland N
    6. Mallick S
    7. Fernandes D
    8. Novak M
    9. Gamarra B
    10. Sirak K
    11. Connell S
    12. Stewardson K
    13. Harney E
    14. Fu Q
    15. Gonzalez-Fortes G
    16. Jones ER
    17. Roodenberg SA
    18. Lengyel G
    19. Bocquentin F
    20. Gasparian B
    21. Monge JM
    22. Gregg M
    23. Eshed V
    24. Mizrahi AS
    25. Meiklejohn C
    26. Gerritsen F
    27. Bejenaru L
    28. Blüher M
    29. Campbell A
    30. Cavalleri G
    31. Comas D
    32. Froguel P
    33. Gilbert E
    34. Kerr SM
    35. Kovacs P
    36. Krause J
    37. McGettigan D
    38. Merrigan M
    39. Merriwether DA
    40. O’Reilly S
    41. Richards MB
    42. Semino O
    43. Shamoon-Pour M
    44. Stefanescu G
    45. Stumvoll M
    46. Tönjes A
    47. Torroni A
    48. Wilson JF
    49. Yengo L
    50. Hovhannisyan NA
    51. Patterson N
    52. Pinhasi R
    53. Reich D

    (2016) Genomic insights into the origin of farming in the ancient near East

    Nature 536:419–424.

    https://doi.org/10.1038/nature19310

    • PubMed
    • Google Scholar
35. 1. Leppälä K
    2. Nielsen SV
    3. Mailund T

    (2017) Admixturegraph: an R package for admixture graph manipulation and fitting

    Bioinformatics 33:1738–1740.

    https://doi.org/10.1093/bioinformatics/btx048

    • PubMed
    • Google Scholar
36. 1. Librado P
    2. Khan N
    3. Fages A
    4. Kusliy MA
    5. Suchan T
    6. Tonasso-Calvière L
    7. Schiavinato S
    8. Alioglu D
    9. Fromentier A
    10. Perdereau A
    11. Aury J-M
    12. Gaunitz C
    13. Chauvey L
    14. Seguin-Orlando A
    15. Der Sarkissian C
    16. Southon J
    17. Shapiro B
    18. Tishkin AA
    19. Kovalev AA
    20. Alquraishi S
    21. Alfarhan AH
    22. Al-Rasheid KAS
    23. Seregély T
    24. Klassen L
    25. Iversen R
    26. Bignon-Lau O
    27. Bodu P
    28. Olive M
    29. Castel J-C
    30. Boudadi-Maligne M
    31. Alvarez N
    32. Germonpré M
    33. Moskal-Del Hoyo M
    34. Wilczyński J
    35. Pospuła S
    36. Lasota-Kuś A
    37. Tunia K
    38. Nowak M
    39. Rannamäe E
    40. Saarma U
    41. Boeskorov G
    42. Lōugas L
    43. Kyselý R
    44. Peške L
    45. Bălășescu A
    46. Dumitrașcu V
    47. Dobrescu R
    48. Gerber D
    49. Kiss V
    50. Szécsényi-Nagy A
    51. Mende BG
    52. Gallina Z
    53. Somogyi K
    54. Kulcsár G
    55. Gál E
    56. Bendrey R
    57. Allentoft ME
    58. Sirbu G
    59. Dergachev V
    60. Shephard H
    61. Tomadini N
    62. Grouard S
    63. Kasparov A
    64. Basilyan AE
    65. Anisimov MA
    66. Nikolskiy PA
    67. Pavlova EY
    68. Pitulko V
    69. Brem G
    70. Wallner B
    71. Schwall C
    72. Keller M
    73. Kitagawa K
    74. Bessudnov AN
    75. Bessudnov A
    76. Taylor W
    77. Magail J
    78. Gantulga J-O
    79. Bayarsaikhan J
    80. Erdenebaatar D
    81. Tabaldiev K
    82. Mijiddorj E
    83. Boldgiv B
    84. Tsagaan T
    85. Pruvost M
    86. Olsen S
    87. Makarewicz CA
    88. Valenzuela Lamas S
    89. Albizuri Canadell S
    90. Nieto Espinet A
    91. Iborra MP
    92. Lira Garrido J
    93. Rodríguez González E
    94. Celestino S
    95. Olària C
    96. Arsuaga JL
    97. Kotova N
    98. Pryor A
    99. Crabtree P
    100. Zhumatayev R
    101. Toleubaev A
    102. Morgunova NL
    103. Kuznetsova T
    104. Lordkipanize D
    105. Marzullo M
    106. Prato O
    107. Bagnasco Gianni G
    108. Tecchiati U
    109. Clavel B
    110. Lepetz S
    111. Davoudi H
    112. Mashkour M
    113. Berezina NY
    114. Stockhammer PW
    115. Krause J
    116. Haak W
    117. Morales-Muñiz A
    118. Benecke N
    119. Hofreiter M
    120. Ludwig A
    121. Graphodatsky AS
    122. Peters J
    123. Kiryushin KY
    124. Iderkhangai T-O
    125. Bokovenko NA
    126. Vasiliev SK
    127. Seregin NN
    128. Chugunov KV
    129. Plasteeva NA
    130. Baryshnikov GF
    131. Petrova E
    132. Sablin M
    133. Ananyevskaya E
    134. Logvin A
    135. Shevnina I
    136. Logvin V
    137. Kalieva S
    138. Loman V
    139. Kukushkin I
    140. Merz I
    141. Merz V
    142. Sakenov S
    143. Varfolomeyev V
    144. Usmanova E
    145. Zaibert V
    146. Arbuckle B
    147. Belinskiy AB
    148. Kalmykov A
    149. Reinhold S
    150. Hansen S
    151. Yudin AI
    152. Vybornov AA
    153. Epimakhov A
    154. Berezina NS
    155. Roslyakova N
    156. Kosintsev PA
    157. Kuznetsov PF
    158. Anthony D
    159. Kroonen GJ
    160. Kristiansen K
    161. Wincker P
    162. Outram A
    163. Orlando L

