Afghan Dataset

A paper, Afghan Hindu Kush: Where Eurasian Sub-Continent Gene Flows Converge by Julie Di Cristofaro, Erwan Pennarun, Stéphane Mazières, Natalie M. Myres, Alice A. Lin, Shah Aga Temori, Mait Metspalu, Ene Metspalu, Michael Witzel, Roy J. King, Peter A. Underhill, Richard Villems, Jacques Chiaroni was published at PLoS One about the genetics of the people of Afghanistan.

Thanks to Mait Metspalu, the data is available online. It consists of:

  • 5 Hazara
  • 5 Pashtun
  • 5 Tajik
  • 4 Turkmen
  • 5 Uzbek

Here are the HarappaWorld Admixture results for the samples in this dataset.

You can check the spreadsheet too.

Tadjik1_44Af and Pashtun2_6Af seem to be outliers and there's a possibility they are mislabeled. I would like to look into these two samples further before I calculate group averages.

You can compare these Pashtun results to HGDP Pathan and HAP Pashtun results.


  1. Looks like paulgill was right, these are very similar to HRP0370 overall (at least across the South Asian, Caucasian, and European components), but there is some variety in the Arctic (Siberian/Beringian/Amerindian) and East Asian figures. These have more NE-Asian and HRP0370 has slightly more SW-Asian.

    • These Afghan samples were collected from the North and East of the country which might make sense of that. HRP0370 is a Durrani/Popalzai Pashtun from Kandahar and that tribe prior to that spent some time in Herat, so they're mostly south/southwestern.

      • AD, what is your opinion about the absence of rajasthani-jatts on these Afghan oracles? I'm curious how its possible that punjabi-jatts and haryana-jatts feature on pashtun and tajik population sharing, but, the rajasthani-jatts are nowhere to be seen?

        Could this mean that the rajasthani-jatts share less cultural history with Central Asians, compared to the other two jatt clans? Or maybe they have an altogether different source of origin?

        Also, I am quite surprised at the near absence of the UP/Delhi muslims in these oracles. Was thinking they would come out top, because the Harappa results prove they have actually preserved their Pashtun heritage! What could be the reasonable explanation for this, I wonder?

        • You misunderstand what Oracle does.

          • Thanks zack. I read your posts on Oracle limitations and understand better now. So, it seems this oracle is accurate only so far as when it comes to informing south asians how close in terms of genetics/genetic components they are to different ethnic groups, and allowing comparisons to be made. So basically, for the afghans who have haryana-jatts at the top of their oracle, it means they share those similar components with each other for whatever reason, but not necessarily to say they share the same exact ancestry or history, right? otherwise, the UP/Pashtuns from India who still have large amounts of Islamic Central Asian ancestry intact would be much closer than the hindu jatts on population sharing.

            in other words, for south asians, this Oracle is good for determining racial purity served by endogamy practices, but not so great for Indians, especially muslims, who have mixed ancestry. O.k., interesting!

          • Close, but I would add that it's also good for figuring out geographical shifts in ancestry signals. I'm not a fan of the term "racial purity" because I don't think there is such a thing, it seems all peoples form a genetic gradient.

        • Rajasthani and Haryanvi jatts are more or less similar genetically. You need to check spreadsheet.

  2. Do you feel Uzbek8_18Af seems to be bit off because of his NE component? Seems more close to Pashtuns.

  3. Awesome, thank you so much Zack! I really appreciate this.

    For anyone interested in the anthropological details, these Pashtuns originate from the provinces of Baghlan and Kunduz, far up north in Afghanistan.They are undoubtedly Ghilzai Pashtuns, the largest Pashtun confederacy in Afghanistan. Baghlan and Kunduz are rather distant from the Pashtun heartland, and these Pashtuns are very recent arrivals, only in the last 150-100 years (and in some cases, only in the last 50 years). They originally came from Loya Paktia, although there are some Kandahari people as well, but not so much in Baghlan and Kunduz. These Pashtuns were exiled to this region by Amir Abdur Rahman. For those of us interested in genetic affinities, we should note something very important. When these Pashtuns arrived in the region, there was some heavy intermarriage with Tajiks and Uzbeks. At first, they tried to remain endogamous, but their relatives down south weren't interested in sending daughters this far up north.

    Although, I'm not sure we see any evidence of this in the results. These Pashtuns don't seem to be highly admixed with Tajiks and Uzbeks, contrary to what anthropologists claim. Although, we'd need samples from Paktia to verify this. Pashtuns from Loya Paktia are very similar to Pashtuns from the Kurram Valley, in terms of both culture and phenotype. In fact, the Kurram Valley is an extension of Loya Paktia! So, if Ghilzai from the heartland turn out to be similar to the HGDP Pashtuns, then we can be sure of some admixture, just like what anthropologists have documented.

    Unfortunately, I don't think samples from the Pashtun heartland are coming anytime soon. In fact, the only reason they sampled isolated Pashtun communities from the north is due to the situation in the Pashtun south being so messed up.

    Zack, perhaps you could do a quick ChromoPainter/FineStructure run? All of the HGDP Pakistani samples+HGDP Hazara, Yunusbayev Tajiks, Behar Uzbeks, Yunusbayev Turkmens, and all of your Afghan-Tajik/Central Asian-Pakistani-Northwest Indian participants. In short, just Central Asians+Northwest South Asians, and no other populations and participants. This would be much quicker, and it would settle the issue of whether Tadjik1_44Af and Pashtun2_6Af are mislabeled. I really hope you do consider this, it would be sincerely appreciated by all of us. Thank you.

  4. Also, here are Tadjik1_44Af Oracle results:

    [1,] "pashtun_harappa_3" "5.022"
    [2,] "pathan_hgdp_23" "11.0842"
    [3,] "kalash_hgdp_23" "11.4926"
    [4,] "burusho_hgdp_25" "12.8513"
    [5,] "tajik_yunusbayev_15" "13.6317"
    [6,] "kashmiri_harappa_2" "14.5434"
    [7,] "bhatia_harappa_2" "15.5462"
    [8,] "gujarati-muslim_harappa_3" "15.914"
    [9,] "punjabi-jatt_harappa_8" "16.3542"
    [10,] "kashmiri-pandit_reich_5" "16.7199"
    [11,] "haryana-jatt_harappa_5" "16.7514"
    [12,] "turkmen_yunusbayev_11" "17.2561"
    [13,] "punjabi_harappa_10" "18.421"
    [14,] "punjabi-arain_xing_25" "18.4525"
    [15,] "sindhi_hgdp_24" "18.7546"
    [16,] "punjabi-brahmin_harappa_2" "18.9876"
    [17,] "singapore-indian-c_sgvp_10" "19.3387"
    [18,] "bene-israel-jew_behar_4" "19.3647"
    [19,] "kashmiri-pahari_harappa_2" "19.4284"
    [20,] "punjabi-ramgarhia_harappa_2" "20.4054"


    [1,] "tajik_yunusbayev_15" "11.406"
    [2,] "turkmen_yunusbayev_11" "14.5357"
    [3,] "pashtun_harappa_3" "15.6957"
    [4,] "uzbek_behar_15" "15.8167"
    [5,] "burusho_hgdp_25" "19.3148"
    [6,] "hazara_hgdp_22" "19.7037"
    [7,] "pathan_hgdp_23" "21.4015"
    [8,] "uyghur_hgdp_10" "21.8867"
    [9,] "kalash_hgdp_23" "22.425"
    [10,] "haryana-jatt_harappa_5" "23.737"
    [11,] "kashmiri_harappa_2" "24.0714"
    [12,] "bhatia_harappa_2" "24.9143"
    [13,] "punjabi-jatt_harappa_8" "25.1523"
    [14,] "gujarati-muslim_harappa_3" "25.6342"
    [15,] "nogai_yunusbayev_16" "25.769"
    [16,] "kashmiri-pandit_reich_5" "26.119"
    [17,] "romanian-b_behar_2" "26.5135"
    [18,] "nepalese-c_xing_7" "26.8667"
    [19,] "nepalese-a_xing_12" "27.145"
    [20,] "punjabi_harappa_10" "27.3047"

    Tadjik1_44Af sure isn't very Tajik-like. To be honest though, Tadjik1_44Af is quite different from the other Pashtuns as well. I'd wait till we see the ChromoPainter/FineStructure results. Should prove very interesting.

    • Try mixed mode oracle. Both these samples seem very different from the others.

      • To be honest, I'm not sure what to think, both these samples have rather interesting results:

        [1,] "30.4% kyrgyz_hodoglugil_16 + 69.6% pashtun_harappa_3" "3.6668"
        [2,] "31.4% kyrgyz_xing_25 + 68.6% pashtun_harappa_3" "3.6882"
        [3,] "32.9% kazakh_harappa_5 + 67.1% pashtun_harappa_3" "3.7903"
        [4,] "44% hazara_hgdp_22 + 56% pashtun_harappa_3" "3.9242"
        [5,] "25.4% mongolian_rasmussen_9 + 74.6% pashtun_harappa_3" "3.9371"
        [6,] "50.2% pashtun_harappa_3 + 49.8% uzbek_behar_15" "3.9931"
        [7,] "42.1% pathan_hgdp_23 + 57.9% uzbek_behar_15" "4.0513"
        [8,] "40.9% kalash_hgdp_23 + 59.1% uzbek_behar_15" "4.1994"
        [9,] "35.4% sindhi_hgdp_24 + 64.6% uzbek_behar_15" "4.294"
        [10,] "34.6% balochi_hgdp_24 + 65.4% uzbek_behar_15" "4.5143"
        [11,] "38.2% bhatia_harappa_2 + 61.8% uzbek_behar_15" "4.5558"
        [12,] "27.7% altaian_rasmussen_13 + 72.3% pashtun_harappa_3" "4.6836"
        [13,] "20.8% oroqen_hgdp_9 + 79.2% pashtun_harappa_3" "4.6941"
        [14,] "35.4% punjabi-arain_xing_25 + 64.6% uzbek_behar_15" "4.8856"
        [15,] "22.4% buryat_xing_25 + 77.6% pashtun_harappa_3" "5.0631"
        [16,] "58.9% pashtun_harappa_3 + 41.1% uyghur_hgdp_10" "5.2159"
        [17,] "21.9% buryat_rasmussen_18 + 78.1% pashtun_harappa_3" "5.3022"
        [18,] "37.7% punjabi-jatt_harappa_8 + 62.3% uzbek_behar_15" "5.3562"
        [19,] "33.8% makrani_hgdp_25 + 66.2% uzbek_behar_15" "5.3735"
        [20,] "31.9% brahui_hgdp_25 + 68.1% uzbek_behar_15" "5.4368"

        [1,] "48.3% punjabi-arain_xing_25 + 51.7% turkmen_yunusbayev_11" "2.8393"
        [2,] "27.2% armenian_harappa_2 + 72.8% burusho_hgdp_25" "3.1486"
        [3,] "70.9% burusho_hgdp_25 + 29.1% turkish_harappa_4" "3.1894"
        [4,] "46.9% kashmiri-pahari_harappa_2 + 53.1% turkmen_yunusbayev_11" "3.3569"
        [5,] "50.8% kashmiri-pandit_reich_5 + 49.2% turkmen_yunusbayev_11" "3.3737"
        [6,] "72.8% burusho_hgdp_25 + 27.2% turk-kayseri_hodoglugil_23" "3.4088"
        [7,] "77.1% burusho_hgdp_25 + 22.9% lebanese-druze_haber_25" "3.4453"
        [8,] "29.4% azeri_harappa_2 + 70.6% burusho_hgdp_25" "3.4619"
        [9,] "73% burusho_hgdp_25 + 27% turk_behar_19" "3.4951"
        [10,] "23.9% azerbaijan-jew_behar_8 + 76.1% burusho_hgdp_25" "3.5129"
        [11,] "48.3% punjabi_harappa_10 + 51.7% turkmen_yunusbayev_11" "3.5164"
        [12,] "54.5% kashmiri_harappa_2 + 45.5% turkmen_yunusbayev_11" "3.5433"
        [13,] "47% singapore-indian-c_sgvp_10 + 53% turkmen_yunusbayev_11" "3.5521"
        [14,] "23.5% armenian_yunusbayev_16 + 76.5% burusho_hgdp_25" "3.5807"
        [15,] "76.4% burusho_hgdp_25 + 23.6% georgia-jew_behar_4" "3.6085"
        [16,] "45.7% punjabi-ramgarhia_harappa_2 + 54.3% turkmen_yunusbayev_11" "3.6197"
        [17,] "79.8% burusho_hgdp_25 + 20.2% druze_hgdp_42" "3.6305"
        [18,] "47.8% sindhi_hgdp_24 + 52.2% turkmen_yunusbayev_11" "3.6391"
        [19,] "47.5% punjabi-brahmin_harappa_2 + 52.5% turkmen_yunusbayev_11" "3.6771"
        [20,] "77.9% burusho_hgdp_25 + 22.1% cypriot_behar_12" "3.6835"

        Zack, I'd say you're the expert, any thoughts on this?

