Admixture (Ref3 K=11) HRP0161-HRP0170

Here are the admixture results using Reference 3 for Harappa participants HRP0161 to HRP0170.

You can see the participant results in a spreadsheet as well as their ethnic breakdowns and the reference population results.

Here's our bar chart and table. Remember you can click on the legend or the table headers to sort.

If the above interactive charts are not working, here's a static bar graph.

HRP0161 is my mom.

HRP0169 is our first 100% Sri Lankan Tamil. Admixture results are close to the other non-Brahmin Tamils.

HRP0170 is a Haryana Jatt whose results match the other Haryana/UP Jatt.


  1. The two Haryana Jatts are an interesting case - they exhibit more European admixture than even the Punjabi Jatts. Rather strange that there would be such a significant difference with regards to this component between the two groups when they are merely from two different political territories i.e states (Punjab vs Haryana). Haryana was part of the greater Punjab region historically, and was made a separate state only in 1966.

    As an aside, Zack, while I don't mean to pry, I can't help but ask on what basis the Brazilian, Belizean, Mexican, Hispanic and European participants (who's ancestry has little to do with South-Asia in a colonial context, not the ones from the British Isles) were accepted into the project.

  2. My guess is that the haryana jats have a particular european component in larger quantity than the punjabi ones. From the Dodecad results it seems to be the east european component. While Punjabi jats seems to have more west asian. This West asian or at least part of it is probably subsumed under the "South Asian" component for Zacks analysis. Btw i provided 2 haryana jat free kits , and this is one of them . The other one is awaiting results. So we should have one more coming up soon.

    • Hi Guys,
      I want to tell everybody that Simranjit providing two free Kits for samples from Haryana Jats and this was an appreciable initiative on his part to do his bit for projects like Harrapa Ancestry Project and he deserves credit for going the extra mile for such studies.

    • Simranjit, it is commendable that you're, as Vikram put it, going the extra mile towards bringing further insight into Jatt genetics. Perhaps you could also fund some Gurjar, Kambhoj, Mohyal and Khatri individuals some time down the line, if it is of interest to you and you have the cash to spare :-).

      Back on topic, I don't think Dienekes'/Dodecad's K=12 v3 ADMIXTURE results should be taken all that seriously. For one, the components are rather Euro-centric. Secondly, there are some odd trends, such as the high Western European scores among the Indian/South Asian_D participants. It seems as if the ADMIXTURE algorithm tends to split the same components at different places when different Ks are used. Thus, I think the old West Asian component is primarily being distributed between the new v3 West Asian, West European and Mediterranean.

      If any sort of Northern European admixture, South Asians tend to have North East European admixture. In this regards, Dienekes' K=12 ADMIXTURE results for selected participants - which he posted a while ago are very plausible and are in line with the theories regarding the genetic origins of Indo-Europeans out there. However, the elevated European score of the other Haryana/U.P Jatt (27%) is more or less in line with his Northern European score in the standard K=10 Dodecad analysis (about 20%, IIRC). However, I still wonder how the elevated European scores of the two Haryanvi Jatts would fit in a historical context..

      • Ah, rats. The Italics again! Zack, if it isn't a bother, kindly fix.

        Anyway, HRP0170's haplogroups are also very interesting - Y-DNA Q1a2 and mtDNA U5a1a1. Y-DNA Q peaks among Siberians (also, Native Americans) and mtDNA U5 among various Europeans. While U5 isn't a complete surprise - in fact we have two participants, HRP107 and HRP111 who are Pakistani Punjabi Rajputs and carry mtDNA U5a1b1 - the y-DNA Q is indeed quite fascinating. And given that it is North Eurasian specific, it only adds further fuel to the speculations regarding Scythian admixture among the Jatts.

      • I do have more kits i hope to be able to pass on to perhaps kambhojs or khatris etc, but it's hard to convince relatives/friends to spit( they look at you like you have a screw loose) . For the haryana jats all i needed to do was request for volunteers on jatland forum and then get a POC to distribute 2 kits for me in india.

        • I have an Arora friend I'd like to test, who, I'm pretty sure will be willing to send in a spit sample. I haven't seen any [active] Aroras on 23andMe, neither have I seen any genetic studies on them. Aroras are another one of those caste groups who have a rather uncertain status, they are generally merchants but are considered Kshatriyas.

          The only problem is, I reside in India, and am a minor. So how I will go about getting the kit to my city (if you have a spare one and are willing to fund it), is the big question here.

          • Messed up the e-mail ID entry, thus the change in the color of the avatar. Sorry 'bout that.

