East Eurasia · primary clusters
The genetic history of East Eurasian populations is a complex topic and still under ongoing debate. Here we summarise the major ancestral clusters that played a significant role in shaping the later populations in the region, under the guidance of the latest studies.
Abbreviations
MMesolithic
NNeolithic
ENEarly Neolithic
EMNEarly–Middle Neolithic
MNMiddle Neolithic
LNLate Neolithic
LNBALate Neolithic – Bronze Age
Northeast Asia
Khairygas_16.7kya. One of the oldest samples recovered from Northeast Asia, this sample may be a near-unadmixed representative of Ancient Paleosiberian (APS) ancestry and is likely a near-unadmixed descendant of a population closely related to the founding population of the Americas. This sample can be modelled as ~97% Native-American-like (represented by Peru_Laramate_900B) and ~7% ANE-related (represented by AfontovaGora3). (Zeng et al. 2023, sup. mat. p. 172–173)
Ust_Kyakhta_14kya. The second-oldest sample in the APS cluster, this individual can be modelled as ~53–69% preceding Khairygas plus an undefined East Asian source, possibly related to KhatystyrCave_M_10.2kya. (Zeng et al. 2023, sup. mat. p. 173)
KhatystyrCave_M_10.2kya. Appearing to be of mixed ancestry, in a 3-way model this individual primarily inherits the majority of its genetic heritage from China_AmurRiver_Mesolithic (~57–78% — alternatively as ~61% from China_AmurRiver_Mesolithic + ~24% from Cisbaikal_LNBA + ~15% from China_SEastAsia_Coastal_EN), with the remaining portion originating from Ust_Kyakhta and another East Asian source (either China_NEastAsia_Inland_EN or China_SEastAsia_Coastal_EN). (Zeng et al. 2023, sup. mat. p. 173–174)
Kolyma_M_10.1kya. While models for Kolyma_M_10.1kya are quite varied, they are highly consistent in suggesting that this population is mostly descended from Ancient Paleosiberians (~65–73% Khairygas_16.7kya-like) and is less admixed with East Asian ancestry compared to Ust_Kyakhta_14kya. The remaining East Asian ancestry of Kolyma_M_10.1kya seems to be derived from either Yumin- or Amur-related sources. (Zeng et al. 2023, sup. mat. p. 174 / 196)
Dzhilinda1_M_N_8.4kya. This individual can be modelled as having all its ancestry coming from Ust_Kyakhta_14kya, or alternatively as an admixture of Kolyma_M_10.1kya or Khairygas with an East-Asian-related source — with much ancestry from KhatystyrCave_M_10.2kya or Transbaikal_EMN_old. (Zeng et al. 2023, sup. mat. p. 176)
Transbaikal_EMN_old. Consists of two individuals within the Transbaikal_EMN cluster (irk007.SG and cta016.SG), both older than the others at approximately 8,300 and 8,800 BP respectively. This genetic make-up can be represented as a two-way admixture between the Yumin individual (China_NEastAsia_Inland_EN) and KhatystyrCave_M_10.2kya. (Zeng et al. 2023, sup. mat. p. 176)
Altai_N_old. The oldest individuals under the Altai_N group label — uniting Kemerovo_N, Firsovo_N, and West_Siberia_N (extremely similar populations found in the Altai region) — older than ~8,000 BP (including individuals I2137 and I13098) can be successfully modelled as a 2-way admixture between a minor fraction of ancestry from AG3 and a major fraction from the three APS populations Khairygas, Ust_Kyakhta_14kya, and Kolyma_M_10.1kya. The same models pass in set B; no simpler models pass. (Zeng et al. 2023, sup. mat. p. 177)
Altai_N. This group — comprising all individuals in the Altai_N group label younger than 8,000 BP — can be modelled as descending completely from the older individuals in this cluster (Altai_N_old). (Zeng et al. 2023, sup. mat. p. 178)
Transbaikal_EMN. Consisting of younger individuals in the Transbaikal_EMN cluster, can either be modelled as descending completely in a one-way model from Transbaikal_EMN_old or via a two-way model with a small additional contribution from China_NEastAsia_Inland_EN (represented by the Yumin individual). (Zeng et al. 2023, sup. mat. p. 178)
China_AmurRiver_N. Consisting of individuals from the Amur River Basin region younger than ~8,000 BP, can be modelled as descending almost entirely (~77–98%) from prior populations of the Amur River region (China_AmurRiver_EarlyN or China_AmurRiver_Mesolithic) in multiple 2-way models with the addition of a small contribution from one other, typically Northeast Asian or APS source. (Zeng et al. 2023, sup. mat. p. 178) The individuals labelled Russia_Boisman_MN and Devils_Gate_N from the Amur Basin region are also associated with this ancestral cluster. Present-day Tungusic populations from the region also show strong correlation with this ancestry.
