There has been some discussion in the comments about the C2 ancestral component at K=12 admixture runs which I called Pakistani/Caucasian.

First of all, we should remember that these "names" of ancestral populations are just rough mnemonics. They are chosen based on the frequencies of the component among modern reference samples. So the names have nothing at all to do with history.

In the case of Pakistani/Caucasian component, I wanted to emphasize the peaks of the component in Pakistan and the Caucasus. As commenters pointed out, the component is also quite high among the Iranians.

However, I have realized that this name, Pakistani/Caucasian, is a hindrance rather than a help for understanding the Admixture results. Also, this component is lower among the Pathan, Sindhis, and Punjabis than it is for Iranians etc. Therefore, the Pakistani part of the name is a bit of a misnomer, considering that the Pakistani populations it is high among comprise only about 5% of the country's population.

On the other hand, I do not like the name "Iranian" for this component. While it was suggested based on the geographical Iranian plateau which extends from the Caucasus to Balochistan, it still is confusing and it doesn't emphasize the peak areas.

Thus, I have renamed "Pakistani/Caucasian" as "Balochistan/Caucasus". I didn't use the shorter Baloch as this component is equally high among the Baloch, Brahui and Makrani, all populations living in the province of Balochistan.


  1. Also I think Balochistan-Caucasian works and is much more precise.

  2. I think that is a good solution. It also helps to emphasize the diversity of Pakistani populations. Since India is the largest country in South Asia and shows such a wide range of people, its diversity is always frong and center. But Pakistani diversity also deserves its own share of emphasis.

  3. Its interesting I didn't understand the resistance to the label "Pakistani".

    However what I've been pleasantly surprised is by the "homogeneity" of the Indus Valley/Pak populations. Its always been drilled on (virtually and abroad) that somehow Pakistanis are either "undifferentiated" from their neighbours or otherwise bear absolutely no similarity to one another.

    Its nice to know that on some level (genetic at the very least) we belong together. We are missing the Afghans though I'd be very interested to see where they end up.

    • The afghan pashtuns would most likely be very similar to the peshwari pathan hdgp sample. They might have slightly more North east asian etc.

      • perhaps but they are authocntous to afghanistan in a way they're not to nwfp.

        pashtunisation (like baluchisation) has only happened in the last millennia or so, their historic homelands were further west.

        i'd suspect the core populations are alot more "central asian"/middle eastern than south asian...

        or you could be right, its anyones guess either way, need more samples!

        • It really depends on where the HGDP Pathan samples are from. If they are from individuals who are from FATA for example, then I expect them to be very similar to Afghan Pashtuns.

        • What do you mean by pashtunisation? do you mean the forming of their identity? I believe they were a distinct group more than a millenia ago as a rig vedic tribe.

          I noticed also that we don't have any rajput samples as well.

          • Pity, I know. However, if my persuasion works, we might have a Makwana and a (Pakistani) Kashmiri Rajput in time.


            I believe NWFP (KP) only really got Pathanised in the last millennia (probably related to the Afghan invasions).

            A third of NWFP is still Punjabi speaking.

            Much as it was a demographic wave of Pathans and Baloch, it was also a cultural absorption; if you drill down into the history of specific tribes the lineages are made apparent (some are Indian/Indic, some are Afghan and some indigenous to their area).

            Their identity in Afghanistan, both the Pashtuns and Balochs (in Iran) probably attests to ancient times as you say in the Vedic/Avesta and most likely Achaemenian.

  4. Just fyi , i'm working on some maps to show the distribution of certain Components like C1 and C2. I will try to get something out soon, maybe later today

  5. I have sent you a sample for most debated component C2(baluchistan/caucasion) , you can post that up while i work on others. Any suggestions would be most welcome.

    Eventually i want to cluster our participants into categories and put them in eventually. Perhaps u can do that sometime. ( the grouping and averaging, i think statewise for now will be fine)

  6. Sim, what are the positions of the populations in your map based upon? Genetics or geography?

    • Geography. ( For example the pathan sample is from peshawar,pakistan). Some i don't know the exact city, so i take the capital city ( for example turks,georgians,lezgin etc ).

  7. Isopleths | Harappa Ancestry Project - pingback on March 17, 2011 at 9:46 am
  8. I have thought about the distribution of the Caucasus/Baluchistan component, and this is what I have come up with. My apologies if anyone already mentioned this.