          • Your Arora friend's a minor? Is it possible to test one of his/her parents?

            Also who are Aroras?

          • Arora are supposed to be Khatris from Aror (Sindh, Indus ).

          • Zack, my friend is almost 20 years old. I am 17 (thus, I am the minor here). What I was saying was that I won't be able to fund even the delivery charges even in the scenario that Simranjit funds the kit apart from that part of the costs. So, we'll have to figure out how to go about delivery if the plan materializes. We can test anyone, either him or one of his parents, I suppose. It'll be better to test him though, I suppose, given that he is well aware of my interest in personal genomics. It'll seem less outlandish to him than his parents :-).

            While Parasar has already answered your question regarding the Aroras, here is the Wikipedia page on the community-

        • Simranjit , do you need any more samples from the Jats side or Gurjar let me know.Maybe i can get some other Jat clan sample or can get someone within the malik clan of Jats but from different place to see how widespread this Q1a2 is.
          i am waiting for any discount by either 23andme or FTDNA to buy more kits and get more Jat samples .

        • The result for other Haryana Jat might also come in the coming days.Besides i can check someone from Sandhu clan in Haryana Jats.

  3. BTW Zack, HRP0031 (who's account I manage from time to time) is also a Sri Lankan Tamil Vellalar on his paternal side. Considering there is now a second Vellalar in the project, it'd be great if you could specify that in the ethnicity spreadsheet.


  4. These Haryana Jatts are worth studying. To have another Haryana Jatt added to the study and to also have such a european percentage is quite interesting, seeing how haryana and punjab were once the same province. My father always told me there was a difference between them but that was just on face value. When insisting on difference i don;t mean in dividing or meaning better but merely genetic. I would love more a variety in samples instead solely jatts but it would be great to get a huge harayana database and compare it with punjabis. You wouldn't think there would be a difference seeing how close the locations are. I know not to get carried away seeing how theres only two or a little more then that but the odds of another high euro score is intriguing at least.

    • Hi Guys , the result for the 2nd Haryana Jat of the two kits provided by Simranjit is depicted by HRP0188

  5. I have calculated the ANI values of non-Brahmin Tamils and listed them below.

    Tamil Vishwakarma HRP0069 44%
    Tamil Vishwakarma HRP0070 43%

    Tamil Nadar HRP0007 47%
    Tamil Nadar HRP0065 49%
    Tamil Nadar HRP0066 47%

    Tamil Muslim HRP0013 50%

    Tamil Vellalar (Sri Lankan) HRP0169 51%

    Tamil Brahmin (1/2), Christian (1/2) HRP0154 53%

    There does appear to be a small but real difference in ANI between the Vishwakarma and the Nadar. Since Tamil Brahmins have ANI of around 60%, the Christian parent of HRP0154 has ANI of 46%.

  6. The ANI of HRP0155 (Sinhalese Govigama) is 49% close to that of HRP0169 (Sri Lankan Tamil Vellalar).

    • Hi Balaji,How do you calculate ANI?

      • I used the following.
        ANI = 32% + SW Asian + European + 0.18 * S Asian
        With this I got close to the same ANI values as Reich et al. did for their reference samples.

    • HRP0031 is Tamil Vellalar Sri Lankan (1/2), Telegu (1/2). This individual has ANI of 46%. If the Vellalar parent's ANI is 51%, the Telugu parent's ANI is 41% about that of a Madiga from Reich et al.

      • Thanks for clarifying Balaji. What would the ASI component be? Would this be the same as Onge?

        • ASI will not be the same as Onge though it is related to it. You can think of ASI as 100% - ANI. However I believe that there is a small East Asian component as well so that ASI = 100% - ANI - EA. However it is hard to get a good estimate for EA.

  7. Zack,

    Do you have a list of all the HGDP samples you excluded?

    I used the search function here, but couldn't find the blog entry about you cleaning up that dataset.

    Also, I just e-mailed M. Metspalu, who was a co-author on that new paper on the genetics of the Caucasus, and it seems the entire dataset might be available on his website within a week or so. He'll let me know when that happens.

    • I used the Rosenberg list for HGDP.

      It's great that the Caucasus data will be online so soon. These guys are pretty good at making their data public. Could you let me know when it's available?

  8. Something's up with the haplogroups spreadsheet. A given participant's haplogroups are being listed for the participant below him/her.