        • Perhaps run the Oracle on other Pashtuns and Tajiks as well and compare the results?

          • Sounds like a great idea.

            [1,] "pashtun_harappa_3" "5.7075"
            [2,] "tajik_yunusbayev_15" "10.2237"
            [3,] "pathan_hgdp_23" "12.338"
            [4,] "kalash_hgdp_23" "13.0999"
            [5,] "burusho_hgdp_25" "13.8984"
            [6,] "turkmen_yunusbayev_11" "15.6134"
            [7,] "haryana-jatt_harappa_5" "16.6"
            [8,] "kashmiri_harappa_2" "16.7805"
            [9,] "bhatia_harappa_2" "16.8362"
            [10,] "punjabi-jatt_harappa_8" "17.5013"
            [11,] "kashmiri-pandit_reich_5" "18.7101"
            [12,] "gujarati-muslim_harappa_3" "18.7698"
            [13,] "punjabi_harappa_10" "20.3215"
            [14,] "punjabi-arain_xing_25" "20.6276"
            [15,] "punjabi-brahmin_harappa_2" "20.7143"
            [16,] "sindhi_hgdp_24" "20.9396"
            [17,] "singapore-indian-c_sgvp_10" "21.1201"
            [18,] "kashmiri-pahari_harappa_2" "21.5074"
            [19,] "nepalese-a_xing_12" "21.8826"
            [20,] "punjabi-ramgarhia_harappa_2" "22.1638"

            [1,] "47.1% punjabi-jatt_harappa_8 + 52.9% turkmen_yunusbayev_11" "2.8552"
            [2,] "84.3% pashtun_harappa_3 + 15.7% uzbek_behar_15" "3.0882"
            [3,] "42.1% nogai_yunusbayev_16 + 57.9% sindhi_hgdp_24" "3.2982"
            [4,] "9.9% kazakh_harappa_5 + 90.1% pashtun_harappa_3" "3.3888"
            [5,] "67.9% pashtun_harappa_3 + 32.1% tajik_yunusbayev_15" "3.4406"
            [6,] "34.3% gujarati-muslim_harappa_3 + 65.7% tajik_yunusbayev_15" "3.4709"
            [7,] "8.9% kyrgyz_hodoglugil_16 + 91.1% pashtun_harappa_3" "3.5404"
            [8,] "9.2% kyrgyz_xing_25 + 90.8% pashtun_harappa_3" "3.557"
            [9,] "12.9% hazara_hgdp_22 + 87.1% pashtun_harappa_3" "3.5584"
            [10,] "8.2% altaian_rasmussen_13 + 91.8% pashtun_harappa_3" "3.5701"
            [11,] "56.3% pathan_hgdp_23 + 43.7% turkmen_yunusbayev_11" "3.5762"
            [12,] "24.8% cochin-jew_behar_4 + 75.2% tajik_yunusbayev_15" "3.5991"
            [13,] "7.3% mongolian_rasmussen_9 + 92.7% pashtun_harappa_3" "3.6527"
            [14,] "6.5% buryat_rasmussen_18 + 93.5% pashtun_harappa_3" "3.6614"
            [15,] "6.5% buryat_xing_25 + 93.5% pashtun_harappa_3" "3.7027"
            [16,] "48.4% haryana-jatt_harappa_5 + 51.6% turkmen_yunusbayev_11" "3.7038"
            [17,] "93.2% pashtun_harappa_3 + 6.8% tuvinian_rasmussen_14" "3.7123"
            [18,] "22.1% kerala-christian_harappa_7 + 77.9% tajik_yunusbayev_15" "3.7701"
            [19,] "88.2% pashtun_harappa_3 + 11.8% uyghur_hgdp_10" "3.8105"
            [20,] "42.6% punjabi-brahmin_harappa_2 + 57.4% turkmen_yunusbayev_11" "3.8429"

            [1,] "pashtun_harappa_3" "4.6476"
            [2,] "kalash_hgdp_23" "11.4737"
            [3,] "pathan_hgdp_23" "11.6546"
            [4,] "tajik_yunusbayev_15" "11.7033"
            [5,] "burusho_hgdp_25" "15.1145"
            [6,] "haryana-jatt_harappa_5" "16.033"
            [7,] "bhatia_harappa_2" "16.3264"
            [8,] "kashmiri_harappa_2" "16.7179"
            [9,] "punjabi-jatt_harappa_8" "17.1222"
            [10,] "turkmen_yunusbayev_11" "17.2124"
            [11,] "kashmiri-pandit_reich_5" "18.576"
            [12,] "gujarati-muslim_harappa_3" "18.8482"
            [13,] "punjabi_harappa_10" "20.3578"
            [14,] "punjabi-brahmin_harappa_2" "20.4347"
            [15,] "punjabi-arain_xing_25" "20.4765"
            [16,] "sindhi_hgdp_24" "20.7708"
            [17,] "singapore-indian-c_sgvp_10" "21.0645"
            [18,] "kashmiri-pahari_harappa_2" "21.5697"
            [19,] "punjabi-ramgarhia_harappa_2" "22.3264"
            [20,] "bene-israel-jew_behar_4" "22.3675"

            [1,] "87% pashtun_harappa_3 + 13% urkarah_xing_18" "2.4413"
            [2,] "86.5% pashtun_harappa_3 + 13.5% stalskoe_xing_5" "2.5391"
            [3,] "12.1% lezgin_behar_18 + 87.9% pashtun_harappa_3" "2.6239"
            [4,] "10.2% chechen_yunusbayev_20 + 89.8% pashtun_harappa_3" "2.8634"
            [5,] "10.6% kumyk_yunusbayev_14 + 89.4% pashtun_harappa_3" "3.0826"
            [6,] "8.4% adygei_hgdp_17 + 91.6% pashtun_harappa_3" "3.1788"
            [7,] "8.5% balkar_yunusbayev_19 + 91.5% pashtun_harappa_3" "3.2119"
            [8,] "8.3% north-ossetian_yunusbayev_15 + 91.7% pashtun_harappa_3" "3.26"
            [9,] "6.8% bulgarian_yunusbayev_13 + 93.2% pashtun_harappa_3" "3.5205"
            [10,] "78.7% pashtun_harappa_3 + 21.3% tajik_yunusbayev_15" "3.5301"
            [11,] "91.3% pashtun_harappa_3 + 8.7% turk-istanbul_hodoglugil_20" "3.5643"
            [12,] "9.2% nogai_yunusbayev_16 + 90.8% pashtun_harappa_3" "3.5738"
            [13,] "6% abhkasian_yunusbayev_20 + 94% pashtun_harappa_3" "3.5767"
            [14,] "93.3% pashtun_harappa_3 + 6.7% romanian-a_behar_14" "3.5926"
            [15,] "5.6% georgian_behar_20 + 94.4% pashtun_harappa_3" "3.6271"
            [16,] "55.2% punjabi-arain_xing_25 + 44.8% stalskoe_xing_5" "3.6396"
            [17,] "91.6% pashtun_harappa_3 + 8.4% turk-aydin_hodoglugil_20" "3.6514"
            [18,] "91.9% pashtun_harappa_3 + 8.1% turk_behar_19" "3.6581"
            [19,] "6.5% georgian_harappa_4 + 93.5% pashtun_harappa_3" "3.664"
            [20,] "74.1% kalash_hgdp_23 + 25.9% turk-aydin_hodoglugil_20" "3.6776"

            [1,] "tajik_yunusbayev_15" "8.3449"
            [2,] "pashtun_harappa_3" "11.0204"
            [3,] "turkmen_yunusbayev_11" "11.612"
            [4,] "burusho_hgdp_25" "18.0774"
            [5,] "pathan_hgdp_23" "18.3118"
            [6,] "kalash_hgdp_23" "18.939"
            [7,] "uzbek_behar_15" "20.4562"
            [8,] "haryana-jatt_harappa_5" "21.0964"
            [9,] "kashmiri_harappa_2" "21.3356"
            [10,] "romanian-b_behar_2" "21.9237"
            [11,] "bhatia_harappa_2" "22.6768"
            [12,] "punjabi-jatt_harappa_8" "22.827"
            [13,] "gujarati-muslim_harappa_3" "23.1701"
            [14,] "iranian_harappa_9" "23.1742"
            [15,] "nogai_yunusbayev_16" "23.5664"
            [16,] "kashmiri-pandit_reich_5" "23.6519"
            [17,] "iranian_behar_19" "24.0936"
            [18,] "stalskoe_xing_5" "24.104"
            [19,] "bene-israel-jew_behar_4" "24.6344"
            [20,] "kurd_harappa_11" "25.0999"

            [1,] "66% pashtun_harappa_3 + 34% uzbek_behar_15" "3.7294"
            [2,] "29.9% nepalese-a_xing_12 + 70.1% turkmen_yunusbayev_11" "3.8155"
            [3,] "14% great-andamanese_reich_4 + 86% tajik_yunusbayev_15" "3.9039"
            [4,] "12.3% onge_reich_9 + 87.7% tajik_yunusbayev_15" "4.0687"
            [5,] "9.3% pulliyar_metspalu_5 + 90.7% tajik_yunusbayev_15" "4.2066"
            [6,] "10.1% santhal_reich_7 + 89.9% tajik_yunusbayev_15" "4.2234"
            [7,] "10% ho_chaubey_5 + 90% tajik_yunusbayev_15" "4.228"
            [8,] "9.1% paniya_chaubey_3 + 90.9% tajik_yunusbayev_15" "4.234"
            [9,] "10% kharia_reich_4 + 90% tajik_yunusbayev_15" "4.2477"
            [10,] "10% savara_chaubey_2 + 90% tajik_yunusbayev_15" "4.2533"
            [11,] "10.2% asur_chaubey_2 + 89.8% tajik_yunusbayev_15" "4.3014"
            [12,] "9.3% irula_xing_22 + 90.7% tajik_yunusbayev_15" "4.3018"
            [13,] "10% kharia_chaubey_2 + 90% tajik_yunusbayev_15" "4.3428"
            [14,] "34.2% haryana-jatt_harappa_5 + 65.8% turkmen_yunusbayev_11" "4.4467"
            [15,] "10% juang_chaubey_2 + 90% tajik_yunusbayev_15" "4.467"
            [16,] "9.9% bonda_chaubey_4 + 90.1% tajik_yunusbayev_15" "4.4907"
            [17,] "10.4% nihali_metspalu_2 + 89.6% tajik_yunusbayev_15" "4.5079"
            [18,] "56.9% burusho_hgdp_25 + 43.1% nogai_yunusbayev_16" "4.5145"
            [19,] "11.3% sahariya_reich_4 + 88.7% tajik_yunusbayev_15" "4.5766"
            [20,] "22.3% bene-israel-jew_behar_4 + 77.7% tajik_yunusbayev_15" "4.6098"

            [1,] "turkmen_yunusbayev_11" "6.1125"
            [2,] "tajik_yunusbayev_15" "12.6561"
            [3,] "pashtun_harappa_3" "17.9006"
            [4,] "uzbek_behar_15" "17.9885"
            [5,] "iranian_harappa_9" "18.6082"
            [6,] "iranian_behar_19" "19.4176"
            [7,] "nogai_yunusbayev_16" "19.5018"
            [8,] "kurd_harappa_11" "20.1934"
            [9,] "azeri_harappa_2" "21.7867"
            [10,] "turk-aydin_hodoglugil_20" "21.8785"
            [11,] "turkish_harappa_4" "21.9728"
            [12,] "stalskoe_xing_5" "22.4985"
            [13,] "iraqi-arab_harappa_4" "23.0893"
            [14,] "romanian-b_behar_2" "23.1431"
            [15,] "kurd_xing_24" "23.1678"
            [16,] "kurd_yunusbayev_6" "23.3239"
            [17,] "hazara_hgdp_22" "23.3478"
            [18,] "turk-istanbul_hodoglugil_20" "23.4212"
            [19,] "turk-kayseri_hodoglugil_23" "24.0535"
            [20,] "kumyk_yunusbayev_14" "24.2738"

            [1,] "13.4% nepalese-c_xing_7 + 86.6% turkmen_yunusbayev_11" "3.531"
            [2,] "9.2% nepalese-b_xing_6 + 90.8% turkmen_yunusbayev_11" "3.5967"
            [3,] "8.5% great-andamanese_reich_4 + 91.5% turkmen_yunusbayev_11" "4.13"
            [4,] "7.7% garo_chaubey_4 + 92.3% turkmen_yunusbayev_11" "4.1776"
            [5,] "16.2% hazara_hgdp_22 + 83.8% turkmen_yunusbayev_11" "4.2031"
            [6,] "7.4% onge_reich_9 + 92.6% turkmen_yunusbayev_11" "4.2303"
            [7,] "8.8% khasi_chaubey_3 + 91.2% turkmen_yunusbayev_11" "4.246"
            [8,] "8% burmanese_chaubey_15 + 92% turkmen_yunusbayev_11" "4.2892"
            [9,] "85% turkmen_yunusbayev_11 + 15% uyghur_hgdp_10" "4.4174"
            [10,] "6.2% hezhen_hgdp_9 + 93.8% turkmen_yunusbayev_11" "4.5482"
            [11,] "6.2% daur_hgdp_9 + 93.8% turkmen_yunusbayev_11" "4.6028"
            [12,] "8.7% bengali_harappa_8 + 91.3% turkmen_yunusbayev_11" "4.6074"
            [13,] "6.2% mongola_hgdp_10 + 93.8% turkmen_yunusbayev_11" "4.6247"
            [14,] "5.3% tibet_simonson_31 + 94.7% turkmen_yunusbayev_11" "4.6431"
            [15,] "6.2% oroqen_hgdp_9 + 93.8% turkmen_yunusbayev_11" "4.6529"
            [16,] "9.8% kyrgyz_xing_25 + 90.2% turkmen_yunusbayev_11" "4.6645"
            [17,] "94.1% turkmen_yunusbayev_11 + 5.9% xibo_hgdp_9" "4.6745"
            [18,] "5% nysha_reich_4 + 95% turkmen_yunusbayev_11" "4.6918"
            [19,] "81.6% turkmen_yunusbayev_11 + 18.4% uzbek_behar_15" "4.7012"
            [20,] "5.4% naxi_hgdp_7 + 94.6% turkmen_yunusbayev_11" "4.7121"

            Seems reasonable. The other Pashtuns are closest to HAP Pashtuns, and pretty close to the HGDP Pashtuns. Other Tajiks are closest to the Yunusbayev Tajiks or Turkmens.