    (2021) The origins and spread of domestic horses from the Western Eurasian steppes

    Nature 598:634–640.

    https://doi.org/10.1038/s41586-021-04018-9

    • PubMed
    • Google Scholar
37. 1. Lipson M
    2. Loh PR
    3. Levin A
    4. Reich D
    5. Patterson N
    6. Berger B

    (2013) Efficient moment-based inference of admixture parameters and sources of gene flow

    Molecular Biology and Evolution 30:1788–1802.

    https://doi.org/10.1093/molbev/mst099

    • PubMed
    • Google Scholar
38. 1. Lipson M
    2. Szécsényi-Nagy A
    3. Mallick S
    4. Pósa A
    5. Stégmár B
    6. Keerl V
    7. Rohland N
    8. Stewardson K
    9. Ferry M
    10. Michel M
    11. Oppenheimer J
    12. Broomandkhoshbacht N
    13. Harney E
    14. Nordenfelt S
    15. Llamas B
    16. Gusztáv Mende B
    17. Köhler K
    18. Oross K
    19. Bondár M
    20. Marton T
    21. Osztás A
    22. Jakucs J
    23. Paluch T
    24. Horváth F
    25. Csengeri P
    26. Koós J
    27. Sebők K
    28. Anders A
    29. Raczky P
    30. Regenye J
    31. Barna JP
    32. Fábián S
    33. Serlegi G
    34. Toldi Z
    35. Gyöngyvér Nagy E
    36. Dani J
    37. Molnár E
    38. Pálfi G
    39. Márk L
    40. Melegh B
    41. Bánfai Z
    42. Domboróczki L
    43. Fernández-Eraso J
    44. Antonio Mujika-Alustiza J
    45. Alonso Fernández C
    46. Jiménez Echevarría J
    47. Bollongino R
    48. Orschiedt J
    49. Schierhold K
    50. Meller H
    51. Cooper A
    52. Burger J
    53. Bánffy E
    54. Alt KW
    55. Lalueza-Fox C
    56. Haak W
    57. Reich D

    (2017) Parallel palaeogenomic transects reveal complex genetic history of early European farmers

    Nature 551:368–372.

    https://doi.org/10.1038/nature24476

    • PubMed
    • Google Scholar
39. 1. Lipson M

    (2020a) Applying f4-statistics and admixture graphs: theory and examples

    Molecular Ecology Resources 20:1658–1667.

    https://doi.org/10.1111/1755-0998.13230

    • Google Scholar
40. 1. Lipson M
    2. Ribot I
    3. Mallick S
    4. Rohland N
    5. Olalde I
    6. Adamski N
    7. Broomandkhoshbacht N
    8. Lawson AM
    9. López S
    10. Oppenheimer J
    11. Stewardson K
    12. Asombang RN
    13. Bocherens H
    14. Bradman N
    15. Culleton BJ
    16. Cornelissen E
    17. Crevecoeur I
    18. de Maret P
    19. Fomine FLM
    20. Lavachery P
    21. Mindzie CM
    22. Orban R
    23. Sawchuk E
    24. Semal P
    25. Thomas MG
    26. Van Neer W
    27. Veeramah KR
    28. Kennett DJ
    29. Patterson N
    30. Hellenthal G
    31. Lalueza-Fox C
    32. MacEachern S
    33. Prendergast ME
    34. Reich D

    (2020b) Ancient West African foragers in the context of African population history

    Nature 577:665–670.

    https://doi.org/10.1038/s41586-020-1929-1

    • PubMed
    • Google Scholar
41. 1. Lipson M
    2. Sawchuk EA
    3. Thompson JC
    4. Oppenheimer J
    5. Tryon CA
    6. Ranhorn KL
    7. de Luna KM
    8. Sirak KA
    9. Olalde I
    10. Ambrose SH
    11. Arthur JW
    12. Arthur KJW
    13. Ayodo G
    14. Bertacchi A
    15. Cerezo-Román JI
    16. Culleton BJ
    17. Curtis MC
    18. Davis J
    19. Gidna AO
    20. Hanson A
    21. Kaliba P
    22. Katongo M
    23. Kwekason A
    24. Laird MF
    25. Lewis J
    26. Mabulla AZP
    27. Mapemba F
    28. Morris A
    29. Mudenda G
    30. Mwafulirwa R
    31. Mwangomba D
    32. Ndiema E
    33. Ogola C
    34. Schilt F
    35. Willoughby PR
    36. Wright DK
    37. Zipkin A
    38. Pinhasi R
    39. Kennett DJ
    40. Manthi FK
    41. Rohland N
    42. Patterson N
    43. Reich D
    44. Prendergast ME

    (2022) Ancient DNA and deep population structure in sub-Saharan African foragers

    Nature 603:290–296.

    https://doi.org/10.1038/s41586-022-04430-9

    • PubMed
    • Google Scholar
42. 1. Lu D
    2. Lou H
    3. Yuan K
    4. Wang X
    5. Wang Y
    6. Zhang C
    7. Lu Y
    8. Yang X
    9. Deng L
    10. Zhou Y
    11. Feng Q
    12. Hu Y
    13. Ding Q
    14. Yang Y
    15. Li S
    16. Jin L
    17. Guan Y
    18. Su B
    19. Kang L
    20. Xu S

    (2016) Ancestral origins and genetic history of Tibetan highlanders

    American Journal of Human Genetics 99:580–594.

    https://doi.org/10.1016/j.ajhg.2016.07.002

    • PubMed
    • Google Scholar
43. Software

    1. Maier R
    2. Isildak U
    3. rmmaier

    (2022) Admixtools, version swh:1:rev:45fcc41768d88559863d6dfae9ccaceaf247f81e

    Software Heritage.

