Ancient DNA reveals the prehistory of the Uralic and Yeniseian peoples

Abstract

The North Eurasian forest and forest-steppe zones have sustained millennia of sociocultural connections among northern peoples, but much of their history is poorly understood. In particular, the genomic formation of populations that speak Uralic and Yeniseian languages today is unknown. Here, by generating genome-wide data for 180 ancient individuals spanning this region, we show that the Early-to-Mid-Holocene hunter-gatherers harboured a continuous gradient of ancestry from fully European-related in the Baltic, to fully East Asian-related in the Transbaikal. Contemporaneous groups in Northeast Siberia were off-gradient and descended from a population that was the primary source for Native Americans, which then mixed with populations of Inland East Asia and the Amur River Basin to produce two populations whose expansion coincided with the collapse of pre-Bronze Age population structure. Ancestry from the first population, Cis-Baikal Late Neolithic-Bronze Age (Cisbaikal_LNBA), is associated with Yeniseian-speaking groups and those that admixed with them, and ancestry from the second, Yakutia Late Neolithic-Bronze Age (Yakutia_LNBA), is associated with migrations of prehistoric Uralic speakers. We show that Yakutia_LNBA first dispersed westwards from the Lena River Basin around 4,000 years ago into the Altai-Sayan region and into West Siberian communities associated with Seima-Turbino metallurgy-a suite of advanced bronze casting techniques that expanded explosively from the Altai¹. The 16 Seima-Turbino period individuals were diverse in their ancestry, also harbouring DNA from Indo-Iranian-associated pastoralists and from a range of hunter-gatherer groups. Thus, both cultural transmission and migration were key to the Seima-Turbino phenomenon, which was involved in the initial spread of early Uralic-speaking communities.

Extended Data Figure 1 |. Sites with newly-reported samples.

This map displays all sites from which samples that fall in all major populations that are the subject of focus in this paper have come. These include all sites 1) whose samples fall on the NEAHG cline, 2) whose samples fall in the Cisbaikal_LNBA cluster or are admixed with it, 3) whose samples fall in the Yakutia_LNBA cluster or are admixed with it, 4) who are a part of the ten-population East Siberian transect described in our qpAdm modelling, and 5) who are Seima-Turbino period individuals. Each site is represented by a pie chart, whose size is proportional to the number of individuals from that site; the white fraction represents previously-published samples, and the black newly-published samples. Our sampling fills geographic and temporal lacunae.

Extended Data Figure 2 |. Chronology of sites and cultures in each geographic region.

Temporal and geographic disposition of cultures from the Mesolithic to the Late Bronze and Iron Ages across Northern Eurasia. Sites whose samples are analyzed in our paper are highlighted in darker boxes, within containing boxes indicating archaeological cultures. Sites whose colors are darker are those that we believe are most securely dated (based on radiocarbon, isotopic, and archaeological evidence).

Extended Data Figure 3 |. PCA with target populations projected onto ancient populations with an especially high fraction of ANE ancestry.

To illuminate the role that levels of ANE ancestry plays in generating variation among the populations we analyze, we use as a basis for another projection 71 shotgun-sequenced ancient individuals from across Eurasia, of which a large proportion are enriched in ANE ancestry and fall outside the range of present-day variation (e.g. individuals from populations like Tyumen_HG.SG or Kazakhstan_Botai.SG; for full list, see Supplementary Information section 4). The North Eurasian Hunter-Gatherer cline forms a curved arc stretching from EHG populations to present-day East Asians; the center of the arc dominated by populations rich in ANE ancestry is moved toward the positive direction in PC2. The individual furthest along the positive direction in PC2 is AG3. Clines formed by later Inner Asian populations, such as present-day Uralic, Turkic, and Mongolic speakers, as well as Late Bronze Age and Iron Age steppe populations such as Scythians and Sarmatians, are distinguished from the NEAHG cline by their much lower values along PC2, suggesting a much lower level of ANE ancestry. This PCA shows that populations along the NEAHG cline, remaining stable for many millennia, were substantially outside the range of present-day genetic variation in Northern Eurasia.

Extended Data Figure 4 |. PCA focusing on East Eurasian populations.