          • And if it might help, my Oracle results:

            [1,] "pashtun_harappa_3" "2.7793"
            [2,] "pathan_hgdp_23" "8.1267"
            [3,] "kalash_hgdp_23" "9.6295"
            [4,] "burusho_hgdp_25" "10.6789"
            [5,] "kashmiri_harappa_2" "11.5076"
            [6,] "bhatia_harappa_2" "12.6449"
            [7,] "punjabi-jatt_harappa_8" "12.6541"
            [8,] "haryana-jatt_harappa_5" "12.9978"
            [9,] "gujarati-muslim_harappa_3" "13.0915"
            [10,] "kashmiri-pandit_reich_5" "13.3197"
            [11,] "punjabi_harappa_10" "14.8043"
            [12,] "punjabi-brahmin_harappa_2" "15.3529"
            [13,] "punjabi-arain_xing_25" "15.4933"
            [14,] "singapore-indian-c_sgvp_10" "15.8319"
            [15,] "tajik_yunusbayev_15" "15.9154"
            [16,] "sindhi_hgdp_24" "15.9538"
            [17,] "kashmiri-pahari_harappa_2" "16.2257"
            [18,] "punjabi-ramgarhia_harappa_2" "17.0701"
            [19,] "nepalese-a_xing_12" "17.4427"
            [20,] "bene-israel-jew_behar_4" "18.5617"

            [1,] "7.2% kerala-nair_harappa_3 + 92.8% pashtun_harappa_3" "1.651"
            [2,] "5.8% ap-reddy_harappa_4 + 94.2% pashtun_harappa_3" "1.6601"
            [3,] "92.7% pashtun_harappa_3 + 7.3% rajasthani_harappa_4" "1.6796"
            [4,] "7.3% kerala-christian_harappa_7 + 92.7% pashtun_harappa_3" "1.6868"
            [5,] "6.4% kerala-muslim_harappa_2 + 93.6% pashtun_harappa_3" "1.689"
            [6,] "5.9% gujarati-a_1000genomes_5 + 94.1% pashtun_harappa_3" "1.7028"
            [7,] "94.7% pashtun_harappa_3 + 5.3% tamil-nadar_harappa_3" "1.7066"
            [8,] "5.6% gujarati-patel_harappa_2 + 94.4% pashtun_harappa_3" "1.7079"
            [9,] "5.8% gujarati-a_hapmap_63 + 94.2% pashtun_harappa_3" "1.7089"
            [10,] "7.6% gujarati_harappa_3 + 92.4% pashtun_harappa_3" "1.7091"
            [11,] "7% ap-brahmin_xing_25 + 93% pashtun_harappa_3" "1.7137"
            [12,] "7.6% maharashtrian_harappa_4 + 92.4% pashtun_harappa_3" "1.7149"
            [13,] "93.7% pashtun_harappa_3 + 6.3% tamil_harappa_3" "1.7151"
            [14,] "94.6% pashtun_harappa_3 + 5.4% tamil-vellalar_harappa_4" "1.7161"
            [15,] "7.2% iyengar-brahmin_harappa_7 + 92.8% pashtun_harappa_3" "1.7187"
            [16,] "5.1% naidu_reich_4 + 94.9% pashtun_harappa_3" "1.7232"
            [17,] "94.3% pashtun_harappa_3 + 5.7% velama_reich_3" "1.7235"
            [18,] "7.7% meghawal_reich_5 + 92.3% pashtun_harappa_3" "1.7264"
            [19,] "94.4% pashtun_harappa_3 + 5.6% velama_metspalu_10" "1.7267"
            [20,] "7.5% brahmin-tamil-nadu_metspalu_2 + 92.5% pashtun_harappa_3" "1.7278"

      • Zack, trouble with these oracles is that one have no option to choose the distance, percentage and population base one wants to compare two kits for, so these results mostly are useless for comparison.

    • who, or what, are the bhatia_harappa? they feature prominent on these oracles.

      • Bhatia are some caste from Rajahistan, I am surprised they are only 25% south indian component, though again they are only 2 samples. Same thing with Haryana Jatts, they need to be sampled more to get proper conclusion.

        • Bhatia are related to Rajputs and Jatts, it's possible their ancestry is linked to Bhattis. These castes are pretty famously endogamous so their having genetics closer to Punjabis is not that surprising. Plus when they have mixed, it's been with Muslim groups usually so their genes are probably shifted to the West rather than the East due to that.

          • O.k. thanks, but, I read zack announced before the two bhatia_harappa samples are sindhi and gujarati? religion-wise, I think they are split between both hindu and muslim sects. I was just confused on why wikipedia said they are from Rajasthan, because the donors are labeled otherwise.

            Yes I think you're right AD, the south asians who consistently have an order of importance on these oracles seem to be the Northwest Indians who are the most endogamous, and it looks like the haryana-jatts who I presume are hindu, outrank the punjabi-jatts (sikh and muslim) not because they share a closer sense of geography with central asians or as some people wrongly think they must have mixed recently with them, but, because they practice a stronger sense of cultural endogamy related to hinduism, so that is it. But other than that looks like this Oracle is not good for people with a mix of ancestry because it doesn't always pick out real ancestry in mixed people, as zack mentioned in one of his earlier posts.

          • Haryana jat samples are too small, I agree these paticular samples are showing more outside influence, but something has to explain this. I really doubt all Haryana jats would show this

            In terms of phenotype, an average Harayana jat is much more (Indic) looking then even a sikh jatt. There are lots of Haryana jats who are marrying girls from outside, even from south india, so their strong cultural endogamy seems to be a myth

          • I don't really know what you mean by endogamous? Yes generally they do marry within the ethnic group but they never marry within the same clan. Generally there should be no relation within the last 6 generation married into the proposed clan(I mean the whole clan not just the family), now I think it is reduced to 3 generations before one can marry into a particular clan. For example I am Gill, Mother Mann, Grandmothers both sides Mahil+Bal and Virk+Wraich great grandmothers Bains+Dhillon, Sandhu+Sidhu, I can't marry into any of those clans.

            Trouble with Jatts is that they often kill their infant girls generally at birth, this now been going on for ages and that results in shortage of Jatt girls and then whomsoever was not able to get a Jatt wife will go and marry one from outside of the Jatt ethnic group. That is the basic reason why they have so much of south Indian and so low NE European component now. And the way it is going on now, it looks like that soon all Jatts be more close to south Indians than anyone else.

          • @Pakistani, looks like that you are back into your phenotypes, here we talk about genotype not phenotype.

            It is said by experts that Jatts came from beyond Afghanistan and Iran through Sindh > Rajsthan > Haryana > Punjab. So the Jatts that you find in Punjab are probably from the earlier invasions thus are in India for longer time.

            Someone have found some people in Sri Lanka and Kerla who are most closely related to Punjabi Jatts and then to the Haryanvi Jatts and they are also related to Croatians, true or not, I don't know, but these are the Buddhist people who were once sent by Ashoka to Sri Lanka, and like Jatts they still are not under Hinduism. Hindu Jatts are not Hindus but Arya Samaji. In Arya Samaj the caste system is abolished and that is the reason why you find Jatts in there.

            Now back to these Buddhists. If they really are Jatts originally from Punjab then the possibility is that Chander Gupta Maurya who was from Punjab was actually a Jatt and again most likely from the Jatts who were mercenaries with Alexander the Great when he invaded India.

          • Paul: Those are practices of Sikh Jatts to my knowledge. Muslim Jatts generally marry other Jatts but also have no qualms about marrying within the family (though marrying within the same clan is rare because most families are multi-clan, mine is mostly Gill, Pansota, Balham, for instance, they have married cousins, but not cousins of the same paternal clan lineage), in line with Islamic custom since Muslim South Asians are big on marrying in the family. In some Muslim Jatt clans a lot of inbreeding has occurred. They also don't engage in infanticide so there isn't a shortage of Jatt women in Pakistan (the opposite I would say, most people seem to be having more girls than boys nowadays). When they do marry outside their caste, if ever, it's probably with Rajput, Khatri/Kamboj/etc, Arain, Pathan, in that order of preference. A lot of mixed marriage is now happening (in the past 20 years) but the genetic results of that won't be seen for another generation. But until that, marrying even non-Jatt women was a big deal and pretty frowned upon.

            I think there was a period of mixed marriages right after some of these groups converted to Islam, then they reverted to their endogamy. That could be why Muslim Jatts average higher Caucasian and lower NE-Euro, or it might just be a geographical quirk, since a lot of Sikh Jatts do hail from lands which are now in Pakistan (i.e, to the West) and Muslim Jatts often hail from lands now in India (to the East), so that's why the gradient got messed up, because of partition in '47. Only one side of my family is local to Pakistan, the Balham clan, which hail from Multan (and part of that was Bhatti Jatt). But my Gill paternal grandfather came from Hoshiarpur and my Pansota ancestors before my great-great-grandparents also (at least it says on Jatt history sites) emigrated from Hoshiarpur to that South Punjab area (which is now mostly Doabi-dialect speaking).

            I don't know what the situation is with Hindu Jatts, whether they marry within the same clan or family or not.

          • Death penalty for marrying within the same clan for both is still in effect, the Jatt Law.

          • Not among the Muslim Jatts in Pakistan at least. I can tell you from experience they don't care about that and don't even know such a law ever existed (or yet exists among their Sikh cousins in India).

          • Sikh or Hindu they have the same Jatt Law.

  5. Are all of the the Lazaridis et al. data publically available? There are some interesting new samples there.

    • Nope. The interesting ones will only be available once the paper is printed in a journal. You will need to sign a form.

  6. Hazara6_66Af is also very interesting. Seems more like the Pashtuns than the other Hazara.

  7. I find it difficult to believe the Uzbeks will have significant amount of South Indian. Does not make sense.

    • Its actually really plausible, take for example R1a1-Y7+. Y7+ is possibly South Asian and migrated out of the region into the Middle East and into Kazakhstan. So there's high possibility that there were constant large migrations out of South Asian into Uzbekistan and into central asia for many years and the last large group being the Romani.

      These are just my thoughts, and they are probably going to change in a few weeks with more reading.

    • If you compare these Afghan Uzbeks to those from Uzbekistan, you can see that the Afghan ones are more South Indian.

    • Would you mind to identify yourself, please. Here we have a refugee Pushtun from U.P. now living in Karachi, I guess. The guy thinks that it is an insult to belong to India and does not realize that if that land had not supported his ancestors then he was not in this world today, what a thankless creature he is. I hope you have nothing to do with him.

  8. Are the South Indian, Baloch, Caucasian, North Euro actual ancestral populations or made up? If they are made up, who are they supposed to be most closely be related to? South India is obvious; how about the others?

  9. ^ south indian is clearly a very ancient component, the reason it is found in afghans at the 10-15% range is because it probably existed in the afghan region before any of the other components arrived. After all that component is related to a population like Andaman Islanders, who are a very ancient people

    Over all though this study clearly shows that afghans are related to each other, then to anyone else.