  9. I list below the ANI for all the Tamil Brahmins.

    HRP0016 Tamil Brahmin 60%
    HRP0079 Karnataka Iyengar Brahmin 60%
    HRP0103 Tamil Brahmin Iyer 60%
    HRP0142 Tamil Brahmin 60%
    HRP0133 Tamil Brahmin 60%
    HRP0017 Karnataka Iyengar 59%
    HRP0014 Tamil Brahmin 58%
    HRP0057 Tamil Brahmin 58%
    HRP0048 Tamil Brahmin 58%
    HRP0075 Tamil Brahmin Iyer 58%
    HRP0041 Tamil Brahmin Iyer 58%
    HRP0072 Tamil Brahmin Iyengar 57%
    HRP0084 Tamil Brahmin 57%
    HRP0084 Tamil Brahmin 57%

  10. ANI for non-Tamil Brahmins.

    HRP0033 Rajasthani Brahmin 71%

    HRP0004 Punjabi Brahmin 70%
    HRP0019 Punjabi Brahmin 68%

    HRP0129 UP Brahmin 67%
    HRP0063 UP Brahmin 65%
    HRP0029 UP Brahmin 64%

    HRP0003 Bihari Brahmin 65%

    HRP0077 Bengali Brahmin 63%
    HRP0054 Bengali Brahmin 60%
    HRP0091 Bengali Brahmin 60%

    HRP0047 Mahrashtrian Desastha Brahmin 60%
    HRP0090 Maharashtra/Madhya Pradesh Saraswat 56%

    HRP0104 Oriya Brahmin 58%

    HRP0025 Karnataka Konkani Brahmin 60%
    HRP0092 Karnataka Brahmin 59%
    HRP0088 Karnataka Kannada Brahmin 56%

    HRP0067 Kerala Brahmin 58%

    HRP0124 Kerala Nair 58%

    HRP0026 Goan Catholic Brahmin 57%

    I have included HRP0124 and HRP0024 who are strictly-speaking not brahmins but who clearly have mostly brahmin ancestry.

    • I'm curious, on what basis are you calculating these scores? Has this formula you're using been posted in some sort of supplementary note to the Reich et al study, or did you derive a formula by chance with coincidentally seemed plausible along the lines of the Reich et al figures? I mean, who is to say that the Pathan samples represent the ANI modal population? The Thathai Bhatia (Sindhi Rajput) participant has even less "Onge" than the Reich Pathans. Why are you separating all of the East Eurasian admixture from the ASI - given that ASI is marginally more related to East Eurasians, at least some of the ASI will definitely show up in a granular analysis as, say, SE Asian.

      • Also, HRP0024 is a Hyderabadi Muslim and HRP0124 is a Nair from Kerala. Why on earth would these be considered as having Brahmin ancestry? I can understand this proposition to certain extent in the case of the Nair given the matrimonial unions between Nair-Namboothiris. But they are, for the most part, socially dominant non-Brahmins. But in the case of the Hyderabadi, your assertion makes little sense.

        • I mistyped HRP0024 in my comments but did not include this individual. I included HRP0026 who is a Goan catholic. This person's ancestors must have been brahmin before converting to Christianity.

      • I used ANI = SW Asian + European + 32% + 0.18* S Asian. I got this by using regression on the Ref. 3 data on the HRP website. Below is a comparison to the ANI values published by Reich et al to values calculated from the formula. Note that I don't use the Onge values or East Asian admixture values.

        Reich ANI calc ANI
        bhil 42.90% 46%
        chenchu 40.70% 44%
        kamsali 44.50% 45%
        kashmiri-pandit 70.60% 68%
        kurumba 43.20% 45%
        lodi 49.90% 50%
        madiga 40.60% 43%
        mala 38.80% 42%
        meghawal 60.30% 60%
        naidu 50.10% 49%
        pathan 76.90% 77%
        satnami 43.00% 45%
        sindhi 73.70% 70%
        srivastava 56.40% 56%
        tharu 51.00% 50%
        vaish 62.60% 62%
        velama 54.70% 52%
        vysya 46.20% 45%

  11. ANI of Pakistanis and NW Indians.

    HRP0035 Punjabi (3/4), Egyptian (1/4) 80%
    HRP0158 Punjabi (3/4), Egyptian (1/4) 79%
    HRP0001 Punjabi (3/4), Egyptian (1/4) 78%
    HRP0161 Punjabi (3/4), Egyptian (1/4) 78%
    HRP0170 Haryana Jatt 76%
    HRP0131 UP/Haryana Jatt 76%
    HRP0093 Punjabi Jatt HRP0006 76%
    HRP0021 Kashmiri 76%
    HRP0085 Thathai Bhatia 75%
    HRP0006 Punjabi Jatt 75%
    HRP0062 Sindhi 74%
    HRP0108 Halai Bhatia 74%
    HRP0008 Punjabi Jatt 74%
    HRP0005 Punjabi Jatt 73%
    HRP0126 Punjabi Jatt 73%
    HRP0039 Sindhi (1/2), Balochi (1/2) 72%
    HRP0162 Punjabi (1/2), Gujarati (1/2) 72%
    HRP0086 Punjabi (1/2), Sindhi (1/2) 72%
    HRP0033 Rajasthani Brahmin 71%
    HRP0004 Punjabi Brahmin 70%
    HRP0107 Punjabi 69%
    HRP0125 Punjabi 69%
    HRP0073 Punjabi 69%
    HRP0012 Punjabi 68%
    HRP0019 Punjabi Brahmin 68%
    HRP0136 Punjabi Ramgarhia 68%
    HRP0044 Kashmiri 67%
    HRP0135 Punjabi Pahari Rajput 67%
    HRP0111 Punjabi 67%
    HRP0106 Punjabi 66%
    HRP0130 Gujarati Ganchi 65%
    HRP0064 Punjabi 61%
    HRP0071 Gujarati 54%
    HRP0058 Gujarati 51%
    HRP0068 Gujarati 50%

    The top first people are Zack and his family and their ANI includes the West Eurasian component of their Egyptian ancestry.

  12. ANI for Carribean, Sri Lanka, Bangladesh and Rest of India.

    HRP0122 Sri Lankan 59%
    HRP0169 Tamil Vellalar (Sri Lankan) 51%
    HRP0155 Sinhalese Govigama 49%
    HRP0031 Sri Lankan (1/2), Telegu (1/2) 46%

    HRP0055 Sourastrian 54%

    HRP0067 Kerala Brahmin 58%
    HRP0124 Kerala Nair 58%
    HRP0152 Kerala Syrian Christian 56%
    HRP0038 Kerala Christian 56%
    HRP0117 Kerala Syrian Christian 55%
    HRP0053 Kerala Muslim Rawther 52%

    HRP0025 Karnataka Konkani Brahmin 60%
    HRP0092 Karnataka Brahmin 59%
    HRP0088 Karnataka Kannada Brahmin 56%
    HRP0078 Karnataka 54%
    HRP0109 Karnataka Lingayat 49%

    HRP0024 Andhra Pradesh (Hyderabadi) 57%
    HRP0159 Andhra Pradesh 55%
    HRP0157 Andhra Pradesh (Hyderabad) 53%
    HRP0076 Andhra Pradesh Reddy 51%
    HRP0168 Andhra Pradesh Gouda 48%
    HRP0009 Andhra Pradesh Reddy 48%
    HRP0061 Andhra Pradesh 48%
    HRP0060 Andhra Pradesh 47%

    HRP0090 Maharashtra/Madhya Pradesh Saraswat Brahmin 56%

    HRP0104 Oriya Brahmin 58%

    HRP0003 Bihari Brahmin 65%
    HRP0112 Bihar/Jharkhand Muslim 64%
    HRP0098 Bihari Muslim 63%
    HRP0123 Bihari Muslim 63%
    HRP0137 Bihari Syed 60%
    HRP0032 Bihari Kayastha 53%

    HRP0077 Bengali Brahmin 63%
    HRP0054 Bengali Brahmin 60%
    HRP0091 Bengali Brahmin 60%
    HRP0114 Bengali/Marathi/Goan 54%
    HRP0050 Bengali/Oriya 53%
    HRP0049 Bengali 51%
    HRP0022 Bengali 49%
    HRP0002 Bengali 49%
    HRP0023 Bengali 49%

    HRP0131 UP/Haryana Jatt 76%
    HRP0119 UP Syed 69%
    HRP0129 UP Brahmin 67%
    HRP0063 UP Brahmin 65%
    HRP0029 UP Brahmin 64%
    HRP0051 UP Mixed Indian/Pakistani 64%
    HRP0056 UP/MP/Marathi 63%
    HRP0139 UP Mughal/Pathan 59%
    HRP0052 UP Kayasth 58%

    HRP0028 Caribbean Indian 56%
    HRP0027 Caribbean Indian 52%

  13. Awaiting results from 23andme any time soon . Jat / rajput ancestors Jammu and Punjab Pakistan . In this region terms are interchangeable except we consider ourselves to be exclusively Jat of the Kalyal clan.