    https://archive.softwareheritage.org/swh:1:dir:38fceb428451354c86a1c41fa7291ea117a15d2b;origin=https://github.com/uqrmaie1/admixtools;visit=swh:1:snp:ec5ab16d631d19a1943b1f874db5a4df0cee6ced;anchor=swh:1:rev:45fcc41768d88559863d6dfae9ccaceaf247f81e
44. 1. Mallick S
    2. Li H
    3. Lipson M
    4. Mathieson I
    5. Gymrek M
    6. Racimo F
    7. Zhao M
    8. Chennagiri N
    9. Nordenfelt S
    10. Tandon A
    11. Skoglund P
    12. Lazaridis I
    13. Sankararaman S
    14. Fu Q
    15. Rohland N
    16. Renaud G
    17. Erlich Y
    18. Willems T
    19. Gallo C
    20. Spence JP
    21. Song YS
    22. Poletti G
    23. Balloux F
    24. van Driem G
    25. de Knijff P
    26. Romero IG
    27. Jha AR
    28. Behar DM
    29. Bravi CM
    30. Capelli C
    31. Hervig T
    32. Moreno-Estrada A
    33. Posukh OL
    34. Balanovska E
    35. Balanovsky O
    36. Karachanak-Yankova S
    37. Sahakyan H
    38. Toncheva D
    39. Yepiskoposyan L
    40. Tyler-Smith C
    41. Xue Y
    42. Abdullah MS
    43. Ruiz-Linares A
    44. Beall CM
    45. Di Rienzo A
    46. Jeong C
    47. Starikovskaya EB
    48. Metspalu E
    49. Parik J
    50. Villems R
    51. Henn BM
    52. Hodoglugil U
    53. Mahley R
    54. Sajantila A
    55. Stamatoyannopoulos G
    56. Wee JTS
    57. Khusainova R
    58. Khusnutdinova E
    59. Litvinov S
    60. Ayodo G
    61. Comas D
    62. Hammer MF
    63. Kivisild T
    64. Klitz W
    65. Winkler CA
    66. Labuda D
    67. Bamshad M
    68. Jorde LB
    69. Tishkoff SA
    70. Watkins WS
    71. Metspalu M
    72. Dryomov S
    73. Sukernik R
    74. Singh L
    75. Thangaraj K
    76. Pääbo S
    77. Kelso J
    78. Patterson N
    79. Reich D

    (2016) The simons genome diversity project: 300 genomes from 142 diverse populations

    Nature 538:201–206.

    https://doi.org/10.1038/nature18964

    • Google Scholar
45. 1. Mathieson I
    2. Alpaslan-Roodenberg S
    3. Posth C
    4. Szécsényi-Nagy A
    5. Rohland N
    6. Mallick S
    7. Olalde I
    8. Broomandkhoshbacht N
    9. Candilio F
    10. Cheronet O
    11. Fernandes D
    12. Ferry M
    13. Gamarra B
    14. Fortes GG
    15. Haak W
    16. Harney E
    17. Jones E
    18. Keating D
    19. Krause-Kyora B
    20. Kucukkalipci I
    21. Michel M
    22. Mittnik A
    23. Nägele K
    24. Novak M
    25. Oppenheimer J
    26. Patterson N
    27. Pfrengle S
    28. Sirak K
    29. Stewardson K
    30. Vai S
    31. Alexandrov S
    32. Alt KW
    33. Andreescu R
    34. Antonović D
    35. Ash A
    36. Atanassova N
    37. Bacvarov K
    38. Gusztáv MB
    39. Bocherens H
    40. Bolus M
    41. Boroneanţ A
    42. Boyadzhiev Y
    43. Budnik A
    44. Burmaz J
    45. Chohadzhiev S
    46. Conard NJ
    47. Cottiaux R
    48. Čuka M
    49. Cupillard C
    50. Drucker DG
    51. Elenski N
    52. Francken M
    53. Galabova B
    54. Ganetsovski G
    55. Gély B
    56. Hajdu T
    57. Handzhyiska V
    58. Harvati K
    59. Higham T
    60. Iliev S
    61. Janković I
    62. Karavanić I
    63. Kennett DJ
    64. Komšo D
    65. Kozak A
    66. Labuda D
    67. Lari M
    68. Lazar C
    69. Leppek M
    70. Leshtakov K
    71. Vetro DL
    72. Los D
    73. Lozanov I
    74. Malina M
    75. Martini F
    76. McSweeney K
    77. Meller H
    78. Menđušić M
    79. Mirea P
    80. Moiseyev V
    81. Petrova V
    82. Price TD
    83. Simalcsik A
    84. Sineo L
    85. Šlaus M
    86. Slavchev V
    87. Stanev P
    88. Starović A
    89. Szeniczey T
    90. Talamo S
    91. Teschler-Nicola M
    92. Thevenet C
    93. Valchev I
    94. Valentin F
    95. Vasilyev S
    96. Veljanovska F
    97. Venelinova S
    98. Veselovskaya E
    99. Viola B
    100. Virag C
    101. Zaninović J
    102. Zäuner S
    103. Stockhammer PW
    104. Catalano G
    105. Krauß R
    106. Caramelli D
    107. Zariņa G
    108. Gaydarska B
    109. Lillie M
    110. Nikitin AG
    111. Potekhina I
    112. Papathanasiou A
    113. Borić D
    114. Bonsall C
    115. Krause J
    116. Pinhasi R
    117. Reich D

    (2018) The genomic history of southeastern Europe

    Nature 555:197–203.