To further uncover possible structure among the East Asian ancestries within the populations that we analyze, we constructed a third PCA, using as a basis 37 East Asian present-day populations that have minimal West Eurasian admixture, and a single West Eurasian population (Norwegian), all genotyped on the Affymetrix Human Origins array (for a full list of populations analyzed, refer to Supplementary Information section 4). We projected all other shotgun-sequenced and hybridization-captured ancient and present-day individuals onto this basis. Once again, the North Eurasian Hunter-Gatherer cline forms a curved arc stretching from West Eurasian populations to present-day East Asians, with the center of the arc deflected toward the AG3 individual. East Asian populations are now differentiated along PC2, with Southeast Asians and East Asian agriculturalists taking on especially negative values along that dimension; populations from the Amur River Basin taking on intermediate values; then populations on the Mongolian Plateau and surrounding areas. A large gap separates these populations from Yakutia_LNBA and Russia_Tatarka_BA, which take on very positive values along PC2, close to present-day Nganasans and a genetically very similar Iron-Age individual from Yakutia who clusters with Nganasans in the previous two PCAs (Yakutia_IA.SG; also see Extended Data Figure 10). As one moves East along the NEAHG cline, their positions along PC2 tend to converge to the values found among populations of the Mongolian Plateau. In contrast, the Dzhilinda1_M_N_8.4kya and Kolyma_M_10.1kya individuals, and the Syalakh_Belkachi, Yakutia_LNBA and Russia_Tatarka_BA populations do not fall on the NEAHG cline and are shifted in the positive direction on PC2, toward the positions occupied by Nganasans, Beringian populations, and Native Americans. Lastly, Uralic populations possess the most positive values among PC2 when compared to Turkic, Mongolic and Tungusic populations.

Extended Data Figure 5 |. PCA focusing on ancient individuals from Northern Eurasia and the Americas.

To understand structure among NEAHG populations and non-NEAHG Siberians, we constructed two PCAs with ancient individuals including all individuals from the NEAHG cline, ancient non-NEAHG Siberians, and a selection of ancient Beringians and Native Americans. Notably, all these populations possess combinations of only WHG, EHG, ANE and East Asian ancestries. No individuals were projected in these PCAs. The first PCA (Extended Data Figure 7A) includes all individuals in the set, and the second (Extended Data Figure 7B) includes only individuals East of the Altai mountains. (A) In the first PCA we highlight several patterns. 1) the North Eurasian Hunter-Gatherer cline forms a curved arc stretching from West Eurasian populations to East Asian populations along PC1 and PC2. In PC3, populations along the NEAHG cline also form a straight line. However, populations rich in East Asian ancestry are differentiated along PC3, with individuals and populations within or closely related to the Cisbaikal_LNBA cluster having the most positive values, followed by those in the Transbaikal_EMN cluster and populations of the Mongolian Plateau, followed by individuals and populations in the Yakutia_LNBA cluster, followed by those from the Amur River Basin, followed by populations from the Bering Straits and the Americas. Notably, all individuals along the NEAHG cline, including individuals rich in East Asian ancestry (e.g. Cisbaikal_EN, Transbaikal_EMN, and all NEAHG individuals from the Krasnoyarsk region) form a straight line in PC3, suggesting a constant source of East Asian ancestry at the East Asian terminus of the NEAHG cline. 2) Khairygas_16.7kya occupies a central position among the other groups rich in East Asian ancestry in East Siberia, Beringia and the Americas, suggesting a lack of shared drift with later populations of the Bering region or the Americas. The situation is different for later populations: Kolyma_M_10.1kya falls among ancient Beringian populations, while the more East Asian-admixed Ust-Kyakhta_14kya and Dzhilinda1_M_N_8.4kya occupy a position in between Syalakh-Belkachi and ancient Bering Straits populations, with the even more East Asian-admixed Syalakh-Belkachi population showing even less of this displacement towards ancient Bering Straits populations. (B) We find a similar pattern in the second PCA, except with an opposite ordering of the clusters along PC3. Our results suggest that the distinctions we discover between groupings produced by the clustering analyses in Supplementary Information section 6 can be recovered in PCA analyses aimed at recovering fine-scale structure, despite underlying similarities in deep ancestry in populations in East Siberia, Beringia, and the Americas—all the products of admixture between ANE and East Asian ancestry.

Extended Data Figure 6 |. Summary of Genetic Changes Taking Place in Northern Eurasia.

Textual summary of our main findings.

Extended Data Figure 7 |. Graphical Summary of Genetic Changes Taking Place in Northern Eurasia.