Cisbaikal_EN. Models of Cisbaikal_EN — which includes individuals from the Cisbaikal region from ~8,000 to ~6,800 BP — show major contributions from Transbaikal_EMN_old (~65–79%) and minor contributions from an APS population (Ust_Kyakhta_14kya ~35% or Khairygas_16.7kya ~31%). (Zeng et al. 2023, sup. mat. p. 178)
Mongolia_N_North. A two-way model for Mongolia_N_North — from the northern region of the Mongolian plateau just south of Lake Baikal at approximately 7,500 BP — as an admixture between Transbaikal_EMN_old (~65%) + China_NEastAsia_Inland_EN (~35%), alternatively with the addition of a small fraction (<7%) of ancestry from an ANE-rich source (either Altai_N_old or an APS-related population such as Khairygas_16.7kya, Ust_Kyakhta_14kya, Dzhilinda1_M_N_8.4kya, or Kolyma_M_10.1kya). (Zeng et al. 2023, sup. mat. p. 178)
Syalakh-Belkachi. The Syalakh-Belkachi population in the Middle Lena river valley in Central Yakutia from 6,800–6,300 BP can be modelled as descending entirely from the Mesolithic Northern Buryatian Dzhilinda1_M_N_8.4kya. In other passing 2-way models with higher p-values, Syalakh-Belkachi continues to demonstrate majority descent from Dzhilinda1_M_N_8.4kya with a minor contribution (~10–20%) from a broadly East-Asian-related population whose identity is poorly resolved. In other passing models, a major contribution from Dzhilinda1_M_N_8.4kya (~76%) and a minor contribution (~24%) from a population rich in Inland Northeast Asian ancestry (Yumin under the China_NEastAsia_Inland_EN label) — either Cisbaikal_EN, Transbaikal_EMN_old, or Mongolia_N_North — is observed. (Zeng et al. 2023, sup. mat. p. 178–179)
This population shows high affinity with the Paleo-Eskimo / Inuit cultures and may represent the major ancestral source contributed to them. A close genetic relationship was found between the Paleo-Inuit individual from the Saqqaq site in Greenland and the Syalakh-Belkachi population. 3-way models for the Saqqaq individual result in ~64–79% Syalakh-Belkachi-like admixture with the remainder from Kolyma_M_10.1kya and an East Asian source rich in Inland Northeast Asian ancestry. (Zeng et al. 2023, sup. mat. p. 204)
In turn, Saqqaq-related ancestry persisted on either side of the Bering Strait among ancient populations related to Eskimo-Aleut speakers. It is present at high levels in individuals from Old Bering Sea sites (up to ~43%) and in an individual from the classic Thule culture (up to ~59%), who is very similar to present-day Inuit. A Native-American-related source is also present in passing models for all of these groups — which in the case of the Old Bering Sea culture suggests back-migration from the New World to Chukotka. (Zeng et al. 2023, sup. mat. p. 212)
Saqqaq.SG ancestry persists along the northern shores of the Sea of Okhotsk in a population from the Tokarev culture from ~3,000 BP (Magadan_BA). This population is very similar to present-day Chukotko-Kamchatkans, and has very complex ancestry, with contributions from at least a Kolyma_M_10.1kya-related source and a Native-American-related source on top of Saqqaq.SG ancestry; the remainder is drawn from East Asian-related sources. (Zeng et al. 2023, sup. mat. p. 212) Ancestry related to that found in Syalakh-Belkachi also persists in an Iron Age Yakutian (~2,600 BP) from the Middle Lena River Valley, who can be modelled as cladal with Yakutia_LNBA. This individual is extremely similar to present-day Yukaghirs and Nganasans. (Zeng et al. 2023, sup. mat. p. 212)
While all ancient populations from Beringia and Arctic North America require a Saqqaq.SG-related source for passing models, ancient Athabaskans surprisingly do not share in this pattern. The models indicate that ancient Athabaskans can be modelled without any ancestry from Saqqaq.SG (i.e., with Saqqaq.SG retained in the reference populations); rather, in a 3-way model where most ancestry comes from a Native-American-related and a Kolyma_M_10.1kya-related source, model p-values are maximised by a minor (~9–13%) contribution from a third source that may be either Cisbaikal_LNBA or the closely-related Ust_Kyakhta_14kya. (Zeng et al. 2023, sup. mat. p. 213)