    https://doi.org/10.1038/nature25778

    • PubMed
    • Google Scholar
46. 1. McColl H
    2. Racimo F
    3. Vinner L
    4. Demeter F
    5. Gakuhari T
    6. Moreno-Mayar JV
    7. van Driem G
    8. Gram Wilken U
    9. Seguin-Orlando A
    10. de la Fuente Castro C
    11. Wasef S
    12. Shoocongdej R
    13. Souksavatdy V
    14. Sayavongkhamdy T
    15. Saidin MM
    16. Allentoft ME
    17. Sato T
    18. Malaspinas A-S
    19. Aghakhanian FA
    20. Korneliussen T
    21. Prohaska A
    22. Margaryan A
    23. de Barros Damgaard P
    24. Kaewsutthi S
    25. Lertrit P
    26. Nguyen TMH
    27. Hung H-C
    28. Minh Tran T
    29. Nghia Truong H
    30. Nguyen GH
    31. Shahidan S
    32. Wiradnyana K
    33. Matsumae H
    34. Shigehara N
    35. Yoneda M
    36. Ishida H
    37. Masuyama T
    38. Yamada Y
    39. Tajima A
    40. Shibata H
    41. Toyoda A
    42. Hanihara T
    43. Nakagome S
    44. Deviese T
    45. Bacon A-M
    46. Duringer P
    47. Ponche J-L
    48. Shackelford L
    49. Patole-Edoumba E
    50. Nguyen AT
    51. Bellina-Pryce B
    52. Galipaud J-C
    53. Kinaston R
    54. Buckley H
    55. Pottier C
    56. Rasmussen S
    57. Higham T
    58. Foley RA
    59. Lahr MM
    60. Orlando L
    61. Sikora M
    62. Phipps ME
    63. Oota H
    64. Higham C
    65. Lambert DM
    66. Willerslev E

    (2018) The prehistoric peopling of Southeast Asia

    Science 361:88–92.

    https://doi.org/10.1126/science.aat3628

    • PubMed
    • Google Scholar
47. 1. Molloy EK
    2. Durvasula A
    3. Sankararaman S

    (2021) Advancing admixture graph estimation via maximum likelihood network orientation

    Bioinformatics 37:i142–i150.

    https://doi.org/10.1093/bioinformatics/btab267

    • PubMed
    • Google Scholar
48. 1. Moreno-Mayar JV
    2. Vinner L
    3. de Barros Damgaard P
    4. de la Fuente C
    5. Chan J
    6. Spence JP
    7. Allentoft ME
    8. Vimala T
    9. Racimo F
    10. Pinotti T
    11. Rasmussen S
    12. Margaryan A
    13. Iraeta Orbegozo M
    14. Mylopotamitaki D
    15. Wooller M
    16. Bataille C
    17. Becerra-Valdivia L
    18. Chivall D
    19. Comeskey D
    20. Devièse T
    21. Grayson DK
    22. George L
    23. Harry H
    24. Alexandersen V
    25. Primeau C
    26. Erlandson J
    27. Rodrigues-Carvalho C
    28. Reis S
    29. Bastos MQR
    30. Cybulski J
    31. Vullo C
    32. Morello F
    33. Vilar M
    34. Wells S
    35. Gregersen K
    36. Hansen KL
    37. Lynnerup N
    38. Mirazón Lahr M
    39. Kjær K
    40. Strauss A
    41. Alfonso-Durruty M
    42. Salas A
    43. Schroeder H
    44. Higham T
    45. Malhi RS
    46. Rasic JT
    47. Souza L
    48. Santos FR
    49. Malaspinas AS
    50. Sikora M
    51. Nielsen R
    52. Song YS
    53. Meltzer DJ
    54. Willerslev E

    (2018) Early human dispersals within the Americas

    Science 362:2621.

    https://doi.org/10.1126/science.aav2621

    • Google Scholar
49. 1. Narasimhan VM
    2. Patterson N
    3. Moorjani P
    4. Rohland N
    5. Bernardos R
    6. Mallick S
    7. Lazaridis I
    8. Nakatsuka N
    9. Olalde I
    10. Lipson M
    11. Kim AM
    12. Olivieri LM
    13. Coppa A
    14. Vidale M
    15. Mallory J
    16. Moiseyev V
    17. Kitov E
    18. Monge J
    19. Adamski N
    20. Alex N
    21. Broomandkhoshbacht N
    22. Candilio F
    23. Callan K
    24. Cheronet O
    25. Culleton BJ
    26. Ferry M
    27. Fernandes D
    28. Freilich S
    29. Gamarra B
    30. Gaudio D
    31. Hajdinjak M
    32. Harney É
    33. Harper TK
    34. Keating D
    35. Lawson AM
    36. Mah M
    37. Mandl K
    38. Michel M
    39. Novak M
    40. Oppenheimer J
    41. Rai N
    42. Sirak K
    43. Slon V
    44. Stewardson K
    45. Zalzala F
    46. Zhang Z
    47. Akhatov G
    48. Bagashev AN
    49. Bagnera A
    50. Baitanayev B
    51. Bendezu-Sarmiento J
    52. Bissembaev AA
    53. Bonora GL
    54. Chargynov TT
    55. Chikisheva T
    56. Dashkovskiy PK
    57. Derevianko A
    58. Dobeš M
    59. Douka K
    60. Dubova N
    61. Duisengali MN
    62. Enshin D
    63. Epimakhov A
    64. Fribus AV
    65. Fuller D
    66. Goryachev A
    67. Gromov A
    68. Grushin SP
    69. Hanks B
    70. Judd M
    71. Kazizov E
    72. Khokhlov A
    73. Krygin AP
    74. Kupriyanova E
    75. Kuznetsov P
    76. Luiselli D
    77. Maksudov F
    78. Mamedov AM
    79. Mamirov TB
    80. Meiklejohn C
    81. Merrett DC
    82. Micheli R
    83. Mochalov O
    84. Mustafokulov S
    85. Nayak A
    86. Pettener D
    87. Potts R
    88. Razhev D
    89. Rykun M
    90. Sarno S
    91. Savenkova TM
    92. Sikhymbaeva K
    93. Slepchenko SM
    94. Soltobaev OA
    95. Stepanova N
    96. Svyatko S
    97. Tabaldiev K
    98. Teschler-Nicola M
    99. Tishkin AA
    100. Tkachev VV
    101. Vasilyev S
    102. Velemínský P
    103. Voyakin D
    104. Yermolayeva A
    105. Zahir M
    106. Zubkov VS
    107. Zubova A
    108. Shinde VS
    109. Lalueza-Fox C
    110. Meyer M
    111. Anthony D
    112. Boivin N
    113. Thangaraj K
    114. Kennett DJ
    115. Frachetti M
    116. Pinhasi R
    117. Reich D