Panel A shows the widespread distribution of individuals with Ancient Paleosiberian (APS) ancestry in Siberia before the Holocene, >10kya. Panel B shows the formation of the NEAHG cline by ~10kya, and the formation of the population on its eastern terminus (Transbaikal_EMN) through admixture between Amur River and Inland East Asian ancestries. Panel C shows the emergence of Cisbaikal_LNBA and Yakutia_LNBA in genetic turnovers in the Cis-Baikal and Northeastern Siberian regions in the Mid-Holocene, and the genetic diversity of Seima-Turbino period individuals ~4.0kya. Panel D shows the genetic gradient between West Eurasian ancestry and Yakutia_LNBA formed by present-day Uralic populations, along with all locations from which present-day populations with Cisbaikal_LNBA ancestry were sampled (grey dots ringed with black), alongside the geographic locations of two late Bronze Age/early Iron Age individuals (grey dots ringed with yellow) with >90% Cisbaikal_LNBA ancestry.

Extended Data Figure 8 |. Populations created by genetic grouping procedure applied over Northeast Siberians.

Details of populations created by the grouping procedure applied to individuals in Northeastern Siberia.

Extended Data Figure 9 |. Statistics of the form f4(Ethiopia_4500BP.SG, Target, “Route 2” population, Cisbaikal_LNBA).

Central Siberian populations from the Yenisei Basin (including Kets and South Siberian Turks) are highlighted in brown, while Arctic North American and Asian populations on either side of the Bering Straits populations are highlighted in blue. Bering Straits populations that are heavily European-admixed (Aleut and Yukagir_forest) are colored dark blue, while Samoyedic populations (Enets, Selkup, and Nganasan) are colored violet. Despite the similarity of the APS-rich populations in this comparison (all being admixtures between APS ancestry and East Asian ancestry), present-day groups of the Bering Straits are always closer to groups with “Route 2” APS ancestry (i.e., Kolyma_M_10.1kya → Dzhilinda1_8.4kya → Syalakh-Belkachi → Yakutia_LNBA), while Central Siberian populations of the Yenisei Basin are always closer to Cisbaikal_LNBA. For the version including a comparison with Ust-Kyakhta, refer to Supplementary Information section 8; Figures S94 & S95.

Extended Data Figure 10 |. ADMIXTURE results.

For details, refer to Supplementary Information section 5.

Extended Data Figure 11 |. The North Eurasian Hunter-Gatherer (NEAHG) Cline and its legacy through admixture in ancient northern Eurasia.

Higher-resolution version of Figure 1, containing the group/population labels of Figure 1C.

Extended Data Figure 12 |. Contribution of Yakutia_LNBA and Cisbaikal_LNBA to Admixed Inner Eurasians (AIEA).