    The pathans of Kurram agency are some what mixed with the (indic) populations, as even some other pashtun members have said here, maybe this is the reason for the difference in the south indian component between these pashtuns and khurram agency pathans

    I think in Pakistan, they need to study Waziri pashtuns, as well as the pashtuns living close to the Dardic/Chitral areas. it would be interesting to see their results

  10. @Pakistani - What I do find striking is that in Northern Afghanistan, ethnic groups don't seem to differ consistently, or even substantially. There seems to be quite a lot of overlap. At least based on these samples. If Tadjik1_44Af and Pashtun2_6Af are correctly labelled, that is quite an interesting find. Not to mention Hazara6_66Af, who also resembles Pashtuns, like Paul pointed out. I think there might be some intense gene-flow across ethnic lines in the North, but anthropologists have already documented that, so I guess this verifies what many scholars have written. Also, it seems that the Central Asian peoples of Afghanistan are more South Asian than their relatives further north in Central Asia.

    But I wouldn't be too hasty in saying that Pashtuns from Afghanistan will be consistently different from the Kurram Agency Pashtuns sampled by HGDP. I'm sharing with quite a few Afghan Pashtuns on 23andMe. I've had the opportunity to see many of these individual's results for HarappaWorld. One Afghan Pashtun is 22%, one is 20% South Indian, another is 19% South Indian, and another is 17% South Indian. In addition, my Ghilzai friend is 12.5% Hazara, and he is still 15% South Indian. His purely Pashtun side would surely have a higher percentage. Conversely, a Pashtun from Charsadda in Pakistan is 17% South Indian, and even though I don't have my father's data, I already know he is also 16%-18% South Indian, based on my mother's results. My father is completely Pakistani. In HAP, we have HRP0281, who is obviously an Afghan Pashtun, and 20% South Indian. We have HRP0373, who is basically 70% Afghan Pashtun+30% Afghan Tajik, and they're 22% South Indian. So we really do need sampling from different parts of the Pashtun region. Not to mention serious sample sizes. The authors deserve a lot of praise for getting the data, and putting it out public, but it is quite unfortunate that they managed to get only 5 samples. At least 20-25 samples are essential for this sort of thing, especially in regions where people do mix and intermarry across ethnic lines, and for ethnic groups with large populations. Pashtuns fit that bill. But, this paper was about haplogroup information, more than anything else.

    I think the ChromoPainter/FineStructure run will shed a lot of light on Tadjik1_44Af and Pashtun2_6Af.

    • ^ true, pashtuns from afghanistan and pakistan can range from 10-25% south indian on average. Pakistani ones logically will be a bit more, however the Khurram agency samples included 2-3 samples which had over 35% south indian. To me these people are probably not of pashtun back ground, or are extreme outliners. Pashtuns of course are very large group of people and not a homogenous one. Many of the pashtun tribes probably have different origins, some could be purely West Eurasian/West Asian, some Dardic, some even NW indic. I guess many more pashtun samples need to be studied for a proper conclusion on them as a group

  11. Also btw one interesting group of people to study are the Hindko's from Peshawar and Abbotabad. They are probably the most NW indic group in south asia and were the original inhabitants of Peshawar valley. I haven't seen any genetic samples on them

    • ancient_pakistani

      You mean speakers of Indo-Aryan languages? Dardic and Hindko are indeed Indic groups.

      • Hindko is indic, quiet similar to punjabi, though dardic languages aren't classified as indic. They are there own group in south/central asia region

        • ancient_pakistani

          Dardic languages are also sub group of Indic (Indo-Aryan) languages. So Dardic people are also Indic. Hindko is dialect of Punjabi so obviously they are Indic.

          • being a sub group doesn;t mean they are indic. Infact most scholars put dardic languages as a bridging language group between indo aryan and indo iranian languages.

            If you have have heard Chitrali Khowar or even Kosher Kashmiri, they sound nothing like north indian languages. They are a distant group

            And yeah, nice name ancient pakistan, I know who you are

          • ancient_pakistani

            Dardic languages are also sub group of Indic (Indo-Aryan) languages. Basically what i mean is hindu origin of these people who speak it. And i don't know who are you? Can you remind me?

          • There is no hindu origin of the dardic people, they had their own religions that they practiced for thousands of years, just like the kalash do today, who have nothing to do with Hinduism. You are the same hindu nationalist who keeps spreading false information about your greater India.

          • ancient_pakistani

            Genetics and linguistic clearly proves they are Indic people. And im not Indian or believe in Hindutvas greater India BS. But i'm also not ashamed of South Indian component in me unlike you.

          • having the south indian component doesn't say they are indian, even afghans have 10-15% of the south indian component, Iranians have it at 5-10%, does that make them indian also? no it doesn't

            again there is no argument on this, dardic people are a distant group of people, both culturally/language and phenyotype wise.

            also stop ruining comment section with your nationalism. I am not going to argue about this further with you

        • @Pakistani, what is your kit #. Try to get your south Indian component out of you if you can. Thankless creature, you owe your very existence to India, that country, I believe, supported your ancestors for centuries. Stop insulting your that south Indian ancestor who gave you that part of your dna, and if it was not for him/her, your family had disappeared from this earth long ago.

          • stop taking things personally, I need to clarify again that I am not pashtun or dardic, I am 100% desi and I dont personally care if I have 100% south indian component. I am only debating with people who make false claims about certain groups because of their own agends

            anyways since this is not the topic here, lets stop the debate

          • ancient_pakistani

            @Paul Gill, unless he is muhajir who make 7% of Pakistan population otherwise i don't know what his ancestors had to do with India. "South India" is component name which peak in people living in South India. Its like saying punjabis mixing with Balochis.

            @ Pakistani, what i meant to say is Dardic people have Vedic origins. I didn't say they come from India or are Indians.

          • Don't you realize that south Indian component have to have come from South Indians, it certainly did not come from Norway? If a person just removes one of his direct ancestor from his lineage, his whole lineage disappears, do you understand what I am trying to say?

          • paul gill, south indian component didn't come from present day south indians, it came from a ancestral population that doesn't exist any more. That ancestral population was related to present day Andaman Islanders. South Indians today themselves are a mix of that ancient component and west asians

          • If south Indian component is Andaman Islands component then why is it called south Indian component. Again it did not get into you directly from Andaman Islanders but by mixing up with south Indians. Yet again, it is part of you and your ancestry, and you can't just ignore that fact.

            Personally,I am mostly interested in my Y lineage and not in the autosomal, which can be changed drastically just in one generation.

          • I didn't say south indian component is Andaman, but rather the ASI population, which makes bulk of the south indian component, was a cousin/related population to the Andaman Islanders. ASI does not exist anymore. The reason of course it is named the (south indian component) is because it is found in south indian tribal populations the most, just like the baloch component is found in Baloch populations the most, but that doesn't mean the baloch component originated in balochistan

          • Whatever, how did that south Indian tribal component got into you when you were not in India for more than a last 2k or so years? That component is around 50ky old, I believe, and you are not an Andamanian. One or more of your ancestors have to have that component before you can have it, am I not right?

          • ASI population existed all over south asia before, infact even outside south asia, in afghanistan and even Iran probably. Andaman Islanders of today are a cousin and related population to the original ASI folks, Andaman Islands have been isolated for very long and that's why the locals there have hardly mixed with outsiders.

            Again the reason south indians have high of the ASI component is simple, because south india overall has witnessed less migrations/invasions compared to northern south asia

          • All that I am trying to say is that all that we have has come from our ancestors and one should have respect for all his ancestors.

          • yes, but Jats should do the same and respect chmar and low caste communities in punjab, they are not different from you and share the same ancestry as you

          • @Pakistani, all should respect each other. A Jatt in Punjab will stand up even against another Jatt if he sees a Chamar being treated unfairly, at least that is my own experience, it is more like Pashtunwali attitude, but a Jatt may do that even when he is not even asked to intervene.

          • not from what I have seen on youtube and read on internet, in fact the caste system is mostly abandoned by brahmins now in Punjab, it is followed by Jatts, they treat Chamars/low castes very savagely, but again, I never understand why? not sure why Jats think they do not belong to India, when they look 100% indian and follow indian culture

            Anyways that's a seperate discussion altogether

  12. wow interesting study Zack, Tadjik1_44Af's fst to kashmiris and gujarati-muslims is smaller than to turkmen_yunusbayev!

    Pashtun2_6AF's Oracle shows just how insular some Afghan tribes are from peninsular South Asia, but I think this really is a Pashtun sample because it has higher affinities to jatts than the Tadjik1_44Af, even though haryana-jatt fst is "23.737" which is a huge distance.

    Tadjik1_44Af is one Central Asian outlier par excellence even if mislabeled as Pashtun! Could a plausible hypothesis be that his Oracle results are picking up on a newer phenomenon of Afghani Tajiks intermarrying across ethnic lines with southern-shifted Pashtuns? Then again what speaks against that sample being Pashtun is that you'd think haryana-jatts should be higher up on single population sharing than they are.

  13. The European element in Punjabis is Indo-Iranian whereas the European elements in other South Asians is, Indo-Aryan. Please discuss.

    • where are you getting this from? also the European element in other south asians is hardly visible, only some north indians brahmins have it around the 10% range, which is probably because many of the brahmins actually migrated from NW south asia to other parts of south asia. Though even the north Brahmins have the south indian component at 37-45%, clearly they have mixed much more in to the local populations of the areas. Brahmins are a very small minority group overall

      • actually the UP muslims also seem to have high Caucasian and European, though that is probably because of recent muslim migrations to the area

      • Yes they did come from Northwest India but left that region for the more fertile Gangetic Plains. This is why you will find Pakistanis have lower North Euro than Brahmins in those Gangetic regions. Even when I look at a typical Kashmiri I can instantly tell they have less North Euro than Brahmins and even some of the middle castes. There is something about the face that tells me they have significant Afghan/Iranian connections in Kashmiris. Pakistanis tend to have more Caucasian, a sign of admixture with Afghans/Iranians. People in the Punjab region tend to have higher Balochi/lower SI but also lower north Euro. This is a reflection of Indo-Aryans moving on from the Northwest region and being replaced by Afghans and Iranians.

        • UP Muslims tend to have lower North Euro than Brahmins - around 5 to 8% compared to Brahmins who have it at around 10 to 13%. In any case, the sample sizes for most groups are way too small at the moment to make any reliable judgement.

          • I meant they came THROUGH the Northwest area not from THAT area. The Indo-Aryans would have obviously had origins outside South Asia.

          • It might also be worth noting that despite Pakistanis having higher Caucasian and Balochi, they still have less North Euro than higher castes.

          • pakistanis dont have lower euro, UP brahmins have 11%, which is about the same as Many pakistanis. Most people stretching from afghanistan to UP have this component at 5-12% at average, which suggests to me this is a really old component. Also Brahmins from what I have seen are no different in their looks from other indians, some maybe fairer, but overall they are not different. Few percentage higher European wont change their look, you have to count the high 40-45% south indian in them, that is their majority component

          • Actually from the few Pakistanis I have seen represented have around 7 to 8% whereas Brahmins from UP, Bengal have up to 13%, with most ranging from 10 to 13%.

            Interestingly, Pakistanis have lower SI and higher Caucasian, similar to Iranians, but Iranians, like Pakistanis, have lower North euro. Even the Parsis have 5% North Euro, similar to Iranians, and we know Parsis are an Iranian population, although some have surely mixed with Indians.

          • I would go as far as to say Pakistanis are increasingly becoming and Indo-Iranian population rather than South Asian. This is because there is a large Pashtun population in Pakistan who have been mixing with Pakistanis and will surely mix further in the future.

          • Most of the time I can tell the difference between Pakistani Punjabi and Indian Punjabies. Some Pakistani Punjabies also give some Turkic vibe, not surprising given Muslim influences for some 700-800 years. Not always the case but quite often I can even tell Muslims and Hindus of South Asia apart in their various respective regions. Muslim invasions have had much more genetic influence in the Muslim population than these test might show, particularly in the regions where Muslim ruled for very long time i.e. in Northern parts of Northern South Asia.

        • Mal-McNija, what nonsense is this comparison? Why are you comparing one community of UP to the whole population of Punjab? And one more thing that you are ignorant of is that Iranians don't have that much European element in them at all.

        • Pashtuns themselves have 11-13% northern euro. same is true for other northern pakistanis, so you theory doesn't make sense

    • Mal-McNija, what nonsense is this comparison? Why are you comparing one community of UP to the whole population of Punjab?

    • And one more thing that you are ignorant of is that Iranians don't have that much European element in them at all.

  14. Good it seems to be realistic and in strong accord with facts on the ground.

    The only race in Kabul that sticks out notably is the Hazaras, but even they can look just like Pashtuns or Tajiks sometimes.

    It's good for the unity of Afghanistan to know that must of us are rather alike fundamentally.

    • agreed, I am actually surprised how "a like" afghans are from this study. Of course the sampling is very small, that has to be kept in mind. We need at least 25-50 samples range from every group to make proper conclusions

  15. @ HRP0286 - The cool thing about these Hazara samples is that they seem to be more South Asian-shifted than the HGDP Hazara, who were sampled from Pakistan (I'm guessing Quetta?). Afghanistan's north does seem to be quite a melting pot, and the genetic differences between various ethnic groups are rather mild, with a lot of overlap. But do the Hazara of Kabul mostly originate from the Hazarajat? If so, the HGDP Pakistani Hazara samples are more representative of the Hazara one would find in Kabul. The Hazara group sampled by Di Cristofaro et al. has to have experienced admixture from other ethnic groups. For example, Hazara6_66 has more South Indian than you, and about the same amount of Baloch! So, I think Hazara6_66 might have some Pashtun ancestry, at least somewhere along the line.

    If I'm not mistaken, your family also lived in the north, but ultimately originates from Kandahar. You must have some personal insight into this topic, so I was really hoping you could answer a question of mine. The HarappaWorld results of these Afghan samples seem to indicate a very complex and interesting ethnic situation in northern Afghanistan. All of the Central Asians of northern Afghanistan are more genetically South Asian than their relatives in the formerly Soviet "Stans" (Turkmens are an exception, they seem to match people in Turkmenistan very well, although not necessarily the Yunusbayev Turkmens, who I think are Iranian Turkmens), and the Pashtuns have results that seem incongruous with their ultimate geographic origins. What I mean by the latter is that Paktia is intimately associated with the Kurram Valley in Pakistan, and belongs to the same geographic and cultural zone as the Kurram Valley. I've been reading up on the ethnography of the region, and most scholars claim that inter-ethnic marriages occur very frequently, but happened most strongly after the Ghilzai Pashtuns experienced a shortage of Pashtuna for marriages. As someone from the region, are the western scholars accurate here, or have they exaggerated this aspect? After looking at the results again, I'm guessing what they wrote is on the mark? Random side note, but I also just saw a documentary on Kunduz province, and the Pashtuns of the area do look different from Ghilzai in Paktia, which might be of some significance, since that's their ultimate region of origin.

    It really saddens me that they couldn't acquire more autosomal samples. What makes this even more depressing, they managed to genotype hundreds of individuals for Y-DNA/mtDNA. We're talking about some very serious sample sizes for uniparental data. But the sample sizes for autosomal data are so low, we can't even decide if anyone is an outlier, mislabeled, or maybe just an example of intra-ethnic variation.

    • I am actually interested to find out when and where did the south indian/ASI component arrive in Afghanistan, or did the ASI population exist there from the beginning? because technically speaking more then 10000 years ago, much of south asia was probably ASI, before west Eurasians started arriving

  16. My family mixed with Khosti and Afridi women in the 1800s. But only married other Mohammadzai women in the 1900s. There was no problem to bring them to Northern Afghanistan from Eastern or Southern Afghanistan.

    I am just saying that my observations and this seem to be matched. Basically there is an Afghan type that all of the groups seem to share and is like an average.

    I am sure that some are like Pakistani Pashtuns, but I believe that most of us are Durrani so we should be similar to HRP0370 and myselves result.

  17. I think it has always been around. Its basically a migration from African to the coast of Arabia and South Asia. Iran and Pakistan still have African migration much more recently in places so it has continued up to present.

  18. For Kandahari Pashtuns, I think the results are consistent, with exception to HRP0373. Besides HRP0373, Kandahari Pashtuns have very similar results. I have a Popalzai friend at 23andMe, and they're 17% South Indian and 21% Caucasian, which is very close to HRP0370, and to yourself. But the Pashtuns sampled in the Di Cristofaro et al. paper aren't from the Kandahar region. Most of these people are from Loya Paktia. Balkh has a mostly Durrani Pashtun minority though. Sari Pul is mixed. Paktiawal Ghilzai shouldn't be too different from the HGDP Pashtuns, if we take into consideration geography and genealogy. And according to what I was reading (Louis Dupree dedicates a lot of time to this topic), there seems to have been a large influx of Tajik and Uzbek women into the Ghilzai Pashtun communities of the north. If Zack does a ChromoPainter/FineStructure run, and if these samples share a substantial amount with the Afghan Tajiks and Afghan Uzbeks, that would indicate some admixture.

    There is some heterogeneity in Pakistan too. I have a Pashtun friend who originates from Charsadda, Pakistan, on 23andMe, and they're 17% South Indian, 22% Caucasian. My father is around 17% South Indian, and 20% Caucasian, and he is from Peshawar. Not to mention the HGDP sample, HGDP00214, who is 12% South Indian, and 21% Caucasian. So I guess there will be some heterogeneity on both sides of the Durand Line. If only HRP0281 could shed some light on they're provincial and tribal origins in Afghanistan, it would help us a lot. Although, we do have a Ghilzai participant, HRP0326. He is 15% South Indian. But he is also 12.5% Hazara, and 12.5% Durrani. Without the Hazara and Durrani ancestry, his South Indian component would be much higher. It is interesting to note that his results are quite similar to these Pashtuns, despite the fact that he is 25% non-Ghilzai. I think Tadjik1_44Af might be an example of a Pashtun without any Tajik or Uzbek admixture, assuming they are mislabeled (which may not be the case).

    Pretty cool, so your family has some distant Khosti and Afridi ancestry in the mix. Have any of your relatives married Uzbeki people?

  19. Yes I have a few half-Uzbek cousins. I dont think mixture is as common as you indicate because relations between Pashtuns and non-Pashtuns has been negative for such a long time. Especially between Ghilzai and non-Pashtuns. I dont mean just bad, but close to violent hatred.

    On the otherhand, there are Hazaras and Aimaqs who are becoming Pashtunized. These people might marry into Pashtun families around Ghor Province and Uruzgan Province.

  20. I'm sure that's the case today, wouldn't doubt that for a minute. For the last 30 years, Afghanistan hasn't seen much ethnic tranquility. But I think the mixture happened long ago, when the lineages first arrived in the north. Here is what one scholar described:

    "When two ethnic groups live in a gray zone and begin to intermarry, the kinship terms reflect the process, and the terms of the dominant group are adopted by the the "lesser" group. Almost always, men of the dominant group will take wives from the lesser, and almost never will a man from the lesser group marry a woman from the dominant...As an example, during Amir Abdur Rahman's forced migration of many sublineages of the Ghilzai Pushtun in the late nineteenth century, the following process developed in several villages with which I am familiar. When the Ghilzai first arrived, they constructed separate villages near the Uzbak or Tajik villages. If mates of close kin could not be found in the north, young men would return to their lineage homes in south-eastern Afghanistan to seek out brides. Very quickly this system broke down. The Ghilzai who remained in the south did not want to see their daughters move away from the localized, extended family unit, for the very foundation of cousin-swapping is reciprocity. So the northern-based Ghilzai males had to look elsewhere for brides, and began to take daughters from the neighboring Uzbak and Tajik, and relations between the involved villages intensified...Post-World War II voluntary migrant Pushtun (primarily Shinwari) to the Kunduz area were at first horrified at the Pushtun-Uzbak or Pushtun-Tajik miscegenation, but some have already begun to follow the same marital patterns as the earlier migrants. These processes, among others have tended to make the nuclear or extended family the most important economic and political institution." (Dupree, "Afghanistan", Pgs 191-192).

    Basically, he's describing a situation with a lot of ethnic mixture, primarily into Pashtuns. Louis Dupree was the greatest Western authority on the history and anthropology of Afghanistan, which is why I think his account is solid. But other anthropologists have described the same phenomenon, pre-Soviet Invasion. There is a lot of old research out there on the Ghilzai of northern Afghanistan.

    Based on what you've described, I'm assuming things don't work this way anymore?

    Rather interesting information about the Hazara. Did something similar to this happen in Maidan Wardak? I know a few Wardago Pashtuns, and they are probably the most East Eurasian-looking of all Pashtuns. Wardak formerly had a huge Hazara population, so I'm guessing that some we're absorbed and integrated?

    In Pakistan, some Pashtun tribes have surely integrated Indo-Aryan people within their ranks. The Yusufzai of Swat, without a doubt, have experienced a substantial Indo-Aryan contribution to their ancestry.

    • I am familiar with him and I have quoted him here before.

      Based on his quote I wouldn't be surprised if the average ghilzai is like the average Pakistani Pashtun.

      • I think that makes the a lot of sense. If HRP0281 is a Ghilzai, that would be rather strong proof for this. Statistically, the most likely origin for HRP0281 is a Ghilzai one. It's quite unfortunate that we may never know their tribal+provincial background.

  21. I'm not sure about the particular situation in Maidan Wardak.

    I wouldn't be very surprised if a lot of the Pashtuns in the areas of Wardak and Laghman were Tajiks or Hazaras who became Pashtuns. It's a lot more common than people realize.

    • A very good point. I think this applies to parts of Pakistan as well, with Indo-Aryan and Dardic populations. For example, the Afridi Pashtuns of Tirah are often blonde or red-haired, and have quite a substantial proportion of lightly colored eyes. None of these traits are common among other Afridi Pashtuns. But Tirah was once home to a large Dardic population. It would thus make some sense that the Afridi of Tirah absorbed and integrated those people. A different example would be the Afridi of Darra Adam Khel. These Afridi are much more deeply pigmented than other Afridi Pashtuns, and some even look like Punjabi people. So it's no surprise that their area was formerly home to Hindko-Punjabi people. I guess many Pashtun people probably have genetic connections with non-Pashtun ethnic groups who live close to them, but in many cases, these things happened so long ago, that nobody remembers these events or processes.

    • I agree, tajiks from laghman have integrated within the society, like my grandfather for instance married a pashtun from kandahar, and was fluent in Pashto, yet a Tajik. Btw do you know a pashtun tribe called abawi? My maternal grandmother was from that tribe. The results show that afghans are alike more than they think.

  22. HRP282, what is the exact history of pashtuns according to you? a lot of non pashtuns are still confused about their origins? Where did they come from? how did they become in so many tribes?

    also yes you are correct, in pakistan it seems some pashtuns have clearly mixed with other local population around, in the extreme north they have mixed with Dardics, also Dardic population actually dominated the whole of northern pakistan, even Swat region. Recently I read that they have found 3000 year old Dardic graves in the Swat region.

    • Has western Afghanistan ever been tested? I am assuming people living there would be similar to Iranians

      • @Pakistani - I'm not aware of any Farsiwan or Pashtun samples from western Afghanistan. But we do have an idea of what to expect.

        The Pashtuns should be very similar to HRP0286, and HRP0370.

        HRP0034 is an eastern Iranian. The Farsiwan of Herat are very similar to eastern Iranians right across the border, in terms of both culture and language. Herat is really just a piece of Iran, that ended up in Afghanistan, for various historical reasons. HRP0034 is probably a good example of what a western Farsiwan will look like, using HarappaWorld. Not too different from western Iranians, although a bit more South Asian. 8% South Indian, and 9% NE Euro. Compared to an average of 4% South Indian, and 4 NE Euro, for the Behar Iranians (who are pretty similar to Iranians from the west of the country). I'd consider a higher NE Euro score an example of greater South Asian affinity, at least in the case of Iranians.

    • Are you saying that they did not cremate but buried their dead then, or was it a result of earthquake etc.?

  23. ancient_pakistani

    Actually Pakistani is right, they are believed to be Dardic people.

  24. Pakistani Dardic people were never hindus, why would they cremate their dead? Only Kashmiri Dardic followed Vedic religion. Also even the early vedic religion which was founded in the pakistan/Kashmir regions was completely different from the hinduism today, which is a Ganga based religion. You only have to read Rig Vedas to find how different the early vedics were from the hindus of today.

    • ancient_pakistani

      I never said Vedic and Hinduism is same and Dardic people "Indians". The only people in Pakistan with Indian origins are Muhajirs who make 7% of population. I was talking in terms of culture. Current day Hinduism is also product of Vedic people when they moved to Ganga valley. So culturally Dardic people are Indic, and genetically they also have 25-30% S-Indian component.

      Pashtuns and Baloch are not Indic despite having considerable S-Indian component especially in case of Pakistani pashtuns. Who have more S-Indian component then NE Euro or Caucasian.

      • Kashmiries have 25-30% of the south indian component, while Kalash have 22%,

        Anyways south indian component has nothing to do with Hinduism or Vedic religion or indic culture, that component has been around before religions even existed, thats why you find that component even in Afghanistan had Iran.

        Also what do Dardic people people share culturally with the indic people? you seem to be really confused, go to Chitral or even just look at the Kalash culture, there is nothing they share with the indic cultures to the south. Even proper Kashmiri culture quiet different

        Also you are wrong that Muhajirs are the only indic people, Punjabis and Sindhis are purely indic people culturally and phenyotype wise. 70% of Pakistan is Indic, which are Punjabis, SIndhis and Muhajirs. Pashtuns, Baloch, Dardic, Bursho people are not classified as Dardics

        I am pretty sure you are not Pakistani, because anyone who knows northern pakistani cultures, knows how much they differ from the pakistani cultures to the south

        • ^ meant, Pashtuns, Baloch, Dardic, Bursho people are not classified as Indics

        • ancient_pakistani

          Punjabis, Sindhis etc didn't come from India. Only Muhajirs did, and we all share Indic culture. Vedic is indic culture, no one consider Dardic central asians or iranis.