    (2019) The formation of human populations in South and central Asia

    Science 365:7487.

    https://doi.org/10.1126/science.aat7487

    • Google Scholar
50. 1. Nelson D
    2. Kelleher J
    3. Ragsdale AP
    4. Moreau C
    5. McVean G
    6. Gravel S

    (2020) Accounting for long-range correlations in genome-wide simulations of large cohorts

    PLOS Genetics 16:e1008619.

    https://doi.org/10.1371/journal.pgen.1008619

    • PubMed
    • Google Scholar
51. 1. Nielsen SV
    2. Vaughn AH
    3. Leppälä K
    4. Landis MJ
    5. Mailund T
    6. Nielsen R

    (2023) Bayesian inference of Admixture graphs on native American and Arctic populations

    PLOS Genetics 1:1010410.

    https://doi.org/10.1371/journal.pgen.1010410

    • Google Scholar
52. 1. Patterson N
    2. Moorjani P
    3. Luo Y
    4. Mallick S
    5. Rohland N
    6. Zhan Y
    7. Genschoreck T
    8. Webster T
    9. Reich D

    (2012) Ancient admixture in human history

    Genetics 192:1065–1093.

    https://doi.org/10.1534/genetics.112.145037

    • Google Scholar
53. 1. Peter BM

    (2016) Admixture, population structure, and F-statistics

    Genetics 202:1485–1501.

    https://doi.org/10.1534/genetics.115.183913

    • PubMed
    • Google Scholar
54. 1. Pickrell JK
    2. Pritchard JK

    (2012) Inference of population splits and mixtures from genome-wide allele frequency data

    PLOS Genetics 8:e1002967.

    https://doi.org/10.1371/journal.pgen.1002967

    • PubMed
    • Google Scholar
55. 1. Posth C
    2. Nakatsuka N
    3. Lazaridis I
    4. Skoglund P
    5. Mallick S
    6. Lamnidis TC
    7. Rohland N
    8. Nägele K
    9. Adamski N
    10. Bertolini E
    11. Broomandkhoshbacht N
    12. Cooper A
    13. Culleton BJ
    14. Ferraz T
    15. Ferry M
    16. Furtwängler A
    17. Haak W
    18. Harkins K
    19. Harper TK
    20. Hünemeier T
    21. Lawson AM
    22. Llamas B
    23. Michel M
    24. Nelson E
    25. Oppenheimer J
    26. Patterson N
    27. Schiffels S
    28. Sedig J
    29. Stewardson K
    30. Talamo S
    31. Wang CC
    32. Hublin JJ
    33. Hubbe M
    34. Harvati K
    35. Nuevo Delaunay A
    36. Beier J
    37. Francken M
    38. Kaulicke P
    39. Reyes-Centeno H
    40. Rademaker K
    41. Trask WR
    42. Robinson M
    43. Gutierrez SM
    44. Prufer KM
    45. Salazar-García DC
    46. Chim EN
    47. Müller Plumm Gomes L
    48. Alves ML
    49. Liryo A
    50. Inglez M
    51. Oliveira RE
    52. Bernardo DV
    53. Barioni A
    54. Wesolowski V
    55. Scheifler NA
    56. Rivera MA
    57. Plens CR
    58. Messineo PG
    59. Figuti L
    60. Corach D
    61. Scabuzzo C
    62. Eggers S
    63. DeBlasis P
    64. Reindel M
    65. Méndez C
    66. Politis G
    67. Tomasto-Cagigao E
    68. Kennett DJ
    69. Strauss A
    70. Fehren-Schmitz L
    71. Krause J
    72. Reich D

    (2018) Reconstructing the deep population history of central and South America

    Cell 175:1185–1197.

    https://doi.org/10.1016/j.cell.2018.10.027

    • PubMed
    • Google Scholar
56. 1. Prüfer K
    2. Racimo F
    3. Patterson N
    4. Jay F
    5. Sankararaman S
    6. Sawyer S
    7. Heinze A
    8. Renaud G
    9. Sudmant PH
    10. de Filippo C
    11. Li H
    12. Mallick S
    13. Dannemann M
    14. Fu Q
    15. Kircher M
    16. Kuhlwilm M
    17. Lachmann M
    18. Meyer M
    19. Ongyerth M
    20. Siebauer M
    21. Theunert C
    22. Tandon A
    23. Moorjani P
    24. Pickrell J
    25. Mullikin JC
    26. Vohr SH
    27. Green RE
    28. Hellmann I
    29. Johnson PLF
    30. Blanche H
    31. Cann H
    32. Kitzman JO
    33. Shendure J
    34. Eichler EE
    35. Lein ES
    36. Bakken TE
    37. Golovanova LV
    38. Doronichev VB
    39. Shunkov MV
    40. Derevianko AP
    41. Viola B
    42. Slatkin M
    43. Reich D
    44. Kelso J
    45. Pääbo S

    (2014) The complete genome sequence of a Neanderthal from the altai mountains

    Nature 505:43–49.

    https://doi.org/10.1038/nature12886

    • Google Scholar
57. 1. Raghavan M
    2. Skoglund P
    3. Graf KE
    4. Metspalu M
    5. Albrechtsen A
    6. Moltke I
    7. Rasmussen S
    8. Stafford TW
    9. Orlando L
    10. Metspalu E
    11. Karmin M
    12. Tambets K
    13. Rootsi S
    14. Mägi R
    15. Campos PF
    16. Balanovska E
    17. Balanovsky O
    18. Khusnutdinova E
    19. Litvinov S
    20. Osipova LP
    21. Fedorova SA
    22. Voevoda MI
    23. DeGiorgio M
    24. Sicheritz-Ponten T
    25. Brunak S
    26. Demeshchenko S
    27. Kivisild T
    28. Villems R
    29. Nielsen R
    30. Jakobsson M
    31. Willerslev E