Higher-resolution version of Figure 3, containing the group/population labels of Figure 3C, and three-letter population codes in Figure 3A. The codes are: ATN, Altaian; ATN_C, Altaian_Chelkan; BSK, Bashkir; BSM, Besermyan; BRY, Buryat; XNB_AR, China_AR_Xianbei_IA; CVS, Chuvash; DUR, Daur; DGN, Dolgan; DGX, Dongxiang; ENT, Enets; EST, Estonian; EVN, Even; EVN_E, Evenk_FarEast; EVN_T, Evenk_Transbaikal; FIN.SG, FIN.SG; LVL, Finland_Levanluhta; SAM, Finland_Saami_Modern.SG; FIN, Finnish; HZN, Hezhen; KLM, Kalmyk; KKP, Karakalpak; KRL, Karelian; KZK, Kazakh; KZK_C, Kazakh_China; BRL, Kazakhstan_Berel_IA; SARM_C, Kazakhstan_CaspianSteppe_Sarmatian; SARM_C.SG, Kazakhstan_CaspianSteppe_Sarmatian.SG; SAKA_K, Kazakhstan_CentralKazakhSteppe_Saka; SARM_K, Kazakhstan_CentralKazakhSteppe_Sarmatian.SG; KRK, Kazakhstan_Karakhanid.SG; KLK_1, Kazakhstan_Karluk_1.SG; KLK_2, Kazakhstan_Karluk_2.SG; KMK, Kazakhstan_Kimak.SG; KPC_1, Kazakhstan_Kipchak1.SG; KPC_2, Kazakhstan_Kipchak2.SG; SAKA_TS, Kazakhstan_Kyrgystan_TianShan_Saka; BRL_P, Kazakhstan_Pazyryk_Berel; TSM, Kazakhstan_Tasmola; SARM_W, Kazakhstan_WesternKazakhSteppe_Sarmatian; KET, Ket; KKS, Khakass; KKS_K, Khakass_Kachin; KMG, Khamnegan; KHT, Khanty; KOM, Komi_Zyrian; KRG_C, Kyrgyz_China; KRG_K, Kyrgyz_Kyrgyzstan; KRG_T, Kyrgyz_Tajikistan; TUR, Kyrgyzstan_Turk.SG; MNS, Mansi; MRI, Mari.SG; SCY, Moldova_Scythian; MGL, Mongol; MGA, Mongola; XNB_M, Mongolia_IA_Xianbei; MDV, Mordovian; NNI, Nanai; NGD, Negidal; NGS, Nganasan; NVH, Nivh; NGI_A, Nogai_Astrakhan; NGI_K, Nogai_Karachay_Cherkessia; NGI_S, Nogai_Stavropol; ORQ, Oroqen; ADB, Russia_Aldy_Bel; BLS, Russia_Bolshoy; MHE_1, Russia_EarlyMedieval_Heshui_Mohe_1; MHE_2, Russia_EarlyMedieval_Heshui_Mohe_2; SARM_S, Russia_EarlySarmatian_SouthernUrals.SG; KRS_o1, Russia_Karasuk_o1.SG; KRS_o, Russia_Karasuk_oRISE.SG; KRS, Russia_Karasuk.SG; SARM_L, Russia_LateSarmatian.SG; SARM_S.SG, Russia_MiddleSarmatian_SouthernUrals.SG; SARM, Russia_Sarmatian; SARM.SG, Russia_Sarmatian.SG; TGR, Russia_Tagar.SG; SAM.DG, Saami.DG; SKP, Selkup; SHR_K, Shor_Khakassia; SHR_M, Shor_Mountain; TTR_A, Tatar_Astrakhan; TTR_I, Tatar_Irtysh_Barabinsk.SG; TTR_K, Tatar_Kazan; TTR_M, Tatar_Mishar; TTR_S, Tatar_Siberian; TTR_Z, Tatar_Siberian_Zabolotniye; TTR_T, Tatar_Tomsk.SG; TTR_V, Tatar_Volga.SG; TDZ, Todzin; TFL, Tofalar; TBL, Tubalar; TKM, Turkmen; TVN, Tuvinian; UDM, Udmurt; SCY_U, Ukraine_Scythian; ULC, Ulchi; UYG, Uyghur; UZB, Uzbek; VPS, Veps; XIB, Xibo; YKT, Yakut; YKG_F, Yukagir_Forest; YKG_T, Yukagir_Tundra; KNY.SG, Russia_Yenisei_Krasnoyarsk_LBA.SG; KNY_o1.SG, Russia_Yenisei_Krasnoyarsk_LBA_o1.SG.

Extended Data Figure 13 |. f4 statistics of the form f4(Ethiopia_4500BP.SG, X, Yana.SG, China_Paleolithic) plotted against f4(AG3, X, Yakutia_LNBA, East Eurasian Population).

China_Paleolithic includes the Tianyuan and Amur_River_33K genomes, “East Eurasian Population” is some population grouping in Siberia or Northeast Asia other than Yakutia_LNBA, and X are Admixed Inner Eurasian populations (AIEA populations) including ancient Central Asian nomads from the Late Bronze to Iron Age down to the Scytho-Sarmatian period, as well as modern or ancient populations that speak languages from the Yukaghiric, Yeniseian (Kets), Uralic, Turkic, Mongolic, Tungusic, and Nivkh language families. Modern Uralic-speaking populations, and ancient putatively Uralic-speaking populations uniformly prefer Yakutia_LNBA to other East Asian ancestries no matter the other population used in the comparison. Furthermore, at any level of admixture between East and West Eurasian ancestries, the population with the greatest affinity to Yakutia_LNBA is always a Uralic-speaking population. f4-statistics therefore highlight the connection between Uralic populations and Yakutia_LNBA ancestry over other sources of East Asian ancestry.

Extended Data Figure 14 |. Characteristic Seima-Turbino artifacts.

1. Double-bladed dagger with a ring-shaped pommel, robbery find, unknown provenance (probable Omsk region or Rostovka). 2. Double-bladed dagger with a horse figurine on the pommel, an accidental find near Shemonaikha, East Kazakhstan. 3., 5., 7. Crook-backed knives with figurines on pommels: 3. from Seyma; 5. from Elunino-1, burial 1, 7. from Rostovka, burial 2. 4. Scapula-shaped celt with goat image, Rostovka, cluster of finds near burial 21. 6. double-bladed plate dagger with a double elk-head figurine pommel, an accidental find near Perm’ (probably associated with the Turbino site). 8. Top of staff with a horse figurine, an accidental find near Omsk. 9a. & 9b. Single-ear long spearhead with a relief figurine of a Felidae predator (tiger or mountain leopard) on the socket (9a. the spear tip,10b. the detail of the socket), an accidental find near Omsk.