          Kashmiris and other dardic have vedic origins.

          • what is the difference between eastern pakistanis and north indians? is there any diference between punjabis on both sides? they are almost the same people, They are Indics!

            Also there is a difference between Vedic and Indic, you seem to confusing the two words. Dardic culture maybe similar to ancient Vedic culture, I am not denying that, but no way that culture is similar to the Indic culture from the plains of North India. There is a huge difference.

            Obviously you haven't traveled anywhere in northern pakistan to know this, if you are pakistani that is

          • People of Dardic origin dont consider themselves indic or indians, I know this for a fact because I had few students in my class from there

            Also even Kashmiries dont consider themselves Indian

          • ancient_pakistani

            There is no difference between Indian punjabis and Pakistani punjabis. Just like there is no difference between Indian Kashmiris and Pakistani Kashmiris. "Indic" basically mean people who speak indo-aryan languages. Which include Punjabis and as far away Sri Lankans. Also "Indic" is just a name derived from river indus. Vedic is indic culture and not the culture of anyone else, especially those outside South Asia. And Kashmiris are part of it.

          • Vedic in not Indic culture, Modern hinduism and indic culture is a product of the Ganga Civilization, not of the ancient Vedic people who lived in Northern Pakistan/Kashmir region. Kashmiries have never considered themselves Indian and other Dardic people like Kalash and others are completely different, their culture is more similar to the Pamiri and Nuristani peoples, not of the people from the plains of India. Anyone with a bit of knowledge of regions knows this

          • ancient_pakistani

            I think you need to read more about vedic people. Their homeland was punjab for thousand years, not Kashmir. Yes ganga valley and vedic is not 100% same but ganga valley hindusism didnt come from thin air. Sanscrit which is vedic language is indo-aryan or indic as you call it. So tell me how can vedic not be indic culture?

            The punjab is called birth place of hindusism. all brahmins trace their ancestry to punjab. They moved from punjab to rest of South Asia.

          • My argument is not about where vedic culture was formed, be it punjab/kashmir/khyber Pakthukhwa, I am saying Ancient Vedic culture and Modern Indic culture are not the same thing. Ancient Vedic people indeed probably were close to cultures of Dardic people of today, because reading the rig vedas you do find a lot of similarities to them. Also I am not saying that hinduism came out of thin here, yes it came out of Vedic religion, however it clearly got mixed with the local pagan Dravidian religions of India. Modern north Indic culture of today is a product of ancient pagan Dravidian cultures and the Vedic culture that came from the north, in fact I would argue that The Indic culture overall is till pretty much Dravidian, only languages are indo aryan. This is the reason you find such a huge difference between lets say Kalash people of Pakistan and the peoples of the indic plains, from Punjab to Ganga. The base of Modern Hinduism and Indic culture is the Ganga

          • ancient_pakistani

            Modern day hindusim as you call it, came to Punjab and Sindh in Gupta Empire time, about 1600 years ago. And there is no debate on homeland of vedic people, its punjab known as sapta sindhu back then. This include Pakistani and Indian punjab.

            Sansrit is original indic language. There is also dravidian people who are not Indic.

          • There are plenty of places in Peshawar valley and KP areas (Swat) that are mentioned in the Rig Vedas. Even Panini the earliest Sanskrit scholar was from Peshawar, so the Majority of the early Vedics lived in the area stretching from Peshawar to Punjab, that was the main home land.

            Anyways again, that is not my argument. Infact you are confirming my argument when you say it was the Gupta empire from the east that spread Modern form of HInduism in the Punjab area. The early Vedic religion which was based in KP/Punjab areas was not the same as today's Ganga's based hindu religion and thus their culture was very different from modern indic culture from the plains

          • ancient_pakistani

            You do realize these people were indic and not pashtuns like today in those areas? Peshawar was indic majority city till recently in last century.

            Swat, peshwar are historically indic cities. And no they were not dardic people. Also do you know who were Hindu Shahis?

          • Swat was Dardic as proven by the graves, maybe indics came there later, not sure. Also again Indic culture spread to these places later on, with Gupta or even Ashoka empire. You can't deny that modern Indic culture is a product of the Ganga region, it has very little to do with the early Vedic civilization.

            Anyways enough with this argument about Vedics/Indics

            Your original assessment calling Dardics as Indics was completely wrong, now even you are differentiating the two

          • ancient_pakistani

            Not really i was not wrong but there is difference between punjabis and dardic. And graves didnt prove anything lol please. Also do you know who were hindu shahis who ruled Afghanistan? Dont tell me they were pashtuns. Also you are confusing linguistic group indic with culture and ethinicty, basically you are ignorant.

          • Herut River also been their main stay and is mentioned numerous times. And there is one thing that there is no such thing as Hinduism, that word comes from Persians. The name is Sanatan Dharma and the very name tells you that it is not a religion, it is pragmatic and is not based on any dogma like the religions of middle east are. Again the path praticed must suit the mind set of the practitioner, all methods found in all the religion of the world are found in there, plus many more, any method or path which is most suitable to a particular person is the best path for that person, all paths are accepted and non is rejected. Again it is MONOISM, not monotheism.

          • You were wrong, admit it, now you are even saying Punjabis are not same as Dardic, before you were saying all people are Indics. lol

            Also read the article again you posted yourself, it says Dardic people dominated the Swat region and much of Northern Pakistan region. HInduism/Buddhism came later to the region

            Also Hindu Shai were a group from Gandhara, who ruled some parts of Afghanistan, so what?? they were probably like HIndko people you find in Khyber Pakhtunkhwa, who are a mixed group culturally, their languge is indic, but their culture is mixed. They aren't exactly punjabi and thier defiantly not related to Indians from the Ganga region

          • ancient_pakistani

            @pakistani, ofcourse there is difference between dardic and punjabi just like there is differences between UP/Bihar and punjabis or Sri Lankans and Punjabis despite speaking indo-aryan language.

            And we hindko are punjabis, we speak punjabi dialect and majority of us have castes likes jats, rajputs etc.

            Hindu shahis descendents are among punjabis, especially Kamboj tribe. You are probably pashtun but i doubt Pakistani pashtun. Otherwise you wouldn't have been so ignorant of our history.

          • A hindko person will never call himself punjabi lol. This has given away your fake identity, you can't be a pakistani at all. They have a seperate identity and are proud of that identity

            Also, no. Punjabis are much similar to people of north India / UP, then lets say they are too Dardic people. Dardic people are similar to Nuristanis/Pamiries. They have nothing to do people of Indic plains

            Anyways, I am done arguing with you, because you have exposed yourself with complete lack of knowledge of Pakistan and it;s ethnic groups

          • ancient_pakistani

            Fake identity? Im not the one ashamed of my identity unlike you. Hindko speak punjabi dialect. Indian punjabi and hindko can speak with each other in punjabi thats how similar dialects are. Also how can Punjabi Indian jatt let alone Pakistani punjabis be similar to UP and Biharis who have 50+ ASI and very little caucasian and NE Euro?

            Are you not the one who say Kashmiris are not Indic despite having 30% ASI? And hindko people will have 30% average ASI just like rest of Punjabis.

          • I happen to know a lot of Hindko people personally and they will never call themselves punjabi, there are blogs written by hindko people on the net, arguing for their separate identity and not being called punjabi, again stop pretending to be someone you are not.

            Also UP people are not 50% ASI. They have a range depending on the ethnic groups. some muslim UP groups are only 30-33% ASI, Brahmins are 37%, while many of the other groups range from 40-43% ASI. The low caste groups might be over 50%, however remember the Punjabi low caste groups, which make a huge chunk of the population in Punjab haven't been tested, they are probably in the range of 40-50% ASI.

            Overall most punjabi groups even here on Harrapa are coming out to be 30-35% ASI/South Indian, which is not too far from UP averages at all. Also looking at the (phenotype, culture, language) most punjabis are very similar to people of UP. This is reason even in Pakistan, Muhajirs and Punjabis are so often confused with each other. A pashtun or a Dardic person will almost never be confused for Punjabi or Muhajir

          • ancient_pakistani

            We Hindko people speak punjabi dialect, Serakis speak punjabi dialect, and potoharis speak punjabi dialect etc. And dalits make just 3-4% of Punjab population, even less i guess. Christians churas are dalits converts in Pakistan. In Indian punjab 35% of are dalits. Punjabi dalits probably still have lower ASI then majority of UP.

            UP-Kshatria which is higher cast is 45% ASI. UP muslims 53%. The ones with 34% are pathans who mixed with locals. UP Sch caste ASI is 58%. Muhajirs and punjabis confused with each other? Hell Fazlu rehman pathan from FATA can easly be confused with Chura dalit in punjab. Have you even been to Peshawar? Or at least have you watched pashtun movies?

            Anyway im done with you, this conversion's is stupid anyway. You basically mean people from punjab to Bangladesh are the same while pashtuns are completely different from inferior indics.

          • I am muhajir myself and I know majority of Muhajirs and punjabis look very similar. It's only some Punjabi groups that may have a different look, mainly around western or northern borders of Punjab. Also Peshawar is not a Pashtun dominant city, it is Hindko city and a lot of pashtun movies, the actors are either punjabi or hindko. Tribal pashtuns are the real Pashtuns and they dont look Punjabi or desi at all

            Also again you are wrong on the low caste population in pakistani punjab. Have you ever been to interior areas of Punjab? Most people there could be confused with Tamils to be honest and the interior people are the majority in Punjab. Go visit the thread on Biodiversity forum, which compares Pakistani punjabi cricketers with Sri Lankan cricketers, most people on the forum said, they looks similar to Sri Lankans, only the Pashtun players in the team looked Iranid. Most Pakistani Punjabis dont have the foreign look at all, like Pashtuns and Dardic people do. Punjabis are 100% Desi and Indian as a group

          • honestly speaking, I find (Sikh) men to be more west asian looking then Pakistani Punjabis as a whole, maybe they are Jats that migrated from the Pakistani Punjab to India in 47, not sure. But Pakistani Overall are very similar to their North Indian Neighbors

    • I guess there is very little chance of finding ancient remains during the Vedic period for autosomal and haplogroup testing.

      • I wonder how the Farmana testing is working out? I hope they can report something of interest, this year.

        • Good point man, they seem to be taking a long time. Nonetheless I think the onus is to get it right the first time. Either that or the results are controversial and they are waiting for the right time to publish it (me speculating)

          • Sorry but what is the farmana testing?

          • Yeah probably it is because they are controversial. Indus civilization was either a west asian like civilization or a south indian Dravidian like civilization. Both of these are controversial issues. Btw why test Farmana and not the skeletons found on the Indus river? Was Farmana even part of the early indus civilization based on the indus river? There were many Skeletons found in cities like Mohanjedaro and Harrapa sites

      • There is this new discourse among some Pakistanis, that of Aitzaz Ahsan's Indus Saga. This postulates a 'Gurdaspur salient' separating Indus from Ganga culture and defining Pakistan as an entity that always existed west of this salient.
        It is a fatuous concept and easily disproved. However, there is no doubt that people of the Ganga Jumna belt have characteristics differing from those of Punjab, as do those South of the Vindhyas from those North of it, or Bengalis from Biharis and Sindhis from Punjabis. All that is rather elementary.
        The Vedic civilization did centre in Punjab for centuries but moved into the Gangetic basin in the first millenium BCE and then further south and east reaching Sri Lanka and Assam eventually. The basis of modern Hinduism is both the Vedas and the various pre-existing indigenous faiths. Projecting modern national animosities into the remote past is absurd, ethnically, religiously and politically.

        • I am not differentiating Punjab from rest of North India, I actually said Punjab region of Pakistan is pretty similar to rest of North India. My argument is about the dardic people of north Pakistan, they are very different from rest of north indic people and their culture, including punjabis or people form the Ganga region.

          • I am still unable to make out what precisely 'Indic' means in the context of this thread, or Dardic for that matter. Dardic is a linguistic term and Indic is undefined. Identities are modern things, no more than a few hundred years old at most. Racial compositions change over the centuries. How are we expecting to uncover the truth if gene mixtures are constantly mislabelled because of identity politics. Two thousand years ago there were no Punjabis, Pakhtuns, Dards or Sindhis. There was however something called ASI. If ASI provides the 'Indic' component then we have some clarity on what Indic means. If however, as has been sometimes argued (in the context of ASI component in the Pakhtuns and others) that ASI is not the same as 'Indic' then some definition of Indic is necessary, The term 'veddoid' was employed in one of the threads. Is veddoid an aspect of ASI? Does Indic mean dark brown and flat nosed. If so it is just repeating a prejudice of the Vedic people, a remarkable example of the persistence of memes along with genes.

