Empress Ashina

2 Genetic relationship between Ashinaand other post‐Iron Age Central/EasternSteppe pastoralistsFrom the Iron Age onward, a succession of nomadic regimesrose and fell in the Eastern Steppe: the Xiongnu (the 3rdcentury BC–the 1st century CE), Xianbei (the 1st century CE—the 6th century CE), Rouran (the 4th century CE–the 6thcentury CE), Türk (the 6th century CE–the 8th century CE),Uyghur (744–840 CE), Khitan (916–1125 CE), and—mostaggressively expansive and successful of any—the Mongolempire (1206–1368 CE). Adjacent to the Eastern Steppe, theCentral Steppe also witnessed the shifting fortunes ofnomadic regimes, with Hun (the 4th century CE–the 6thcentury CE), Wusun (the 2nd century BC–the 5th century BC),Kangju (140| BCE), Turkic Karluk (the 7th century CE–13thcentury CE), Kimak (the 8th century CE–the 11th century CE),Kara‐Khanid (the 10th century CE–the 13th century CE), andKipchak (the 11th century CE). The genetic relationshipbetween the Türkic Khanate and these Eastern/CentralSteppe nomadic populations has long been uncertain.We found that Ashina was genetically different fromCentral Steppe pastoralists of the Hun, Wusun and Kangjuperiod (|Z| score of f4(X, Mbuti; Ashina, post‐Iron Age Central/Eastern Steppe) >3; Fig. S9). Xiongnu was a geneticallyheterogeneous population with some samples havingreceived abundant genetic influence from ANA and YellowRiver farmers (such as lateXiongnu_han; Table S2E) butothers harboring closer connection with West Eurasian(sucha as early Xiongnu, lateXiongnu, and lateXiongnu_sar-matian; Fig. S10) (Jeong et al., 2020). Ashina showed a closerelationship with lateXiongnu_han. The Ashina shared themost genetic affinity with Mongolic Rouran and Xianbei,followed by Khitan in outgroup‐f3statistics (Table S2A). Thiswas consistent with non‐significant f4(X, Mbuti; Ashina,Rouran/Xianbei /Khitan/Heshui_Mohe) (|Z|<2) and pairwiseqpWave (P>0.01) (Figs. S11, S12, S15), which indicated thatAshina formed a genetic clade together with Rouran, Xianbei,Khitan and Mohe. Among nomadic populations after theTürkic Khanate, Eastern Steppe Khitan and Mohe, andCentral Steppe Kazakhstan_GoldenHordeAsian (an individualfrom Jochi Khan's Golden Horde army from the Ulytaumountains) showed genetic similarity with Ashina (non‐significant Z‐values in f4and P>0.01 in pairwise qpWave;Figs. S12, S14, S15). By contrast, Uigur, Karluk, Kimak,Kipchak, Karakhanid and Mongols were genetically quite separated from the Ashina sample (significant Z‐values in f4and P<0.01 in pairwise qpWave; Figs. S12B, S13).Different genetic relationships between Ashina and thepost‐Iron Age Eastern/Central Steppe nomadic populationswere also evident in the ancestral compositions shown inADMIXTURE, supervised ADMIXTURE and qpAdm (Figs. 1C,1D; Table S2E): among the Eastern Steppe pastoralists, someof the Xiongnu groups (earlyXiongnu_rest, lateXiongnu_Han,and lateXiongnu), Rouran, Xianbei, Khitan, Mongol andHeishui_Mohe harbored dominating East Eurasian ancestryfrom 82.9% to 99.8% and additional West Eurasian ancestry.In contrast, the early West Xiongnu (earlyXiongnu_west) andlate Sarmatian Xiongnu (lateXiongnu_Sarmatian) derivedancestry mainly from West Eurasian; for example, earlyWest Xiongnu exhibited 68.4% Afanasievo‐related ancestry.Among the Central Steppe pastoralists, Wusun, Kangju, and Tianshan Hun derived a majority of their ancestry (62.4%–73%)from Western Steppe nomadic Afanasievo groups with the remainder (37.6%–27%) characterized as BMAC (the Bactria‐Margiana Archaeological Complex) and East Eurasian. TheTurkic Karluk, Kipchak, and Karakhanid could be modeled derived 35%–50.6% of ancestry from Afanasievo, 10.5%–21.7%from BMAC, and 38.9%–49.4% from YR_IA. The proportion of East Eurasian ancestry increased in the Medieval pastoralists(CentralSteppe_Medieval_Nomad and Kazakhstan_Golden-HordeAsian) from 67.3% to 82.5%, but a historical Kazakhstanindividual (Kazakhstan_His) displayed a different geneticprofile due to a primary ancestry at 75.5% derived fromWestern Steppe pastoralists.The Ashina had not shown close genetic affinity with early Medieval or Central Steppe Türk as reflected in the lowervalues of outgroup—f3(Ashina, early Medieval or Central-Steppe Türk; Mbuti) and the significant values of f4(X, Mbuti;Ashina, early Medieval or CentralSteppe Türk) (|Z|>3)(Figs. S5, S8). Although earlyMed_Turk showed a geneticaffinity with a Rouran individual (Fig. S4), earlyMed_Turkharbored a different genetic profile from the Ashina individual, as shown in the PCA and pairwise qpWave. The early Medieval Türk (earlyMed_Turk) derived the major ancestry from ANA at a proportion of 62.2%, the remainder from BMAC (10.7%) and Western Steppe Afanasievo nomad(27.1%) (Figs. 1C, 1D; Table S2E). The geographically remoteCentral Steppe Türk (Kyrgyzstan_Turk and Kazakhstan_Turk)could be modeled as an admixture of ANA (Mongolia_N_-North), BMAC, and West Steppe pastoralists (Afanasievo)(P=0.0196) (Fig. S5; Table S2E). In contrast to Ashina,Central Steppe and early Medieval Türk exhibited a high but variable degree of West Eurasian ancestry, indicating there was a genetic substructure of the Türkic empire.

www.researchgate.net/publication/366965287_Ancient_Genome_of_Empress_Ashina_reveals_the_Northeast_Asian_origin_of_Gokturk_Khanate