Figure 1 |. The North Eurasian Hunter-Gatherer (NEAHG) Cline and its legacy through admixture in ancient northern Eurasia. (Top) Map.

A higher-resolution version of this image, with all population labels indicated, can be found in Extended Data Figure 12. Sampling locations of all individuals, and selected samples not on the cline mentioned in the text. (Center) PCA. We project ancient and present-day data onto variation from 122 genotyped present-day Eurasian and Native American populations selected to have minimal sub-Saharan African and Oceanian admixture. We observe a continent-spanning NEAHG cline, as well as a cline for Uralic populations stretching from European and Bronze Age Steppe populations to present-day Nganasans, Yakutia_LNBA individuals, and the ST-period site of Tatarka. (Bottom) Admixture Proportions. The first row of bar graphs presents qpAdm estimates of ancestry related to four sources (Russia_AfontovaGora for ANE, China_AmurRiver_LPaleolithic_19K for East Asian, Russia_HG_Elshanks for EHG, and Romania_IronGatesMesolithic for WHG) for all populations on the NEAHG cline. 84 out of 93 have passing models (p>0.01); populations that do not have a star above the bar plot. In these cases we show the model with the highest p-value. The second row of graphs display estimated admixture proportions for all 8 sources in the legend (expanding to include Tarim_EMBA1, Altai_N_9kya, Iran_GanjDareh_N and CHG); a pink dot above the bar plot indicates that all passing qpAdm models have Tarim_EMBA in the sources. A cross indicates a population used as a source (Altai_N_9kya; Russia_MiddleVolga_Elshanka_Chekhalino_4_10kya). For the Elshanka individual, we replaced the EHG source with Russia_Veretye_Mesolithic.SG.

Figure 2 |. Middle Holocene populations and admixture events that formed them.

(A, Top) Statistics of the form f4(Ethiopia_4500BP, Target, China_Paleolithic, Yana_UP), vs. f4(Ethiopia_4500BP, Target, X, Yana_UP), where X are ancient Native Americans or populations from the Bering Straits. The Target population’s position on the y-axis is proportional to its ratio of ANE and East Asian ancestry. Kolyma_M_10.1kya, MiddleVitim_Dzhilinda1_M_N_8.4kya, Syalakh-Belkachi, and Yakutia_LNBA are shifted left, indicating they share more drift with ancient Bering Straits groups than other populations with similar ratios of ANE and East Asian ancestry. (B, bottom left) Statistics of the form f4(Ethiopia_4500BP, X, China_NEastAsia_Inland_EN, China_AmurRiver_Mesolithic 14K), against f4(Ethiopia_4500BP, X, China_Paleolithic, MA1_HG) (left top) and f4(Ethiopia_4500BP, X, China_Paleolithic, Peru_Laramate_900BP) (left bottom), where X are ancient populations in Northeast Asia and Siberia. These statistics detect differentiation between an Inland East Asian-related source (proxied by the Yumin hunter-gatherer China_NEastAsia_Inland_EN), and an Amur-River-related source (represented by the China_AmurRiver_Mesolithic_14K). Populations from the Amur River region always have high affinity to China_AmurRiver_Mesolithic_14K, while those on the Mongolian Plateau and the Baikal area share more affinity with Yumin. The earliest strongly East Asian individual in Siberia, the Mesolithic MiddleLena_KhatystyrCave_M_10.2kya, is extremely Amur-River-related; other Northeastern Siberian groups high in APS ancestry, such as MiddleVitim_Dzhilinda1_M_N_8.4kya, Kolyma_M_10.1kya, and Syalakh-Belkachi, have both affinities; Cisbaikal_LNBA has extreme Inland Northeast Asian-relatedness. Affinity to China_NEastAsia_Inland_EN increases among agriculturalist populations along the Yellow River Valley. (C, bottom right) Schematic of population relationships in Northeast Asia and East Siberia, deduced from qpAdm in a ten-member transect from ~17kya to ~4kya. Major findings are: 1) That the MiddleLena_Khaiyrgas_16.7kya population is a near-unadmixed representative of an “Ancient Paleosiberian” (APS) population with Native American affinities, 2) APS ancestry persisted through two routes, and 3) the East Asian ancestry of Siberians derives from an Amur Basin-related source and an Inland East Asian-related source.