    (2014) Upper Palaeolithic Siberian genome reveals dual ancestry of native Americans

    Nature 505:87–91.

    https://doi.org/10.1038/nature12736

    • PubMed
    • Google Scholar
58. 1. Raghavan M
    2. Steinrücken M
    3. Harris K
    4. Schiffels S
    5. Rasmussen S
    6. DeGiorgio M
    7. Albrechtsen A
    8. Valdiosera C
    9. Ávila-Arcos MC
    10. Malaspinas A-S
    11. Eriksson A
    12. Moltke I
    13. Metspalu M
    14. Homburger JR
    15. Wall J
    16. Cornejo OE
    17. Moreno-Mayar JV
    18. Korneliussen TS
    19. Pierre T
    20. Rasmussen M
    21. Campos PF
    22. de Barros Damgaard P
    23. Allentoft ME
    24. Lindo J
    25. Metspalu E
    26. Rodríguez-Varela R
    27. Mansilla J
    28. Henrickson C
    29. Seguin-Orlando A
    30. Malmström H
    31. Stafford T
    32. Shringarpure SS
    33. Moreno-Estrada A
    34. Karmin M
    35. Tambets K
    36. Bergström A
    37. Xue Y
    38. Warmuth V
    39. Friend AD
    40. Singarayer J
    41. Valdes P
    42. Balloux F
    43. Leboreiro I
    44. Vera JL
    45. Rangel-Villalobos H
    46. Pettener D
    47. Luiselli D
    48. Davis LG
    49. Heyer E
    50. Zollikofer CPE
    51. Ponce de León MS
    52. Smith CI
    53. Grimes V
    54. Pike K-A
    55. Deal M
    56. Fuller BT
    57. Arriaza B
    58. Standen V
    59. Luz MF
    60. Ricaut F
    61. Guidon N
    62. Osipova L
    63. Voevoda MI
    64. Posukh OL
    65. Balanovsky O
    66. Lavryashina M
    67. Bogunov Y
    68. Khusnutdinova E
    69. Gubina M
    70. Balanovska E
    71. Fedorova S
    72. Litvinov S
    73. Malyarchuk B
    74. Derenko M
    75. Mosher MJ
    76. Archer D
    77. Cybulski J
    78. Petzelt B
    79. Mitchell J
    80. Worl R
    81. Norman PJ
    82. Parham P
    83. Kemp BM
    84. Kivisild T
    85. Tyler-Smith C
    86. Sandhu MS
    87. Crawford M
    88. Villems R
    89. Smith DG
    90. Waters MR
    91. Goebel T
    92. Johnson JR
    93. Malhi RS
    94. Jakobsson M
    95. Meltzer DJ
    96. Manica A
    97. Durbin R
    98. Bustamante CD
    99. Song YS
    100. Nielsen R
    101. Willerslev E

    (2015) Population genetics: genomic evidence for the Pleistocene and recent population history of native Americans

    Science 349:aab3884.

    https://doi.org/10.1126/science.aab3884

    • PubMed
    • Google Scholar
59. 1. Reich D
    2. Thangaraj K
    3. Patterson N
    4. Price AL
    5. Singh L

    (2009) Reconstructing Indian population history

    Nature 461:489–494.

    https://doi.org/10.1038/nature08365

    • PubMed
    • Google Scholar
60. 1. Reich D
    2. Patterson N
    3. Kircher M
    4. Delfin F
    5. Nandineni MR
    6. Pugach I
    7. Ko AM-S
    8. Ko Y-C
    9. Jinam TA
    10. Phipps ME
    11. Saitou N
    12. Wollstein A
    13. Kayser M
    14. Pääbo S
    15. Stoneking M

    (2011) Denisova admixture and the first modern human dispersals into Southeast Asia and Oceania

    The American Journal of Human Genetics 89:516–528.

    https://doi.org/10.1016/j.ajhg.2011.09.005

    • Google Scholar
61. 1. Reich D
    2. Patterson N
    3. Campbell D
    4. Tandon A
    5. Mazieres S
    6. Ray N
    7. Parra MV
    8. Rojas W
    9. Duque C
    10. Mesa N
    11. García LF
    12. Triana O
    13. Blair S
    14. Maestre A
    15. Dib JC
    16. Bravi CM
    17. Bailliet G
    18. Corach D
    19. Hünemeier T
    20. Bortolini MC
    21. Salzano FM
    22. Petzl-Erler ML
    23. Acuña-Alonzo V
    24. Aguilar-Salinas C
    25. Canizales-Quinteros S
    26. Tusié-Luna T
    27. Riba L
    28. Rodríguez-Cruz M
    29. Lopez-Alarcón M
    30. Coral-Vazquez R
    31. Canto-Cetina T
    32. Silva-Zolezzi I
    33. Fernandez-Lopez JC
    34. Contreras AV
    35. Jimenez-Sanchez G
    36. Gómez-Vázquez MJ
    37. Molina J
    38. Carracedo A
    39. Salas A
    40. Gallo C
    41. Poletti G
    42. Witonsky DB
    43. Alkorta-Aranburu G
    44. Sukernik RI
    45. Osipova L
    46. Fedorova SA
    47. Vasquez R
    48. Villena M
    49. Moreau C
    50. Barrantes R
    51. Pauls D
    52. Excoffier L
    53. Bedoya G
    54. Rothhammer F
    55. Dugoujon JM
    56. Larrouy G
    57. Klitz W
    58. Labuda D
    59. Kidd J
    60. Kidd K
    61. Di Rienzo A
    62. Freimer NB
    63. Price AL

    (2012) Ruiz-linares a

    Reconstructing Native American Population History. Nature 488:370–374.

    https://doi.org/10.1038/nature11258

    • Google Scholar
62. 1. Rogers AR

    (2019) Legofit: estimating population history from genetic data

    BMC Bioinformatics 20:526.

    https://doi.org/10.1186/s12859-019-3154-1

    • PubMed
    • Google Scholar
63. 1. Scally A
    2. Durbin R

    (2012) Revising the human mutation rate: implications for understanding human evolution

