Visualizing West Eurasia

By Razib Khan | January 5, 2011 2:49 pm

Over at his blog Dienekes Pontikos has taken some public data sets and his own Dodecad samples and generated a massive MDS plot of West Eurasian populations. The MDS is fine as it goes. It illustrates clearly that when you visualize an individual on a plot defined by the two largest dimensions of variation in the total data set clusters naturally emerge. Some of of them are totally expected. For example, the cluster of Ashkenazi Jews. But, some of the relationships need to be interpreted with care. The similar position of Sicilians with Ashkenazi Jews does not mean that these two populations are identical. Rather, their ancestral components exhibit similarities in such a manner that in a representation constrained to a few dimensions they shake out similarly. You can view the full thumbnail by clicking it, but I thought that for purposes of intuitive comprehension it would be useful to “cut out” the outlines of the distributions, and label them by geography. I added Ashkenazi Jews because I thought readers would be interested, but omitted the other Jewish groups.

CATEGORIZED UNDER: Genetics, Genomics

Comments (15)

  1. SV

    It seems reasonable to estimate that if there were samples from Afghanistan, they would form a bridge between the Iranian and Pakistani samples, to some degree.

  2. probably. it is an interesting illustration of the difference between genes & culture: the former can be much more continuous than the latter. it looks like the pakistani iranian speakers are much more like pakistan indo-aryan speakers than iranian speakers from the west, the kurds and farsi speakers. as you note, filling the whole with afghans of various ethnicities, central asian tajiks, and iranians from khorasan would be be of interest.

  3. Nava

    Who are those green circle and squares at the top right of the graph, if i may ask.

  4. pconroy


    If you look at the legend, the are as follows:

    Green Circle = Saudi
    Green Diamond = Bedouin

  5. It is somewhat surprising that one doesn’t see separate coastal and Danubian forks in the data, because that was one of the original big divides in the process of spreading farming to Europe, and because there pretty distinct divides (e.g. between Y-DNA haplogroups R1a and R1b) in some of the uniparental markers. Both European forks seem to spread out from a gateway in Greece, and the plot suggests more East-West axis admixture in Europe than one would expect.

    It isn’t too surprising but still worth noting that the Caucuses, while geographically compact and fairly modestly populated, have almost subcontinental levels of genetically diversity. The region is almost as thinly connected to the rest of West Eurasia genetically as it is linguistically in the autosomal markers. But, perhaps (as Dienekes has suggested) it is a common Caucuses component that is pulling the Northern European and Pakistani components towards it via Indo-European contributions. However, if that interpretation is correct, the extent to which Bronze Age and early Iron Age Indo-Europeans were replacing existing populations would have had to have been very great to have that effect.

  6. One expects, but does not know, if Direction3 had been included in the analysis, Ashkanazis and Sicilians point clouds would separate out on the 3rd axis.

  7. Jim

    So it’s settled. Ashkenazim are (basically) white Europeans. Agreed?

  8. So it’s settled. Ashkenazim are (basically) white Europeans. Agreed?

    i don’t agree. it isn’t “basically.” AJ are special in a variety of ways. but some people are invested in AJ being white europeans because of their racialist sympathies, and some are invested in them not being white europeans. you lot will continue to hold to your position no matter the PCA or bar plots i present from what i can tell, so wuteva. i.e., people can look at the same data visualization and come to exactly opposite viewpoints.

  9. oh, any retard discussion about whether ashkenazi jews are white or not will be deleted. i think they’re white personally, but i don’t give a shit if you agree or don’t agree. this stuff is moron bait, and i’m rather avoid that particular swarming of fish.

  10. pconroy

    What’s interesting to me is that the British and Germans are so close, and surprisingly the Slovenians are right by them. Then the Irish and Orcadians bridge the gap between the British/German and the Scandinavians.

    My Mother is under the IRI – so typically Irish by this analysis, while my Father is on top of the SCD, so pulls more towards Scandinavia – maybe some unknown Viking (or Finnish/Russian/Chuvash??) ancestry.

    My wife’s Sicilian grandmother is right on top of the SIS towards the Sephardic Jews – I was expecting her to be nearer the Greeks. So maybe this is reflective of the fact that she has relatives with North African mtDNA and others with connections to Spain and Spanish names, so she may have Spanish/Moorish ancestry, which is pulling her away from the Greeks.

  11. Pumice

    I’m surprised that the Bnei Menashe, who hail from eastern India and show noticeable Mongoloid features cluster soo close to the Brahui,Balochi and Pathan.
    If it was lack of focus in South Asia groups for this particular plot??, then how are the Kalash clustering further away from the Brahui,Baloch and Pathan than the Bnei Menashe when they (Kalash) are geographically closer than the BMJ to the aforementioned 3 Pakistani ethnic groups?

  12. Pumice

    I meant to include the Makrani when mentioning the Brahui,Balochi and Pathan clusters and the omission is unintentional.

  13. onur

    Dienekes writes: “The people who lived in Anatolia 1500 years ago were not Turks. Hell, most of them were not Turks until quite recently when the Turkish Republic decided to forge a Turkish nation out of the heterogeneous Muslim population of the Ottoman Empire.”

    Dieneke, as a self-identity the “Turkish” identity itself is much more recent than you think, as I’ve explained in another thread of this blog:

    the “Turkish” identity of Anatolian + Rumelian (=Balkan) + Cypriot “Turks” is the result of nationalism imported from the West during the late 19th and early 20th centuries by the Ottoman intelligentsia, and it spread among the masses only after the establishment of the Turkish Republic and with the nationalist reforms of Kemal Ataturk. Before all these, there was no “Turkish” ethnicity or nation, Seljuqs and Ottomans were called “Turks” only by the West, but no individual or group in the Seljuq and Ottoman empires from the sultans to the lowest levels of the society applied the “Turkish” identity for themselves, the word “Turk” was just an insult and not an ethnic term (in fact, there was no concept of ethnicity or nation in the Seljuq and Ottoman empires, the Millet system was a religious system, not ethnic). Before the spread of the “Turkish” identity among the masses beginning from Ataturk, Turkish-speaking Muslims of the former Ottoman lands didn’t call themselves “Turk”, but only “Muslim”.

    unlike the “Arab” identity, the “Turkish” identity was never applied by the very people designated with that identity (not even by the conquerors from Central Asia) before the spread of nationalism from the West. Before nationalism, the “Turkish” identity was an identity of the other (and often a despised other), not oneself.

    “Turkmen” identity, OTOH, was used by the conquerors from Central Asia and by their descendants (not pure of course, there are apparently no pure descendants of them) for themselves, but its use has always been rather limited to a minority among the Turkish-speaking Muslims of the former Ottoman and Iranian lands.

    The country name “Turkey” (and its other versions in other languages) was also only used by the West (for the territories ruled by “Turkish” rulers) before the spread of nationalism among “Turks”. Seljuqs of Anatolia called their own country “Rum” (=Rome) and Ottomans called their own country “Osmaniyye” (=Ottoman state). The name “Türkiye” (Turkish version of Turkey) is a late 19th century borrowing from the Italian word “Turchia” (then used for the Ottoman Empire) and was subsequently applied to the republic founded by Kemal Ataturk.

  14. I’m surprised that the Bnei Menashe, who hail from eastern India and show noticeable Mongoloid features cluster soo close to the Brahui,Balochi and Pathan.

    i assume that is a cording error.

  15. Navatar

    It would interesting to see other races East Eurasians and Africans on the same plot, just to see where in what proximity to whom they would cluster.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar