In the heartland of Eurasia: the multilocus genetic landscape of Central Asian populations

The analysis of genetic variation reveals that Central Asian diversity is mainly shaped by linguistic affiliation, with Turkic-speaking populations forming a cluster more closely related to East-Asian populations and Indo-Iranian speakers forming a cluster closer to Western Eurasians.

I have often commented on the fact that Central Asians were mainly formed by the pendulum of Western Eurasians (Caucasoids) moving east with Indo-Iranian languages during prehistory and the later historical westward movement of Turkic speakers. There were other movements besides these, e.g., the Tocharians represent a non-Indo-Iranian eastward Caucasoid movement, while the Mongols represent a non-Turkic westward Mongoloid movement.

Central Asians are therefore today a variable mix of Caucasoids and Mongoloids, formed over the last few millennia, although the constituent elements are still present and recognizable. The Turkicization of the region was, to a large extent, the result of language shift among Iranian populations (Sakas-Scythians), but not without some genetic contribution from the original Turks who were a Mongoloid people akin to their linguistic Altaic cousins, the Mongols. This Mongoloid component is attenuated westward, reaching its minimum among Anatolian and Balkan Turkish speakers.

With respect to the admixture proportions (figure top left) presented in the paper, I have a couple of quick comments:

Using modern south Asians as representatives of a source population of Central Asia is problematic, as modern south Asians are admixed, comprised of Caucasoids and indigenous South Asians. While South Asia may have been a population source during remote periods of the Paleolithic, in the more recent post-Neolithic times when Central Asian populations were formed, South Asia was a population sink.

The use of only a few autosomal markers does give a broad overview of the east-west components in these populations, but it should be noted that the use of few markers tends to overestimate minority ancestral components.

Even with such a small number of markers, it is evident that the separation of groups at the population level is possible, as the correspondence analysis indicates: green/European, red/East Asian, blue/Indo-Iranian from Central Asia, orange/Turkic from Central Asia.

The paper includes STRUCTURE results for K=2 to K=6. Below is the STRUCTURE run for K=6:

While less distinct than what we would get with more markers, the emergence of several clusters of individuals is apparent (from left to right: East Asian, Turkic, Central Asian Iranian, South Asian, West Eurasian, Sub-Saharan). Notice how Hazaras and Uyghurs are islands of the Turkic component in the Central/South Asian cluster, and how some Uzbeks are Iranian-like while others are Turkic-like. I am reminded of an older study which found how mythology was used among some Uzbek groups to create a common ancestry for groups of unrelated origin.