    Nature Reviews. Genetics 13:745–753.

    https://doi.org/10.1038/nrg3295

    • PubMed
    • Google Scholar
64. 1. Schiffels S
    2. Durbin R

    (2014) Inferring human population size and separation history from multiple genome sequences

    Nature Genetics 46:919–925.

    https://doi.org/10.1038/ng.3015

    • PubMed
    • Google Scholar
65. 1. Schiffels S
    2. Haak W
    3. Paajanen P
    4. Llamas B
    5. Popescu E
    6. Loe L
    7. Clarke R
    8. Lyons A
    9. Mortimer R
    10. Sayer D
    11. Tyler-Smith C
    12. Cooper A
    13. Durbin R

    (2016) Iron age and anglo-saxon genomes from East England reveal British migration history

    Nature Communications 7:10408.

    https://doi.org/10.1038/ncomms10408

    • PubMed
    • Google Scholar
66. 1. Seguin-Orlando A
    2. Korneliussen TS
    3. Sikora M
    4. Malaspinas A-S
    5. Manica A
    6. Moltke I
    7. Albrechtsen A
    8. Ko A
    9. Margaryan A
    10. Moiseyev V
    11. Goebel T
    12. Westaway M
    13. Lambert D
    14. Khartanovich V
    15. Wall JD
    16. Nigst PR
    17. Foley RA
    18. Lahr MM
    19. Nielsen R
    20. Orlando L
    21. Willerslev E

    (2014) Paleogenomics: genomic structure in Europeans dating back at least 36,200 years

    Science 346:1113–1118.

    https://doi.org/10.1126/science.aaa0114

    • PubMed
    • Google Scholar
67. 1. Shinde V
    2. Narasimhan VM
    3. Rohland N
    4. Mallick S
    5. Mah M
    6. Lipson M
    7. Nakatsuka N
    8. Adamski N
    9. Broomandkhoshbacht N
    10. Ferry M
    11. Lawson AM
    12. Michel M
    13. Oppenheimer J
    14. Stewardson K
    15. Jadhav N
    16. Kim YJ
    17. Chatterjee M
    18. Munshi A
    19. Panyam A
    20. Waghmare P
    21. Yadav Y
    22. Patel H
    23. Kaushik A
    24. Thangaraj K
    25. Meyer M
    26. Patterson N
    27. Rai N
    28. Reich D

    (2019) An ancient harappan genome lacks ancestry from steppe pastoralists or Iranian farmers

    Cell 179:729–735.

    https://doi.org/10.1016/j.cell.2019.08.048

    • PubMed
    • Google Scholar
68. 1. Sikora M
    2. Pitulko VV
    3. Sousa VC
    4. Allentoft ME
    5. Vinner L
    6. Rasmussen S
    7. Margaryan A
    8. de Barros Damgaard P
    9. de la Fuente C
    10. Renaud G
    11. Yang MA
    12. Fu Q
    13. Dupanloup I
    14. Giampoudakis K
    15. Nogués-Bravo D
    16. Rahbek C
    17. Kroonen G
    18. Peyrot M
    19. McColl H
    20. Vasilyev SV
    21. Veselovskaya E
    22. Gerasimova M
    23. Pavlova EY
    24. Chasnyk VG
    25. Nikolskiy PA
    26. Gromov AV
    27. Khartanovich VI
    28. Moiseyev V
    29. Grebenyuk PS
    30. Fedorchenko AYu
    31. Lebedintsev AI
    32. Slobodin SB
    33. Malyarchuk BA
    34. Martiniano R
    35. Meldgaard M
    36. Arppe L
    37. Palo JU
    38. Sundell T
    39. Mannermaa K
    40. Putkonen M
    41. Alexandersen V
    42. Primeau C
    43. Baimukhanov N
    44. Malhi RS
    45. Sjögren K-G
    46. Kristiansen K
    47. Wessman A
    48. Sajantila A
    49. Lahr MM
    50. Durbin R
    51. Nielsen R
    52. Meltzer DJ
    53. Excoffier L
    54. Willerslev E

    (2019) The population history of northeastern Siberia since the Pleistocene

    Nature 570:182–188.

    https://doi.org/10.1038/s41586-019-1279-z

    • Google Scholar
69. 1. Skoglund P
    2. Posth C
    3. Sirak K
    4. Spriggs M
    5. Valentin F
    6. Bedford S
    7. Clark GR
    8. Reepmeyer C
    9. Petchey F
    10. Fernandes D
    11. Fu Q
    12. Harney E
    13. Lipson M
    14. Mallick S
    15. Novak M
    16. Rohland N
    17. Stewardson K
    18. Abdullah S
    19. Cox MP
    20. Friedlaender FR
    21. Friedlaender JS
    22. Kivisild T
    23. Koki G
    24. Kusuma P
    25. Merriwether DA
    26. Ricaut F-X
    27. Wee JTS
    28. Patterson N
    29. Krause J
    30. Pinhasi R
    31. Reich D

    (2016) Genomic insights into the peopling of the southwest Pacific

    Nature 538:510–513.

    https://doi.org/10.1038/nature19844

    • PubMed
    • Google Scholar
70. 1. Soraggi S
    2. Wiuf C

    (2019) General theory for stochastic admixture graphs and F-statistics

    Theoretical Population Biology 125:56–66.

    https://doi.org/10.1016/j.tpb.2018.12.002

    • PubMed
    • Google Scholar
71. 1. Speidel L
    2. Forest M
    3. Shi S
    4. Myers SR

    (2019) A method for genome-wide genealogy estimation for thousands of samples

    Nature Genetics 51:1321–1329.