          • @Onlooker - Completely unrelated to the topic at hand, which, by the way, is extraordinarily nauseating (the whole Indic-Dardic thing is really making the site look bad. But I have to say, you are the only sane voice in a highly eccentric and irrelevant conversation. Hats off to you, and I mean that), but the cool thing is that Iranians are around 10%-20% ASI. I guess you could say that ASI isn't really restricted to "Indian" people. The fact that Iranic speakers may range from 30% (Pashtuns) to 10% (Kurds, and the more West Eurasian of the Iranians) shows that ASI is more indicative of pan Indo-Iranian ancestry.

            Don't understand why though. I guess the people who spread Indo-Iranian languages may already have had ASI ancestry. Or perhaps there is a Paleolithic East Eurasian substrate across the Iranian plateau, and into South Asia proper. I'm guessing the latter makes more sense. Perhaps ASI-like people lived as far west as Mesopotamia.

          • HRP0282,

            What is the source of your ASI ancestry figures, especially those of the Iranic groups?

          • Hi Onur,

            A good friend of mine was working with some of the software put out by the Reich lab, and experimenting with different data-sets and tools. They found that the Behar Iranians, once purged of African-admixed samples, turn out to be around 80% West Eurasian, 20% East Eurasian. The range is around 80%-90%. The HGDP Pashtuns turn out to be around 70% West Eurasian, 30% East Eurasian. I'm assuming all of that East Eurasian ancestry is ASI for the Iranians and HGDP Pashtuns. I think this is reasonable for the Behar Iranians, since they don't show any evidence of East Asian admixture. Assuming Kurds are on the West Eurasian end of Iranian variation, I just placed them at 90% West Eurasian, but in their case, I don't have actual numbers. I guess some Kurds might be around 20% ASI as well, or perhaps lower than 10%, but I honestly don't know in their case.

            On another note, Dienekes once tried creating Admixture "zombies" for ANI and ASI. It was a rather successful experiment. His percentages correlated very nicely, and almost exactly, with the Reich et al. paper. Interestingly, he found that the Behar and Dodecad Iranians were around 7% ASI. When he applied the same method to all of the HGDP Pashtuns (22 samples, excluding the Hazara-like outlier), he obtained an ASI estimate of 18%. So the average HGDP Pashtun seems to be around 10% more ASI than Iranians, consistently. This is always a good thing to keep in mind, an approximately 10% difference. This probably accounts for why most Pashtuns don't really appear strongly South Asian in comparison to Iranians, since 10% isn't really enough for phenotypic differentiation.

            The Moorjani et al. paper pushed up ASI estimates by quite a bit, so the Iranian percentages will rise proportionally. If the Moorjani et al. paper used all of the HGDP Pashtuns (rather than just 15 samples), their estimate of ANI would be around 75%. So I think the same method applied to Iranians would yield around 85% West Eurasian. But the latter is obviously guesswork, I have no evidence for this whatsoever, and I might be wrong on this count.

            Nevertheless, the 30% ASI for the HGDP Pashtuns, and 20% ASI for Iranians, are actual numbers obtained using Admixtools, and f-statistics.

            I think the high percentages of ASI in the Iranian plateau point to the persistence of an ancient East Eurasian substrate, shared with South Asia. And I also think the Indo-Iranians (if we can treat them like an actual individuated population, which may not be historically accurate) might already have had some minor ASI admixture.

          • HRP0282,

            Are you sure Kurds, non-Turkic-speaking Iranians and especially Pashtuns do not have any Mongoloid admixture? Their Mongoloid component/South Indian component ratios are much higher than that of the average South Asian. So some portion of the East Eurasian-related ancestry of Kurds, non-Turkic-speaking Iranians and especially Pashtuns is likely to be Mongoloid ancestry rather than ASI ancestry, though their East Eurasian-related ancestry is surely mostly ASI ancestry rather than Mongoloid ancestry.

            BTW, what about Central Asia, East Eurasia, Siberia, the rest of West Asia, the Caucasus, Europe, Africa (including the north), Oceania and the Americas? Did your friend say anything about any ASI ancestry or something similar in those regions?

          • How much of ASI is east asian? I doubt Pashtuns have any more east asian ancestry which is already showing, around 4-5%

            Also the reason I think ASI is found even in Iranians is because ASI were the first groups of people living in Asia. ASI/South Indian like other components is probably a mix of several different types of people. Perhaps the ASI in the Iran/Afghanistan area is somewhat different to lets say ASI in India/South India. We have to look at everything and breakdown these components further to understand how things work, especially in relation to phenotype

          • Onur,

            You make a very good point. I never thought of it that way.

            I could ask my friend about what they found, analyzing other regions.

          • Kurds do not have any South Indian (if they do, it doesn't translate into phenotype). I have come across some Kurds in my area and none look like they have any South Indian but a few do show small signs of Mongol admixture. With Pashtuns, it depends, if they are from Pakistani side of the border then some can give some Indic vibe, to their otherwise overall Pashtun look.

          • @Mal - They aren't too different from Iranians. If the average Iranian is around 20% ASI, than I'm sure the average Kurd will be close to this, perhaps around 10% ASI, but I haven't seen numbers for them. But around 20% ASI for Iranians is an actual percentage, an actual percentage obtained by analyzing the data.

            Phenotype means nothing at this scale. I look like a Kurd, Turk, or someone from the Caucasus (especially Azerbaijani people). West Asians always assume I'm from their part of the world. Two of my siblings look like they're from the Balkans. And at no point in my life has anyone ever assumed that I'm from South Asia, and no South Asian has ever recognized at first glance that I originate from their part of the world. But I'm not West Asian, and my siblings aren't from the Balkans.

            And yes, the Durand Line is a mystical thing indeed. Once you cross it, magical stuff happens. All of a sudden, to it's west, Pashtuns become "West Asians". And all of a sudden, to it's east, Pashtuns become "South Asians". In fact, I think the British deliberately worked these things through with the Amir, so that all of the "South Asian" Pashtun tribes of Yaghistan would end up under their jurisdiction. 😉

          • Well, from the database on here, the few Iranians and Afghans all have less than 5% South Indian, nothing like the 20% you mention.The Kurds 1% and 2% SI, which seems right. From my experience, the kurds look like Turks. I know an Iraqi Kurdish family well. The daughter looks Turkish and the father looks Arab. On the whole, though, the Kurds in my area look similar to Turks from Turkey I would say.

          • Sorry, I meant to say the few Iranians and Kurds not Afghans all have less than 5% SI. In the case of Kurds barely any.

            The Pashtuns in places like Peshawar and the general NWFP have probably mixed to some extent with Punajbis (Indic peoples).

          • HRP0282,

            Did you communicate with your geneticist friend on the issues I raised?


            Would you mind if I asked you where you are from?

          • As a pakistani I can confirm that pakistani pashtuns on average are far more desi looking than afghan pashtuns overall. I am not surprised at all that afghan pashtuns genetic results are confirming what in already known by us in punjab that afghan pashutns are whiter and more foreign looking than pakistani desi pashtuns from KPK where mochis, nais and mirasis also call themselves pashtuns just because they happen to speak pashto.

            And from my experience in europe , the kurds of turkey look mostly like turks from turkey but sometimes just a little bit darker than turks. There is absolutely no overlap between kurds from eastern anatolia and persian people of iran despite linguistic similarities.

          • Supari, it depends which pashtuns you are talking about. From my experience in Pakistan, Pashtuns in Peshawar do give a more desi vibe, however there maybe a reasonable explanation for that. Peshawar was Hindko city and I am pretty sure Pashtuns living there are mixed with Hindko people. I heard many of the Pashtuns in Peshawar dont even know their tribes, this is a big clue

            Now if you look at the Pashtuns living in lets say Waziristan or Bajaur agency in the north and some of the other agencys, those people look much more similar to their Afghan brothers and dont give any desi vibe. So Duraand line theory that HRP0282 is talking about probably only applies to Peshawar valley and probably even Kurram Valley where the original Pathan samples were taken. From the pics I have seen of Kurram agency, people there do look a bit desified, not sure why.

            Overall again, it probably depends on the individual tribes, some tribes may be more south asian genetically, while others not so much. Thats why they needs to really label which tribes are being tested

        • Indic doesn't mean = ASI or the south indian component. This component is thousands of years old before any type of culture or language existed, and even Iranians have this as HRP282 has said.

          Indic is basically a cultural, historical and a language reference, in fact you can probably use the word (Desi) it means the same thing as Indic. Dardic/Pashtuns and other northerners are not desi people, while Punjabis and other Indians are. It is very simple to figure this out, not sure why are people making it complicated

          • I was trying to get some clarity from the posts. If Indic means Indian let us use the term Indian, or Desi if you prefer. If Indic or Indian is a cultural, linguistic and historical term only it can have nothing to do with genetic types, and as we know, genetic variability across the old British India is considerable. Desis also include UP Pathans and Syeds of course.
            Indian by your definition is a meme not a gene. People are searching for their genetic origins here, which is a good thing if it helps understand the interconnectedness of people, but not if the purpose is 'racial'.
            ASI, which can find relationships with only Andaman groups probably shows that the people coming out of Africa split off in the middle east somewhere before migrating south into India. In the days before monarchs when people lived in tribes these tribals found their way about from the mid east to South India and the Andamans. The conflicts on this thread about who is Indian and who is not are meaningless in a genetic context. You may say Pathans are not Indians and no one will dispute you because that is a matter of fact political statement. Some commentators here however insist that 'Indian' 'Indic' or 'Desi' are not cultural markers but distinguishable by genes, and that is the source of the trouble. It then comes down to what percentage of ASI makes you non Indian or non Desi. Pathans with upto 20% are not Desi some say, and I have no quarrel with that at all, but is that an arbitrary political conclusion or a scientific one? We are then getting into the business of phenotypes which say that there is a non desi look about Pathans. So there is, but would you stake scientific theory on that? Kashmiris also have a non desi look but their ASI is higher than that of Punjabi and Haryanvi jats. So what do we conclude?

          • Onlooker, I am talking about real life scenarios. I agree none of these talks are related to genetics, however in real life most people look at your phenotype, culture, language etc.. to judge where you are from, not genetics.

            Also regarding Kashmiries or Jats, the samples are too small to make any sort of conclusions. Jats are arguably much more Desi in everything (phenotype, culture, language) even though their ASI levels might be around the same as Kashmiries. Thats why I said ASI or South Indian components doesn't specifically mean being Indic, because this component is a very old component, spread very widely

  25. Actually found this article published just few days ago, apparently they could not test the Farmana samples because of high acid ground destroyed all dna:

    ^ This is more the reason to test the desert sindh samples from Mohanjedaro, the skeletons there are much older and ground is much drier

  26. ancient_pakistani


  27. Question: I see that there are some South Asian components like South Indian & Baloch amongst these Pashtuns. However, they still form less than 50% of the admixture runs.

    I see a lot of arguing here, but to what extent are these South Asian populations? These is obviously some admixture, but these mostly look like Central Asian populations, and therefore tell us very little about the populations, ancient or modern, of the subcontinent.

    I am probably missing something, so please enlighten me.

    • Isn't Baloch component a mix of West Asian/South Asian?

      • It is, but the west Asian component seems to be shared with peoples throughout the subcontinent.

        On the other hand, Pathans show a large number of components that don't show up in a appreciable amounts elsewhere further East or South.

        • Their Baloch components are lower, while their Northeast European, Northeast Asian, Siberian and Caucasian components are higher.

          • It is only that the component are higher or lower but most of them are found in all other people, you explanation is useless. These variations are also found within any community to a degree.

    • Also Pashtuns defiantly came from outside the region, looking at their culture/Identity. But overtime they probably mixed with other populations. Thats why they show south asian admixture

      • Quite right. So why is it that they are being used in any reference to the genetics of the subcontinent? I am quite surprised that they are being used in ASI-ANI runs, with the assumption that all non-ASI = ANI. Their west Eurasian components look pretty different to those found in South Asia.

      • Pakistani.

        What is your kit #. All Pashtuns owe their look to Indus, the savior of Pashtun look. Personally I think Pashtuns are Scandinavians, the children of the Vikings.

        • Paul Why do Jats from haryana have such a Indian phenotype? I have been going through the net searching about Jats from Haryana, I have seen many videos and pics and none of them even remotely look to any other outside population. Most are very darkskinned to be honest and can easily fit any where even in South India. At least with Sikh Jatts this is not the case, Sikhs also look Indian but with a notable West asian influence in their phenotype

          • We need to split phenotype and genetic ancestry here. Appearance does not equate to ancestry. Phenotypes are strongly influenced by selectional factors, and can change within generations. Most studies genetic ancestry focus on markers that are not influenced by these factors, since they don't necessarily 'code' for anything physically noticeable.