    Baluchi is a Northwestern Iranian language, right? If its position in Southeastern Iran and Southwestern Pakistan is the result of recent migrations, the language might have been brought with a population of Northwestern Iranian origins, i.e., closer to the Caucasus. Hence, populations in Baluchistan would have an elevated portion of this component in comparison to other Pakistanis. This is hardly my area of expertise, so please tell me if that is reasonable?

    • Baloch migrated from western Iran, if I remember my history correctly. However, what about the Brahui? They speak a Dravidian language.

      • I was under the impression that despite the linguistic divide, the populations were not very distinct - see raz's post below which confirms my impressions. I do not know the origin of the Dravidian languages or what their likely migration path was.

  9. The most widely accepted region of origin for the Baluch is the Caspian sea area. It would be very interesting to see the results for Iranians of Rashti or Talysh origin.
    It's also important to keep in mind that, for the most part, Brahui/Baluch is purely a linguistic distinction in Baluchistan. Almost all in Baluchistan proper can speak both. Although Brahui is syntactically related to Dravidian (or Elamite) languages, it is more than 80% Baluchi/Persian in vocabulary including the basics such as numbers.
    That said, the Dravidian origin for the language remains a puzzle given it's location...

    • Their Y data looks slightly different form their neighbors, but not significantly so.

      The Baloch race: A historical and ethnological sketch - "The Arab conquest of Karman took place in A.h. 23, or only sixty-five years after the death of Naushirvan. The conquest was carried out by 'Abdu'llah, under the orders of the Khiilif'Umar; and all the accounts agree that the Arabs found the mountains of Karman occupied by a race known as Koch (in Arabic Qufj or Quf?), and some add the Baloch. None of the authorities are contemporary or nearly so. The earliest writers who deal with the subject are: Al-Bilazuri, who died in A.h. 279 (a.d. 802); Tabari,1 who wrote about A.h. 820 (a.d. 982); Mas'udi, whose work is dated A.h. 832 (a.d. 948); and Istakhri, circa A.h. 840 (a.d. 951). The first two of these,"in describing the conquest, only mention the Koch or Qufs; while Mas'udi and Istakhri, whose works are geographical and deal with their own times, speak both of Koch and Baloch."

      There is also a tradition narrated by Firdausi that that the Baloch accompanied the Ashkanians on campaigns in the Turan regions: "Next came shrewd Ashkash, Endowed with prudent heart and ready brain.
      His troops were from Baluchistan and Kutch [Koch?], And very rams to fight. No one had seen Their backs in battle or one finger mailless; Their banner was a pard [Tiger] with claws projecting."

  10. Excellent work I kept remembering their origins from Aleppo, Syria (which incidentally is now a very Kurdish city).

    Personally I would be surprised by the Brahui numbers in that case. Also the other Pakistani (and Indian) populations have this admixture, just not as high.

    Personally my pet theory is that this population arose in Iran/Khorasan, some thousands years ago, (northern Iran is hilly and densely populated; I'd imagine this to be the focal origin) and pushed out to the sparsely populated spaces of Caucasus & north western South Asia.

    Brahui could be some missing link in Elamo-Dravidian, the plot thickens.

    • The Iran/Khorasan theory looks quite logical. The Brahui too have traditions that they are from Aleppo with an Arab chief Chakur.

      On Elamo-Dravidian, the evidence is unclear at best. The Brahui language may just be remnants of the Rashtrakuta/Chalukya incursions from the northern Karnataka region. These Karnata chiefs were ruling large chunks of northern India and Nepal. The similarity of the Malto, Kurukh, Brahui, Oraon, Dhangar dialects lends some support. El Beruni mentions that Kannada speaking soldiers in Merv and Samarkand as a constituent part of Mahmud's army.

  11. Humayun Libi Luwi

    To have an idea about the possible origin of that component (wich I think should be named Western Asian or Anatolo-Mesopotamian) please take a look at the very recent (october 2010) pdf below

  12. Humayun Libi Luwi

    Is the european admixture amongst those populations due to the legacy of the British?

    • Anglo-Indians have British admixture. I highly doubt others have any detectable British influence.

Trackbacks and Pingbacks: