Razib Khan One-stop-shopping for all of my content

April 24, 2019

The Genetics Of The St. Thomas Christians, part 2

Filed under: Indian Genetics,Nasrani,St. Thomas Christians — Razib Khan @ 10:34 pm

Last year I posted The Genetics of the St. Thomas Christians. Recently I got some more samples. Of these, four were clearly self-identified as Southist/Knanaya Christians (as opposed to Northist Christians). The Knanaya are a bit different in their traditions than the broader much larger St. Thomas Christian community.

In the PCA above the bottom left are Middle Eastern groups. Druze and Yemeni Jews. Toward the top are Lithuanians. Green are Iranians. From the bottom right, up a diagonal axis, you see south-north Indian cline from low caste Telugu Christians, to Jatts from Punjab.

It does seem in relation to the other more generically labeled St. Thomas Christians the four Knanaya show some noticeable Middle Eastern shift.

Here’s an admixture plot:

The St. Thomas Christians have more yellow “Druze” cluster than other South Indians, with variation (the Knanaya have more).

I ran some Treemix. Didn’t detect major gene flow, but the Knanaya group was different from the other St. Thomas Christians, having a closer position to West Eurasians. When I ran a three-population test, it was the St. Thomas Christians (Nasrani above), and not the Knanaya, which registered admixture with Middle Eastern groups. It’s probably an artifact that the latter was not detected.

The Christians of Kerala are very similar to other peoples of Kerala. But, I now think it is more than 50% likely that they do have detectable Near Eastern ancestry above what you should expect.


January 14, 2018

The genetics of the St. Thomas Christians

Filed under: Indian Genetics,Nasrani,St. Thomas Christians — Razib Khan @ 10:08 pm

First, I have to say I appreciate everyone who keeps sending data to the South Asian Genotype Project. Basically, I’m automating the pipeline, finding ways to merge data from a host of sources, but also figuring out how to refine the analysis.

But until then, today I decided to do some more manual analysis of three St. Thomas Christian samples I have (also called Nasranis). The reason is that there were some questions on Twitter in relation to the genetics of this group, and though three is not a great sample size, it’s better than nothing.

The St. Thomas Christians are a diverse group of people of various denominations in the southern state of Kerala who have various origin stories. Though today the St. Thomas Christians have a range of denominational and sectarian affinities, their origins probably have something to the Church of the East.

These Christians also have origins about among the local Brahmin community, Jews, and West Asian settlers. To be honest, whenever people tell me about the Brahmin origins unless they were recent converts I discount this because there are about ten times as many St. Thomas Christians in Kerala as there are Brahmins. There is a small Jewish community in the area, and this region of India was long part of the Indian Ocean trade network of the Arabs.

I merged the three Nasrani samples with a lot of other populations. Zooming in on the South Asians, if you look at the PCA plot to the left (click it), you’ll see that they are not in the same cluster as the South Indian Brahmins (Brahmins from the four South Indian states are very similar to each). But, in comparison to non-Brahmin South Indians, they do seem Brahmin shifted.

As I have observed before these South Indian Brahmins can be thought of as more than 50% North Indian Brahmin, but the remainder being South Indian non-Brahmin. Aside from exotic exceptions (Parsis, Bengalis), most South Asians exist on an ANI-ASI “cline,” with lower caste South Indians being at one end of the cline (more ASI), and populations in the far northwest, such as the Kalash, being at the other end (more ANI). The PCA would suggest that the Nasrani are more ANI-shifted than a generic South Indian group, but less so than South Indian Brahmins.

Using Treemix to detect gene flow events, what I found is that the Nasranis look like a generic South Indian group. There’s no evidence of gene flow from Middle Eastern populations (Jews, Persians).

I did some f-3 tests and there isn’t anything conclusive I see to suggest Middle Eastern gene flow into the Nasranis fro that avenue.

Finally, I ran ADMIXTURE in supervised mode. Here are the average results for a set of South Asian populations (mean values):

Group Druze Georgian Han Iranian Telugu Yemenite Jew
Bangladeshi 1% 2% 12% 1% 83% 1%
Chamar 0% 0% 3% 0% 97% 0%
Gujurati_Patel 0% 1% 0% 10% 89% 0%
UP Kshatriya 0% 3% 1% 21% 76% 0%
Nasrani 0% 4% 1% 12% 83% 0%
Pathan 0% 4% 1% 55% 40% 0%
Piramalai_Kallar 0% 0% 2% 0% 97% 0%
SI_Brahmin 0% 4% 1% 16% 78% 0%
Telugu_Reddy 0% 3% 0% 0% 94% 3%
UP_Brahmin 0% 4% 1% 26% 69% 0%
UP_Kayastha 0% 0% 1% 20% 79% 0%
Velama 1% 1% 0% 2% 96% 0%
West_Bengal_Kayastha 0% 0% 7% 8% 85% 0%

In these data the Nasrani do look shifted in the same direction as South Indian Brahmins, though less so. Observe that there is no clear Middle Eastern signal in the Nasrani above and beyond what you see in South Asians. This, despite the fact that Indian Jews show a very strong signal of admixture from the Middle East. At this point I am confident in rejecting Nasrani St. Thomas Christian origins in a converted Jewish community, or one with a large degree of West Asian admixture.

Though the genetic profile of these three individuals does not support clear descent from South Indian Brahmins, I can not reject the model of Brahmin admixture into this community. On the contrary, a plausible model would see to be that various South Indian groups, including Brahmins, contributed to the Nasrani community over the centuries.

To be continued….

Powered by WordPress