Genetics' random truths

By Razib Khan | May 27, 2012 1:05 am

Update: Please do not take the labels below (e.g., “Baloch”) as literal ancestral elements. The most informative way to read them is that they indicate populations where this element is common, and, the relationship of proportions can tell us something. The literal proportion does not usually tell us much.

End Update

I was browsing the Harappa results, and two new things jumped out at me. Zack now has enough St. Thomas Christian samples from Kerala that I think we need to accept as the likely model that this community does not derive from the Brahmins of Kerala, as some of them claim. Their genetic profile is rather like many non-Brahmin South Indians, except the Nair, who have a peculiar attested  history with the Brahmins of their region.

But that’s not the really interesting finding. Below is a table I constructed from Zack’s data.

Ethnicity Language S.Indian Baloch Caucasian NE.Euro
Karnataka Brahmin Dravidian 47% 38% 4% 6%
Karnataka Hebbar Iyengar Brahmin Dravidian 49% 36% 5% 5%
Karnataka Iyengar Dravidian 48% 39% 3% 5%
Karnataka Iyengar Brahmin Dravidian 48% 37% 3% 7%
Karnataka Kannada Brahmin Dravidian 51% 35% 3% 5%
Karnataka Konkani Brahmin Dravidian 47% 37% 2% 6%
Kerala Brahmin Dravidian 43% 39% 4% 6%
Tamil Brahmin Dravidian 46% 40% 3% 6%
Tamil Brahmin Dravidian 47% 40% 3% 5%
Tamil Brahmin Dravidian 48% 39% 9% 4%
Tamil Brahmin Dravidian 47% 38% 6% 4%
Tamil Brahmin Dravidian 48% 37% 6% 5%
Tamil Brahmin Dravidian 48% 37% 3% 5%
Tamil Brahmin Dravidian 48% 35% 5% 6%
Tamil Brahmin Iyengar Dravidian 47% 38% 6% 4%
Tamil Brahmin Iyengar Dravidian 47% 35% 6% 6%
Tamil Brahmin Iyengar Dravidian 50% 35% 2% 8%
Tamil Brahmin iyer/iyengar Dravidian 48% 38% 2% 5%
Tamil Brahmin iyer/iyengar Dravidian 48% 38% 4% 5%
Tamil Brahmin iyer/iyengar Dravidian 47% 37% 2% 5%
Tamil Brahmin iyer/iyengar Dravidian 47% 37% 6% 8%
Bengali Brahmin IE 43% 35% 4% 10%
Bengali Brahmin IE 45% 35% 2% 11%
Bengali Brahmin IE 44% 35% 5% 11%
Bihari Brahmin IE 39% 38% 5% 11%
Maharashtra/Madhya Pradesh Saraswat Brahmin IE 47% 39% 1% 6%
Mahrashtrian Desastha Brahmin IE 46% 38% 8% 5%
Oriya Brahmin IE 47% 36% 0% 9%
Punjabi Brahmin IE 33% 41% 13% 10%
Punjabi Brahmin IE 35% 40% 8% 11%
Rajasthani Brahmin IE 32% 38% 9% 15%
Sindhi Pushtikar/Pushkarna Brahmin IE 31% 36% 12% 10%
UP Brahmin IE 37% 38% 2% 14%
UP Brahmin IE 41% 37% 7% 11%

I was curious about the distribution of the “Northeast European” component in South Asia. This element is almost entirely lacking in non-Brahmin South Indians (except for the Nair), but, it is present in non-Brahmin Indo-European speaking Indians, including Biharis and Bengalis. And interestingly, it is present in the same rough fraction in North Indian and South Indian Brahmins regardless of locale, ~5 percent in the former case, and ~10-15 percent in the latter. I initially divided them into two language classes, but noticed that the Maharashtra samples were more like the South Indians.

Then I remembered something random: there is a tradition dividing Indian’s Brahmin communities in two, on a north-south split. The above partition does not perfectly reflect the oral history and custom, but it is very close. The Brahmins of South India are a particularly homogeneous lot. I’d bet that their diversity is a function of cultural evolution and adaptation to local circumstances, not disparate origins. Rather, they derive from some initial migration from a specific North Indian Brahmin community, and then admixed somewhat with another South Indian population (explaining their profiles being closer to the Southern average than that of Northern Brahmins).

Finally, most readers will be aware that I broadly accept the outline in Reconstructing Indian History. But, I do think there were multiple waves of northwest population intrusions into South Asia. In particular, I think the demographically preponderant wave was probably West Asian, while a later group brought some Eastern European ancestry into the mix as well. I think this explains nicely the fact that North Indian Brahmins have a “South Indian” cline but not a Northeast European cline (compare Bengali Brahmins to Punjabi Brahmins, and you’ll see what I mean). One possible model is that a very rapid sweep of an Indo-European speaking population may have occurred across the North Indian plain, overlain upon a local set of populations which had an ANI-ASI cline. The Genographic Project reportedly is going to present results which suggest that the Indian caste system pre-dates the arrival of the Aryans. That would comport well with this model, where earlier groups of northwesterners established a caste-like system, which the Aryans, who later formed the core of the twice-born castes, simply suited to their own needs upon arrival. If you look at Zack’s results using public data sets a very low proportion of “Northeastern European,” equivalent to what you see in South India, is found in a few groups:

– The Dravidian Brahui and Baloch

– Tribes and Dalits

– Austro-Asiatic populations

MORE ABOUT: India, Indian
  • Andrew S.

    This is very interesting to me, as my mother is Indo-Guyanese, and my father is Tamil-Sri Lankan. I should spring for four 23andMe kits (my younger brother, myself, my mother, my father) but I’m so cheap; not surprising given my recent ancestry. 😛

  • Nathan

    Is this Genographic Project working on information provided in part by K. Thangaraj and L. Singh of Centre for Cellular and Molecular Biology (Hyderabad) & Banaras Hindu University respectively? This is my hunch as you mention that caste is now being claimed to be pre-Aryan.

  • AV

    Considering the Indo-Iranian world as a whole, the Iranians and Kurds are about as Northeast European in various runs (HarappaWorld, Eurogenes hunter-gather, Eurogenes intra-North Euro analysis, Eurogenes K12b, Dodecad K12b) as Uttar-Pradeshi agriculturalist castes; Rajasthani Meghawal and Meena; and finally, Brahmins south of the Vindhya range (i.e South-Indian Brahmins). Iranians and Kurds, apart from the North-European genetic overlay, are rather similar to their immediate Northern Middle-Eastern neighbors, as evidenced by their IBD scores and sharing in various ChromoPainter results. Tajiks, Pashtuns, Kalash, Kashmiris, Punjabis and Uttar-Pradesh Brahmins exhibit the component in question doubly more than Kurds and Iranians.

    This strongly points to the fact that both Iran and India were already too thickly populated by farming groups for a semi-nomadic/pastoral group like the Indo-Iranians to penetrate the region and leave substantial signals of their genetic signature, while Afghanistan, Pakistan and North-Western India received more of it, as evidenced by the ~10-15% North-European admixture in the latter regions, which should be a good predictor for Afghan Pashtuns as well. Which is why probably why the generic Indo-Aryan speaking populace of India and typical Iranians have about the same Northeast European admixture. In addition to that, the Yunusbayev et al. Tajiks are also ~20% North_European.

    The Kalash, Pashtuns, Kashmiris, Punjabis, Rajasthani upper-castes and Uttar-Pradesh upper castes all seem to have around the same percentage of the North-East European component. Of course, they differ in their percentage of other West-Eurasian components, thus making the Kalash and Pashtuns the most West-Eurasian of the mentioned groups and the eastern Gangetic belt upper-castes the lest. The Eurogenes Hunter-gather vs. Farmer ADMIXTURE test for example pegs the Baltic hunter-gather score of the Kalash at 13%, the Pashtuns at 12.8% and U.P Brahmins at 13.72%. DMXX at the Vaêdhya research blog takes note of a similar trend in his write-up on the intra North-Euro Eurogenes analysis;

    One curious phenomenon is the similar West European-North Sea-Northeast European component proportions across the Turkmen, Uzbeks, Kyrgyz, Pathans, Uttar Pradesh Brahmins, Altaians and the Uyghur.

    As you’ve gone onto hypothesize, in the case of North-Indian upper castes; the Aryans, after sprinkling their genes around in Northwest South-Asia, very likely pressed themselves into a group that was to become the upper castes of North-India once they travelled further east (North India proper and the like) – in the sense that they likely gave their culture to an already pre-existing upper class and thereby were absorbed into them, albeit leaving an appreciable amount of genetic impact, too. Thus, the Northeast European fraction is relatively uniform throughout not only Indo-Aryan speaking Brahmins, but also ethnicities further west in general, while the other main West-Eurasian components (one centered in Balochistan and the other in the Caucasus) have a west-east cline.

    All in all, the genetic demographics of South-Asia strongly points out to the overwhelmingly increasing validity of the Kurgan hypothesis and the Pontic-Caspian dispersal point of proto-Indo-Europeans.

    PS. I’d also like to note that the Kalash don’t seem to be an Indo-Iranian relic as previously thought. They simply seem to be an inbred and genetically homozygous version of their Pashtun neighbors, going by their ADMIXTURE results per HarappaWorld and Eurogenes hunter gatherer vs. farmer test. I think this inference was made by Razib himself based on a phylogenetic tree by Uğur Hodoğlugil et al. showing the Kalash clustering with the Pashtuns and somewhat close to the Burusho.

  • Onur

    Iranians and Kurds, apart from the North-European genetic overlay, are rather similar to their immediate Northern Middle-Eastern neighbors, as evidenced by their IBD scores and sharing in various ChromoPainter results.

    AV, which Northern Middle Eastern neighbors do you mean specifically?

  • AV

    Assyrians, Armenians, Turks, etc. If I’m not wrong, Chromo-Painter and fineSTRUCTURE analyses at the Dodecad Project pretty much illustrated the same.

  • Onur

    Assyrians, Armenians, Turks, etc. If I’m not wrong, Chromo-Painter and fineSTRUCTURE analyses at the Dodecad Project pretty much illustrated the same.

    Yes. But the North European genetic overlay you mention is higher in Turks (even excluding Balkan Turks) than not only the one in Armenians and Assyrians but also the one in Kurds and Iranians.

  • Razib Khan

    #6, did someone check to see the north europ & northeast asian had a negative correlation? i’m wondering about rumelian turks….

  • Onur

    did someone check to see the north europ & northeast asian had a negative correlation?

    Among whom?

  • Paul Givargidze

    @ AV

    “Assyrians, Armenians, Turks, etc. If I’m not wrong, Chromo-Painter and fineSTRUCTURE analyses at the Dodecad Project pretty much illustrated the same.”

    I agree that there is probably significant shared ancient ancestry between the groups you mention. However, I do not believe it has been established through either of the two programs you refer to. At least as far as Assyrians are concerned. Heat maps from the most recent Dodecad fineSTRUCTURE and Chromo-Painter runs, respectively, below:

    Assyrian populations are 0, 11, and 19.



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