    https://doi.org/10.1038/s41588-019-0484-x

    • PubMed
    • Google Scholar
72. 1. Tambets K
    2. Yunusbayev B
    3. Hudjashov G
    4. Ilumäe AM
    5. Rootsi S
    6. Honkola T
    7. Vesakoski O
    8. Atkinson Q
    9. Skoglund P
    10. Kushniarevich A
    11. Litvinov S
    12. Reidla M
    13. Metspalu E
    14. Saag L
    15. Rantanen T
    16. Karmin M
    17. Parik J
    18. Zhadanov SI
    19. Gubina M
    20. Damba LD
    21. Bermisheva M
    22. Reisberg T
    23. Dibirova K
    24. Evseeva I
    25. Nelis M
    26. Klovins J
    27. Metspalu A
    28. Esko T
    29. Balanovsky O
    30. Balanovska E
    31. Khusnutdinova EK
    32. Osipova LP
    33. Voevoda M
    34. Villems R
    35. Kivisild T
    36. Metspalu M

    (2018) Genes reveal traces of common recent demographic history for most of the uralic-speaking populations

    Genome Biology 19:139.

    https://doi.org/10.1186/s13059-018-1522-1

    • PubMed
    • Google Scholar
73. 1. Terhorst J
    2. Kamm JA
    3. Song YS

    (2017) Robust and scalable inference of population history from hundreds of unphased whole genomes

    Nature Genetics 49:303–309.

    https://doi.org/10.1038/ng.3748

    • PubMed
    • Google Scholar
74. 1. Vallini L
    2. Marciani G
    3. Aneli S
    4. Bortolini E
    5. Benazzi S
    6. Pievani T
    7. Pagani L

    (2022) Genetics and material culture support repeated expansions into paleolithic Eurasia from a population hub out of Africa

    Genome Biology and Evolution 14:evac045.

    https://doi.org/10.1093/gbe/evac045

    • PubMed
    • Google Scholar
75. 1. van de Loosdrecht M
    2. Bouzouggar A
    3. Humphrey L
    4. Posth C
    5. Barton N
    6. Aximu-Petri A
    7. Nickel B
    8. Nagel S
    9. Talbi EH
    10. El Hajraoui MA
    11. Amzazi S
    12. Hublin J-J
    13. Pääbo S
    14. Schiffels S
    15. Meyer M
    16. Haak W
    17. Jeong C
    18. Krause J

    (2018) Pleistocene North African genomes link near eastern and sub-Saharan African human populations

    Science 360:548–552.

    https://doi.org/10.1126/science.aar8380

    • PubMed
    • Google Scholar
76. 1. Wang CC
    2. Reinhold S
    3. Kalmykov A
    4. Wissgott A
    5. Brandt G
    6. Jeong C
    7. Cheronet O
    8. Ferry M
    9. Harney E
    10. Keating D
    11. Mallick S
    12. Rohland N
    13. Stewardson K
    14. Kantorovich AR
    15. Maslov VE
    16. Petrenko VG
    17. Erlikh VR
    18. Atabiev BC
    19. Magomedov RG
    20. Kohl PL
    21. Alt KW
    22. Pichler SL
    23. Gerling C
    24. Meller H
    25. Vardanyan B
    26. Yeganyan L
    27. Rezepkin AD
    28. Mariaschk D
    29. Berezina N
    30. Gresky J
    31. Fuchs K
    32. Knipper C
    33. Schiffels S
    34. Balanovska E
    35. Balanovsky O
    36. Mathieson I
    37. Higham T
    38. Berezin YB
    39. Buzhilova A
    40. Trifonov V
    41. Pinhasi R
    42. Belinskij AB
    43. Reich D
    44. Hansen S
    45. Krause J
    46. Haak W

    (2019) Ancient human genome-wide data from a 3000-year interval in the Caucasus corresponds with eco-geographic regions

    Nature Communications 10:590.

    https://doi.org/10.1038/s41467-018-08220-8

    • PubMed
    • Google Scholar
77. 1. Wang C-C
    2. Yeh H-Y
    3. Popov AN
    4. Zhang H-Q
    5. Matsumura H
    6. Sirak K
    7. Cheronet O
    8. Kovalev A
    9. Rohland N
    10. Kim AM
    11. Mallick S
    12. Bernardos R
    13. Tumen D
    14. Zhao J
    15. Liu Y-C
    16. Liu J-Y
    17. Mah M
    18. Wang K
    19. Zhang Z
    20. Adamski N
    21. Broomandkhoshbacht N
    22. Callan K
    23. Candilio F
    24. Carlson KSD
    25. Culleton BJ
    26. Eccles L
    27. Freilich S
    28. Keating D
    29. Lawson AM
    30. Mandl K
    31. Michel M
    32. Oppenheimer J
    33. Özdoğan KT
    34. Stewardson K
    35. Wen S
    36. Yan S
    37. Zalzala F
    38. Chuang R
    39. Huang C-J
    40. Looh H
    41. Shiung C-C
    42. Nikitin YG
    43. Tabarev AV
    44. Tishkin AA
    45. Lin S
    46. Sun Z-Y
    47. Wu X-M
    48. Yang T-L
    49. Hu X
    50. Chen L
    51. Du H
    52. Bayarsaikhan J
    53. Mijiddorj E
    54. Erdenebaatar D
    55. Iderkhangai T-O
    56. Myagmar E
    57. Kanzawa-Kiriyama H
    58. Nishino M
    59. Shinoda K
    60. Shubina OA
    61. Guo J
    62. Cai W
    63. Deng Q
    64. Kang L
    65. Li D
    66. Li D
    67. Lin R
    68. Shrestha R
    69. Wang L-X
    70. Wei L
    71. Xie G
    72. Yao H
    73. Zhang M
    74. He G
    75. Yang X
    76. Hu R
    77. Robbeets M
    78. Schiffels S
    79. Kennett DJ
    80. Jin L
    81. Li H
    82. Krause J
    83. Pinhasi R
    84. Reich D

    (2021) Genomic insights into the formation of human populations in East Asia

    Nature 591:413–419.

    https://doi.org/10.1038/s41586-021-03336-2

    • Google Scholar
78. 1. Yan J
    2. Patterson N
    3. Narasimhan VM

    (2021) MiqoGraph: fitting admixture graphs using mixed-integer quadratic optimization