          • I agree with what your saying, but I still find odd some one with 18% Northern Euro ancestry wont even show any hint of (West Eurasian) like physical features in them

            I mean of course these samples are very small and perhaps the tested individuals do show those features, but then that would mean these numbers dont really represent the bulk of harayana jats, at least from the videos and pics I have seen

            I wish there was way to look at pics of the individuals being tested here. I guess that would be invasion of privacy though

          • Because that ancestry is really old and there are even people with 50% (i.e, Barack Obama) European who don't show a lot of signs of it. Depends on which genes you got. The stereotypical phenotype for Jatts does include a body morphology similar to Europeans though, which is what the British commented on in regards to Jatts/Rajputs in general (their height, stature, and beards reminded them of other Europeans).

            My mom is HRP0374 and she's pretty much white in skin tone (in spite of the Pakistani tan), could have passed for a Mediterranean or Iranian/Afghan in her youth. But her admixture places her all the way in India.

            My admixture (HRP0349) resembles Uttar Pradesh area Indians but no one has pegged me as anything other than Punjabi and Pakistani at that though some do admit my eyes seem kinda Asian-ish or like Nepalis (but these people haven't really seen too many of those, my eyes are just angled a bit and look small). I was in Sikh-heavy areas of the Toronto area a while ago and they all recognized me as Punjabi and would ask if I was from Lahore.

            So the lesson to take away from this is to not get hung up on linking admixture to phenotype. There isn't much of a link. A person with 25% or even 50% European might not look like you photoshopped a quarter of modern European in or whatever you're imagining. But there are people with very low European admixture (like Iranians) who look outright European.

          • I think aniasi has already explained it to you Genotypes don't always translate into Phenotypes. I know that you are well aware of the dominant and recessive genes. There are many SNPs involved in phenotypes just as there are many SNPs involved in the eye color. The Jatts who don't look much western have obviously picked up some of those dominant SNPs while in India and their story is no different than the Pashtuns who don't look very Pashtun like even when those pashtun may actually be very close to other Pashtuns genetically. Another factor is the environment, natural selection is in play in this case. Hariyana is hotter than Punjab and Punjab is hotter than west of Indus area and as Jatts mostly spend their life out in the sun unlike Brahmin, Khatri, Bhatias and the like so here you find the evolution in working.

          • AD, are you punjabi? what are your components percentages if you dont mind?

            Also like I mentioned before, generally there isn't a lot of difference in phenotype between Pakistani Punjabis and West UP muhajir folks. I have many friends from Punjabi and muhajir back ground, most of time they can not tell each other a part.

            Paul Gill, lets assume your theory is correct, if Haryana Jats are more (Indian) in appearance because of their Indian SNP's, why doesn't this apply to other people? including many sikh jatts as a I mentioned who do have west asian influence in their appearance, is it because Sikhs mixed with outsiders recently and Haryana ones didn't??

          • Yes, my ancestry is Punjabi Jatt going back as far as anyone remembers (up to great-great-great grandparents). My numbers are in the Harappa spreadsheet which you can view on the right sidebar, I am HRP0349 and my mom is HRP0374. Our numbers are more East-shifted than most Punjabis and most calculators, oracles, and other such things place us in North India (although Dr. McDonald placed us in Pakistan and in that Harappa map I made we're in the tail area (southeast) of Punjabi Jatts where they overlap with Brahmins, Kshatriya, Uttar Pradesh and non-Jatt Punjabis, so it's still within the range of Punjabi genotypes).

            We're definitely not of a UP Muslim background though, several branches of our family were Sikhs and I have a bunch of Sikh Jatts on my 23andMe relative finder. UP Muslims wind up being close to Punjabis in genotype and phenotype somewhat because of mixing Pathan ancestry with local North Indian groups. At least one branch of my family is local to Multan (Balham clan). Most of the rest emigrated from Hoshiarpur to the Faisalabad area (Gill and Pansota).

            Haryana Jatts like other people in Uttar Pradesh and East of Punjab have slightly lower West Asian/Baloch admixture and this might also be reflected in phenotype because of selection over the generations. Punjabi Jatts generally have higher West Asian (some as high as mid-40s), often higher Caucasian too. In fact, I would bet that aside from body morphology, the NE-Euro in South Asian populations probably has no relevance to their phenotype at all. It's too old. The modern European look evolved out of pretty recent selection, including their pigmentation genes.

          • Hi AD,

            Personally, I've found that Punjabi people are easily distinguishable from Muhajir people. Most Punjabi people "look" strongly northwest South Asian. In fact, they pretty much define what it means to "look" northwest South Asian. Most Muhajirs typify the north Indian phenotype very well, and most don't resemble northwest South Asians. To be honest though, both groups aren't really comparable. Muhajir people are extremely heterogeneous when it comes to geographic origins within India. Bihari Muslims, UP Muslims, Muslims from the Delhi area, Memons from Gujarat, Bohra Muslims from Gujarat, Muslims from Madhya Pradesh, Muslims from Kerala, Muslims from Tamil Nadu, and even Bengalis, are all considered Muhajirs IRL, in Karachi (at least in my experience). Caste origins are equally diverse, not to mention the occasional individual with distant West Asian/Pashtun admixture. Honestly, they don't really constitute an actual ethnic group, not by a long shot. I think "Muhajir" identity is politically based, and before the whole fiasco of ethnic/economic tension started, most Muhajirs didn't see themselves as a single community.

            Although, I wouldn't consider the "Baloch" component to be West Asian. Here is a spreadsheet with results Dienekes obtained for a very successful experiment The Baloch and Brahui are only 1%-2% less "ASI" than the HGDP Pashtuns! For HarappaWorld, the Baloch are 9% less "South Indian", and the Brahui are 11% less "South Indian". That is quite a large gap between "South Indian" percentages, and the actual "ASI" percentages. So, I'm guessing the "Baloch/Gedrosia" component subsumes some "ASI". That might explain why having more of the "Baloch" component drags one towards South Asia on PCA-MDS plots.

          • AD, Thanks for the info!

            HRP0282, have you ever been to Karachi? anyways you are right, it really depends where the Muhajirs are coming from. I am specifically talking about Muhajirs from Delhi and West UP areas, those are indistinguishable from Pakistanis Punjabis most of the time. You can't tell them apart from people in cities like Rawalpindi, Islamabad, Lahore etc...

            Biharies/Bengalis and south Indian Muhajirs obviously have different phenotypes.

          • Among Punjabis though, I have noticed there are few groups like Sikh Jatts who have a bit West Asian shifted Phenotype, of course they look Indian, but still with clear West Asian like Influence

            Also in Pakistan I have seen some Gujjar people, specially from northern Punjab who almost resemble Pashtuns on occasions.

            Btw, have Gujjars from Pakistan ever been tested?

          • But there are people with very low European admixture (like Iranians) who look outright European.

            Relatively so. Then again, nearly all West Asians have a relatively or sometimes absolutely average European look. That is because the native West Asian genes are genetically close to the European genes and Europe and West Asia constitute a genetic continuum (their relationships are clinal and smooth).

          • AD and HRP0282, this Pakistani is none other than the same Muhajir with agenda who been arguing here for a while now. You can give him all the explanations in the world, but they will never be enough, because he is a my way or the highway kind of a personalty.

    • ancient_pakistani

      There is a lot of diversity in looks and skin tone of pashtuns. Pakistani pashtuns are South Asians with Iranic linguistic group.

      • Pashtuns from both sides are the same, there is not pakistani or afghan pashtun. Just like Punjabis from both sides are same. Clearly your knowledge is very limited when it comes to Pakistani groups. Clearly you are not pakistani

  28. @AD

    "But there are people with very low European admixture (like Iranians) who look outright European"

    because Iranians have on average have 40% Caucasian and 11-15% SW Asian taking them west. 8-12% northeast-euro in punjabi-jatts is swamped by their higher value 28%-30% S. Indian component, so how would they look like europeans in the first place. Looks like those British were lying.

    My advice is don't take the British too seriously, they said jatts esp. Sikhs came from europe using pseudoscience anthropology because of their colonial interests. Not from conviction. Sikhs were just more open to British demands because of their own anti-muslim interests, and that's why British flattered them with myths and fables about coming from greece, russia or whatever to win them over in their armies. One thing is the British never talked about the anthropology of muslim aristocracy who have iranian and pashtun ancestors because muslims hated them so from there you can see it's all about flattery through self-interests.

    I found a video of of punjabi, sindhi and rajasthani jatts. Pakistani may not be wrong in saying they look more Indian than their genotype shows

  29. ^ Sikhs Jatts look like the video shown, of course they look Indian, but they do have a West Asian type vibe which other Indian groups lack. I am not sure how Pakistani Jatts look, I dont remember meeting one

    I was actually talking about Haryana Jatts, those guys dont show any West Asian vibe from all the videos and pics I have seen.For example Virender shewag, the indian cricket player is a Haryana Jatt, he can basically fit in central or even South India.

    This is also a pic of Haryana jatts:

    ^ they differ a lot of from their Sikh Punjabi neighbors

    • @Pakistani, sikh jatts don't look west asian even for their west asian vibe. That vibe is pointing to something more stable and ancient, and also they aren't the only Indians who have that west asian vibe.

      Indian Muslims across western UP, rajasthan, gujarat, esp in bohris and surti muslims, and shias with iranian ancestors everywhere in the subcontinent even in south india and bengal have that west asian vibe, but the sikh-jatts still look different from them so their west asian is a different source from that of those Indian muslims.

    • I doubt sikh jatts have any west asian vibe, at least I do not see any such vibe in them in europe. Pakistani punjabi muslims on a whole have far higher incidence of west asian phenotypes than any sikhs that I have seen so far, this is also confirmed due to higher Baloch and Caucasian component among muslim punjabis of pakistan as compared to indian punjabis. Moreover , the jatt identity at least in most pakistani punjabi muslims (native to west punjab) was very fluid so they are not necessarily jatt in the rigid sense since alot of jatt clans in pakistan claim rajput ancestary in different areas. For a pakistani punjabi like me , Jatt, Rajput and gujjar idetities are quite mixed up in pakistani part of punjab so they should not be compared literally with corresponding identities in india where they are much more rigid. Genetic results of pakistani punjabi muslims jatts, rajput and gujjars all are showing similar high Baloch and Caucasus and lower NE euro component compared to indian sikh or hindu jatts which again shows the fluidity of these identities among muslims of pakistani punjab depending on the region.

      For example I know this Minhas guy from gujrat district of pakistan who says that they were hill rajputs just a century ago from lower tracts of mountains encircling kashmir but now they identify as jatts in gujrat district and intermarry with them. Same applies to many clans througout pakistani punjab , it is very different than indian punjabis.

  30. @HRP0282

    which part of karachi you from? Muslims from the border area of India such as Rajasthan and north Gujarat memons are indistinguishable from Sindhis. UP/Muslims blend into Punjab fine because of intermarriages with Punjabis. in karachi all those so-called "muhajirs" have Punjabi or Sindhi family members (its rare to find one who doesn't) so how could they stick out exactly? the only muhajirs who stick out are biharis and tamils, and you might know all about them because the worst muhajir conflicts were the ones instigated by Pashtuns against biharis in the 1980s.

    Pashtuns are better fit to be called muhajirs in karachi because so many of them are strangers, who arrived there recently from up north fleeing the war whereas the rajasthanis gujaratis and parsis have a presence that predates the formation of pakistan. They had never called themselves nor did anyone call them muhajirs, the term muhajirs came into vogue only after UP urdu-speakers migrated to pakistan because of language conflicts.

  31. I would like to add something regarding origin of some part of south asian components in pashtuns of afghanistan. As we all know the afghan pashtuns are predominantly from durranis and Ghilzais supertribes which have numerous subclans under them. A very interesting thing about many Ghilzai afghans is that they are "kochis" or semi-nomads. A lot of Ghilzai clans have historically been nomadic merchants also called Povindah/Lohani/Lawani tribes such as Kharoti, Nasirkhel, Niazi and even to some extent sulemankhel. And there is a considerable agreement among the researchers that Lohani/Povindah tribes are partially descended from ancient sindhi hindu Lohana/Bhaiband merchants who were known to carry trade as far as central asia and it is believed that many of the central asian nomadic Lohana"s were cutoff from their sindhi hindu Lohana bretheran when the central asia and modern afghanistan came under the rule of ghaznawid turks, these lost central asian hindu lohana/Bhaiband merchants may be ancestors of modern Ghilzai Lohani/Povindah nomadic tribes. Surprisingly the attan dance and style of clothing of many of Ghilzai kochi tribes is still very similar to the Lohana hindu merchants of sindh and indian Gujarat.

    So Mr. Imran khan of pakistan who belongs to Niazi barach of Lohani might have sindhi hindu Lohana ancestors some 1000 years ago, something that definitely would not please him.