Post by Admin on Mar 6, 2023 21:39:06 GMT
3 Genetic relationship between ancientGöktürk and modern Turkic‐speaking populations Among the pastoralist Khanates in Eastern Steppe, it has been argued that the spread of Mongolic languages was connected with Xianbei, Rouran, Khitan and Mongol groups,while the Xiongnu, Türk and Uyghur were seen as linked withdiffusion of the Turkic languages. In the latter case, twowaves of diffusion have been hypothesized: the Bulgharic Turkic diffusion, beginning in the Hunnic period, instigated bythe earlier expansion of the Xiongnu, and followed up by the demic expansion associated with the Türkic Khanate(Nichols, 2011). During the second and third centuries CE,the Central Steppe populated by Iranian‐speaking groups was gradually replaced by an increasingly Turkic‐speaking population (de la Vaissière, 2005). The genetic relationship between the ancient Göktürk and modern Turkic‐speaking populations has remained a matter of controversy. In PCA,Turkic‐speaking groups were scattered along the east‐westcline in PC2, while Ashina was positioned in the NortheastAsians cluster. Modern populations showing the most similargenetic profile with Ashina were Tungusic, and secondly Mongolic; this relationship was revealed in PCA, unsuper-vised ADMIXTURE and outgroup‐f3statistics. Notably, these results did not provide evidence of genetic similaritybetween ancient Türkic people and present‐day Turkic‐speaking groups.To assess the degree of relation to Ashina, we compared Tungusic and Mongolic populations with Turkic populations by formal f4statistics (Fig. S16). We found that Tungusic and Mongolic‐speaking groups were more related to Ashina when compared with Turkic speakers, that is, f4(Mbuti,Ashina; Tungusic/Mongolic, Turkic) <0(Z<−3, except for Yakut and Dolgan), f4(X, Mbuti; Ashina, Tungusic/Mongolic)~0 (0 <|Z|<3). We further observed that Tungusic populations shared the highest genetic similarity with Ashinaas reflected in f4(Mbuti, Ashina; Tungusic, Mongolic) with the exception of Evenk_FarEast. Turkic‐speaking populationsharbored significantly divergent genetic profiles when compared with Ashina, that, f4(X, Mbuti; Ashina, Turkic)(Z>3 when X included West Eurasians; Z<−3 when Xincluded East Eurasians). We observed significant differ-entiation between Turkic‐speaking populations and ancient populations associated with the diffusion of Turkic languages(Fig. S17), indicating the spread of Turkic languages wasmainly driven by culture factors rather than demic diffusionand population integration.Unsupervised ADMIXTURE clustering analysis revealed awest‐east admixture pattern among Turkic populations. We also conducted a supervised ADMIXTURE clustering analysisto investigate whether the Ashina‐related ancient group had left a genetic legacy in modern Turkic populations. Theancestral proxies selected were Ashina, Mongolia_N_North,YR_LN, and Russia_Sintashta_MLBA based on unsupervised ADMIXTURE. We observed the proportions of ancestry related to ANA (Mongolia_N_North/Ashina) were diverse among Turkic‐speaking groups. ANA ancestry was absent among the westernmost Turkic populations (Fig. 1D).We performed f4(Turkic, Mbuti; Ashina, East Eurasian) tofurther determine whether the Ashina‐related lineage issufficient to explain the East Eurasian ancestry in Turkic populations, where East Eurasian included people who might not be influenced by the West Eurasian‐like ancestry. The Z‐scores of f4(X, Mbuti; Ashina, Turkic) tended towardspositive f4(most Z>0), showing that the Ashina was genetically closer to East Asians than Turkic groups(Fig. S18). We systematically explored qpAdm‐based admix-ture models of Turkic populations and compared them with Tungusic and Mongolic samples exhibiting close proximity toAshina (Table S3).
Ashina was consistent with deriving from one single source with some Mongolic and Tungusic populations (Bonan, Dongxiang, Evenk_Transbaikal, Oroqen,Tu, and Ulchi) but not with Turkic populations. The two‐way and three‐way admixture models of West and East Eurasians including Ashina failed for most Turkic populations, but onlysucceeded in Dolgan, Salar, Tuvinian, Yakut, Chuvash Altaian,Altaian_Chelkan, Khakass, Kazakh_China, Kyrgyz_China andKyrgyz_Tajikistan, revealing the genetic heterogeneity within Turkic populations. That contrasted with Tungusic andMongolic speakers that showed a dominating contribution from the Ashina‐related lineage in one‐way/two‐way admix-ture models. The results of qpAdm and supervised ADMIXTURE demonstrated a limited contribution fromAshina in Turkic‐speaking populations and the continuity ofANA ancestry showing by Ashina in Northeast AsianTungusic‐and Mongolic‐speaking populations.4
Implications In summary, we have unveiled the first genomic profile of the ancient Türkic royal family. Our genomic analyses of EmpressAshina revealed Göktürk's Northeast Asian origin (97.7%Northeast Asian ancestry and 2.3% West Eurasian ancestry),refuting the western Eurasian origin and multiple origin hypotheses. We found Ashina shared most genetic affinitywith post‐Iron Age Tungusic and Mongolic Steppe pastoralists,such as Rouran, Xianbei, Khitan, and Heshui_Mohe, and showed genetic heterogeneity with other ancient Türkic people,suggesting the multiple sources of the Türkic Khanate populations. Furthermore, the limited contribution from ancientGöktürk found in modern Turkic‐speaking populations once again validates a cultural diffusion model over a demic diffusionmodel for the spread of Turkic languages.
Ashina was consistent with deriving from one single source with some Mongolic and Tungusic populations (Bonan, Dongxiang, Evenk_Transbaikal, Oroqen,Tu, and Ulchi) but not with Turkic populations. The two‐way and three‐way admixture models of West and East Eurasians including Ashina failed for most Turkic populations, but onlysucceeded in Dolgan, Salar, Tuvinian, Yakut, Chuvash Altaian,Altaian_Chelkan, Khakass, Kazakh_China, Kyrgyz_China andKyrgyz_Tajikistan, revealing the genetic heterogeneity within Turkic populations. That contrasted with Tungusic andMongolic speakers that showed a dominating contribution from the Ashina‐related lineage in one‐way/two‐way admix-ture models. The results of qpAdm and supervised ADMIXTURE demonstrated a limited contribution fromAshina in Turkic‐speaking populations and the continuity ofANA ancestry showing by Ashina in Northeast AsianTungusic‐and Mongolic‐speaking populations.4
Implications In summary, we have unveiled the first genomic profile of the ancient Türkic royal family. Our genomic analyses of EmpressAshina revealed Göktürk's Northeast Asian origin (97.7%Northeast Asian ancestry and 2.3% West Eurasian ancestry),refuting the western Eurasian origin and multiple origin hypotheses. We found Ashina shared most genetic affinitywith post‐Iron Age Tungusic and Mongolic Steppe pastoralists,such as Rouran, Xianbei, Khitan, and Heshui_Mohe, and showed genetic heterogeneity with other ancient Türkic people,suggesting the multiple sources of the Türkic Khanate populations. Furthermore, the limited contribution from ancientGöktürk found in modern Turkic‐speaking populations once again validates a cultural diffusion model over a demic diffusionmodel for the spread of Turkic languages.