Yakutia_LNBA. Comprising individuals dating to ~4,500–3,200 BP from a region stretching from the Kan river valley in southeastern Krasnoyarsk to the Middle-Lower Lena valley in Central Yakutia and the Kolyma River basin in far northeastern Yakutia close to Beringia, Yakutia_LNBA can be modelled as a three-way admixture between Syalakh-Belkachi (~50%), Transbaikal_EMN (~42%), and a source related to Amur-River-related populations (~8%). (Zeng et al. 2023, sup. mat. p. 182) The Yakutia_LNBA genetic cluster has strong correlation with Uralic-speaking populations. The recent comprehensive analysis shows that this ancestry distinguishes Uralic-speaking populations from their non-Uralic-speaking neighbours; Uralic peoples derive the majority — and in some cases all — of their East-Asian-related ancestry from this source. (Zeng et al. 2023, sup. mat. p. 214–215)
Cisbaikal_LNBA. A model of Cisbaikal_LNBA as deriving most of its ancestry (~86%) from Ust_Kyakhta_14kya, with some ancestry from Inland Northeast Asians (~14%), is the only model passing with a high p-value. The origin of Cisbaikal_LNBA is not clear and is thought to be descended from an unsampled population that is related to but younger than Ust_Kyakhta_14kya. (Zeng et al. 2023, sup. mat. p. 184) Cisbaikal_LNBA ancestry — represented mostly by individuals associated with the Glazkov Culture — shows strong correlation with Yeniseian speakers and is constantly required as a source when modelling populations in the Yenisei Basin such as Yeniseian-speaking Kets, Uralic-speaking Enets and Selkup, and Turkic-speaking populations (Tuvinian, Tofalar, Tubalar, Altaian, Altaian_Chelkan, Khakass, Khakass_Kachin, Shor, Shor_Mountain, and Todzin), which is not the case for other Turkic or Uralic peoples. (Zeng et al. 2023, sup. mat. p. 219–220)
China_NEastAsia_Inland_EN. Represented by the Yumin (~8,000 BP) individual. This ancestry seems to have significantly contributed to many later populations in Northeast Asia.
Jomon Ancestry. Represented by 8,000–3,000-year-old hunter-gatherer individuals in Japan. The oldest individual sampled to date is Higashimyo, from Kyushu, Japan. Populations associated with this ancestry contributed partially to present-day Japanese populations. (Yang et al. 2022)
Southeast Asia
China_YellowRiver_N. Represented by Neolithic individuals from the Yellow River region, China. This ancestry is associated with the Proto Tibeto-Sinitic languages and is the major ancestral source that contributed to present-day South and East Asian populations (Han Chinese, Tibetan, Korean, Japanese, etc.). The Neolithic individuals grouped under the China_NEastAsia_Coastal_EN label are also closely related to the Neolithic Yellow River ancestry. (Yang et al. 2022)
Fujian Ancestry. The ancestry described in Yang et al. 2020 and 2021, consisting of 4 groups based on the Neolithic-era samples obtained from Fujian, China:
- China_SEastAsia_Coastal_EN
- China_SEastAsia_Coastal_LN
- China_SEastAsia_Island_EN
- China_SEastAsia_Island_LN
This ancestry is found in high levels in Austronesian populations and is thought to be associated with Austronesian expansions. It is also one of the most significant ancestries shared in high levels among present-day East and Southeast Asians.
South Asia
Ancient Ancestral South Indian (AASI). AASI is another "ghost" population, detected by indirect methods. Ancestry associated with the AASI lineage was found at low levels in almost all present-day Indian populations, particularly southern Indians. This ancestry has deep relationships with the other ancient East Eurasian ancestries; Onge or Irula populations are frequently used as a proxy to represent this ancestry. (Yang et al. 2022)
Sources
- Zeng, T. C., et al. (2023). Postglacial genomes from foragers across Northern Eurasia reveal prehistoric mobility associated with the spread of the Uralic and Yeniseian languages (pre-print). bioRxiv. https://doi.org/10.1101/2023.10.01.560332
- Yang, M. (2022). A genetic history of migration, diversification, and admixture in Asia. Human Population Genetics and Genomics, 2(1), 0001. https://doi.org/10.47248/hpgg2202010001
- Yang, M., et al. (2020). Ancient DNA indicates human population shifts and admixture in northern and southern China. Science, 369, 282–288. https://doi.org/10.1126/science.aba0909