Figure 3 |. Contribution of Yakutia_LNBA and Cisbaikal_LNBA to Admixed Inner Eurasians (AIEAs).

A version with all population labels indicated is in Extended Data Figure 11. (A) PCA of f₄-statistics. A version with all population labels indicated is in Supplementary Information section 13. PCA of statistics of the form f₄(Ethiopia_4500BP, AIEA, AG3, East Asian) measure the affinity between an AIEA population’s East Asian ancestry and a panel of tested East Asian populations: China_AmurRiver_N, Mongolia_N_North, Transbaikal_EMN, Cisbaikal_LNBA, or Yakutia_LNBA. PC1 is correlated with proportion of any type of East Asian ancestry. At a given proportion of East Asian ancestry, ancient and present-day Uralic-speaking populations shift in PC2 in the direction suggesting disproportionate Yakutia_LNBA-relatedness. PC3 highlights similarity to Cisbaikal_LNBA (right), with most affinity in Yeniseians, South Siberian Turks, Samoyeds, and two Upper Yenisei outliers (~3.0–2.9kya, RISE497.SG and RISE554.SG, which our archaeological research suggests are from the Lugavskaya culture). (B) Cisbaikal_LNBA contribution to present-day populations. Populations with >4% Cisbaikal_LNBA ancestry, large black dots. Likely Lugavskaya culture outliers of the Minusinsk Basin, white stars. (C) Ancestry modeling. Top row gives qpAdm results for AIEA populations. One orange dot above the bars indicates that all East Asian ancestry can be modeled as Yakutia_LNBA; two indicate that—additionally—all passing models include Yakutia_LNBA among the sources. We also performed qpAdm with Cisbaikal_LNBA among the references and sources (Supplementary Information section 11); a grey dot indicates that all passing models include Cisbaikal_LNBA in the sources. The bottom row displays ADMIXTURE results. Almost all Uralic-speaking populations have East Asian ancestry nearly exclusively assigned to the Yakutia_LNBA component; Yeniseians, South Siberian Turks and Samoyeds are the only populations with appreciable levels of the Cisbaikal_LNBA-related component. The two likely Lugavskaya culture outliers of the Minusinsk Basin are the only individuals with almost all their ancestry assigned to the Cisbaikal_LNBA component.

Figure 4 |. Genetics of the Seima-Turbino Phenomenon.

(A) Map of Seima-Turbino sites and finds. Chernoozerye-1, Rostovka, Satyga-16 and Tatarka from which we have genetic data are marked in red circles; other important sites are numbered: 1, Seima. 2, Reshnoe. 3, Turbino. 4, Kaninskaya Cave. 5, Satyga-16. 6, Rostovka. 7, Samus-4. 8, Shaitanskoe Ozero. 9, Tatarka. (B) ADMIXTURE and qpAdm proportions for 16 Seima-Turbino-period individuals. 9 Rostovka, 2 Satyga-16, 1 Chernoozerye-1, 4 Tatarka. top row: selection of qpAdm models using a distal set of sources, with the simplest passing (p>0.01) qpAdm model with the highest p-value always displayed (same modeling as Figure 3B). No model passes for I32899 at p>0.01 as indicated by an asterisk above the bar). center row: qpAdm results for a more proximal set of sources, with individuals from Tatarka used as the source for Yakutia_LNBA and populations from Late Neolithic or Eneolithic Western Siberia (between the Urals and the Altai) as the source for NEAHG; Yakutia_LNBA has been added to the references. Ancestry from the population at Tatarka suffices to account for all the Yakutia_LNBA-related ancestry of the ST individuals even when Yakutia_LNBA is among the references. Samples that can be modeled with their East Asian ancestry derived completely from a Yakutia_LNBA-related source are highlighted by a single orange dot above their qpAdm bar charts, while those that require such a source in every passing model are highlighted by two such orange dots. One individual (I32816) requires Cisbaikal_LNBA ancestry among the sources (gray dot); the model displayed for this individual is the simplest passing model that contains both Cisbaikal_LNBA and Yakutia_LNBA ancestry among the sources. In both sets of qpAdm, the individual (I6787) from Chernoozersky-1 requires a contribution from a source from far West of the Urals (WHG ancestry. bottom row: ADMIXTURE proportions at K=18.