Genomes of the Golden Horde elites and their implications for the rulers of the Mongol Empire

TL;DR

Ancient DNA from four elite medieval mausoleum burials in Ulitau, central Kazakstan, supports a Mongolian Plateau origin for Golden Horde elites and links three males to the C3* Y-chromosome lineage previously hypothesized for Genghis Khan. The study cannot confirm the identity of the individual buried in the Mausoleum of Joshi or the exact paternal lineage of Genghis Khan.

Key findings

• The four Kenggir individuals — CKZ001 (male, Alasha Khan), CKZ002 (male, Joshi), CKZ003 (male, Ayakkamir), and CKZ004 (female, Bolghan Ana) — were dated to 1286–1398 cal AD for CKZ002/CKZ003, 1300–1369 cal AD for CKZ004, and 1726–1814 cal AD for CKZ001. The three males shared paternal lineage C2a1a3 (C3*/C2a1a3-F1918), defined by the authors as C2a1a3a6-ZQ369; mtDNA lineages were D4c2b, C4a2a1, C4a1a+195, and G2a1g, respectively.
• Autosomal analyses showed the Kenggir individuals derived most of their genetic makeup from Ancient Northeast Asians (ANA), with a second component resembling Berel Scythians for all but CKZ001; CKZ001 was modeled with Ancient North Eurasian (ANE) ancestry. They could also be modeled as ANA plus Kipchaks, and PCA placed them near modern Altaic-speaking Siberian/northeast Asian groups.
• Kinship analyses found no first- or second-degree relationships among the four. CKZ003 was closely related to CKZ001 and CKZ002; CKZ001 and CKZ003 shared six IBD segments totaling 121.6 cM, while CKZ001 and CKZ002 shared only one ~9 cM segment. The claim that CKZ004 was the daughter of CKZ002 was not supported.
• A medieval IBD network connected the Kazak Steppe and Mongolian Plateau: pairs CKZ003–DA28, DA28–MRI001, CKZ002–CKZ003, and CKZ002–SHG003 shared 2–3 IBD segments, each at least 10 cM on average. CKZ001 had 9 ROH segments totaling 83.2 Mb, comparable to a Yakut individual.

Methods

• Ancient DNA was extracted from tooth, petrous bone, talus, and fibula samples of four individuals excavated from medieval mausoleums in the Ulitau Massif, central Kazakstan; two sites were excavated in this study and two Soviet-era sites were re-excavated. Radiocarbon dating was performed by Beta Analytic, and sequencing/library preparation was done in Japan.
• Genotypes were called at 1,240 K SNP capture sites; read depths were ~14.6× (CKZ001), ~3.9× (CKZ002), ~0.79× (CKZ003), and ~3.9× (CKZ004). The four individuals were merged with 400 AADR individuals for a 404-individual dataset, and PCA projected them onto a space calculated from 997 modern Eurasian individuals.
• Analyses included smartPCA; outgroup f3 and f4 statistics via qp3Pop/qpDstat; qpAdm ancestry modeling with nine reference populations; Y-chromosome haplotype placement via pathPhynder; mtDNA haplotype calling via Haplogrep and RAxML-NG; kinship via READ and ancIBD; IBD analysis after GLIMPSE2 imputation; and ROH analysis via hapROH.

Limitations

• The paper states it could not reach an unambiguous conclusion about the true identity of the male buried in the Mausoleum of Joshi and had insufficient evidence to support or refute his descent from Genghis Khan; this would require genetic data from the 12th–13th century pedigree of Genghis Khan. It also notes that CKZ001 dated to the 18th century CE, not the Golden Horde period, and that radiocarbon dates are not infallible.
• The paper states that Y-chromosome tree external node positions can be prone to error because ancient samples have missingness; CKZ003 had the lowest read depth, and more markers contradicted than supported the terminal branches of CKZ001 and CKZ002. It also notes qpAdm can produce equally likely ancestry models with different reference/source combinations, and that only two Kipchak samples were available and they were genetically distinct.
• The paper states IBD cannot identify exact relationships beyond third-degree relatives. Evident from the text: the Mausoleum of Joshi may not house Joshi himself, and the Kenggir individuals might belong to a lineage other than the Altan Urug; the authors also note no consensus on Genghis Khan’s paternal lineage and that the specific Kenggir lineage proposed remains unconfirmed.