HarappaWorld HRP0352-HRP0374

I have added the HarappaWorld Admixture results for HRP0352-HRP0374 to the individual spreadsheet.

Do note that the admixture components do not necessarily represent real ancestral populations. Also, the names I have chosen for the components should be thought of as mnemonics to ease discussion. I chose them based on which populations in my data these components peaked in. They do not tell anything directly about ancestral populations. The best way to look at these admixture results is by comparing individuals and populations. Finally, the standard error estimates on these results can be about 1%. Therefore, it is entirely possible that your 1% exotic admixture result is just noise.

I have also updated the group averages.


  1. This time, the results have been very interesting and surprising.

    The Iranian Bandari participant (HRP0355) is a South Asian! Their largest component is the "Baloch", which they have at a comparable level to HAP's Pashtun participants, and their "South Indian" is at a high 13%. They have slightly more "South Indian" than "SW Asian"! One HGDP Pashtun has less "South Indian" at 11.80% (HGDP00214, I'm not counting HGDP00220, whose results are too unusual). HRP0286 isn't too different in this respect. This individual is very different from the other Iranian participants, and very different from any public samples. If this individual had higher "NE Euro" in comparison to their "SW Asian", and they were not identified as Iranian, I'd assume this individual was a Pashtun. I guess there is a geographic cline in South Asian affinity. I always found it strange that South affinity would have a abrupt peak in Afghanistan. Based on HRP0355's results, I'd say eastern Iran is also rather South Asian shifted, and could even be considered a part of greater South Asia. Then again, HRP0355 could be a outlier. More participants from the region are needed.

    The Rajasthani Jats are more similar to the Haryana Jats. Does this imply that Punjabi Jats have been less endogamous?

    The Nepali Brahmin (HRP0361) is very East Eurasian shifted in comparison to the Nepali Khas Brahmin (HRP0362).

    The Baloch Khetran individual (HRP0363) also has very surprising results. Rather different from the HGDP Baloch, Brahui, and Makrani. If this individual is representative, I guess the Khetran are justified in claiming Pashtun origins. The Khetran claim they are originally Pashtuns, but have long ago been integrated into the Balochi tribal system. HRP0363's "Caucasian" and "South Indian" components are more in line with Pashtun ancestry. Nevertheless, their Balochi affinity is also apparent, as no Pashtun outside of the HGDP data-set scores anything close to 40% "Baloch". Also, HRP0363's "NE Euro" and "SW Asian" percentages are almost equal (6.8% and 5.3%, respectively), which is definitely unusual for a Pashtun individual, so Balochi ancestry is apparent.

    Finally, HRP0373 also has surprising results. In a rough sense, HRP0373 is 75% Durrani Pashtun, and 25% Afghan Tajik. I would expect such an individual to have a much lower "South Indian" percentage. All the other Durrani Pashtuns seem to have more "Caucasian" than "South Indian". And 25% Tajik ancestry should reduce South Asian affinity even further. But HRP0373 is slightly more "South Indian" than myself. I guess Durrani Afghan Pashtuns have more heterogeneity than I expected, and Afghan Tajiks are significantly more South Asian shifted than Tajikistani Tajiks. The Kashmiri ancestry is too small to shift things.

    • I would definitely want to see more samples from Gulf persians before concluding that the single Bandari sample here was not an exception. Not only does this individual have South Indian admixture that is appreciably higher than the closest Baloch population (Makrani baloch at ~ 10 %), their SW asian percentages seem lower than I would expect from a person from the gulf region. That said, I would not be surprised if typical Bandari south Indian percentages were in the 6-10% range. This would fit in with the numbers seen in the Khorasani Persians. It would also support the theory that this admixture was contributed to both the Baloch, as well as south-eastern Persians by the Jadgal (Sindhi Jatt) population that historically populated most of the southern Balochistan region stretching to the periphery of the Persian gulf. The Jadgal are the only Indo-Aryan speaking people found in modern Iran.

      The Khetrani Baloch results are interesting and definitely seem to agree with the generally accepted view that the Khetrani are 'assimilated' Baloch. It would be useful to know the geographic origin of their family/clan since the Khetrani can be found in Sindh/Panjab/Balochistan.

      • I'm very surprised at this Bandari individual's results. I can't believe they are basically a liminal South Asian, rather than a West Asian. Really interested if other people from the area are similar.

        The Khetrani individual is from Vehowa. Supposedly, the Barkhan Khetrani Baloch are less keen on claiming Pashtun origins than those in Vehowa. Is there any reason for this? Is the Baloch identity of the Barkhan Khetrani older?

        • I guess in the case of both individuals, a ChromoPainter/FineStructure run will be more illustrative.

          Concerning the Baloch, I wonder how much more variation we can find. The Baloch are very assimilative people, and unlike Pashtuns, Baloch tribal identity is not contingent on patrilineal descent from a common ancestor.

          • Not big news. Take a look at Iranian Zoroastrians from Yazd and its becomes clear that Iran is a South Asian country or at least highly South Asian influenced place.

          • Concerning the other Durrani person:

            Its a big mystery when and how Tajik, who have no affinity to marry with Pashtuns, entered our bloodlines. I can't find even a single instance of it happening in the past 300 years looking onthe royalark family trees. It must have happened a very very long time ago.

          • The Baloch do make a distinction between tribes of purely Baloch origin and those tribes accorded 'hamsayeh' status. All self-identify as Baloch to outsiders. However, the tribal identity of those tribes recognized as being of Baloch origin is definitely contingent on patrilineal descent from a common *Baloch* ancestor. In fact, a large number of Baloch tribes, particularly in Eastern Balochestan/Sind, trace their ancestry to a single tribe/ancestor (Rind).

            The Khetrani results really do match well with their reputed origin of being a Jatt tribe with Pashtun and Baloch admixture. The Pashtun admixture is most apparent in the non-negligible east-eurasian percentages. The HGDP Baloch samples display essentially no east-eurasian admixture.

    • I don't really consider the overall profile of the Iranian to be South Asian. Deviating from a typical Iranian, yes, but South Asian is a stretch. My brother's sample is the Parsi submission and we have similar numbers. To give some perspective, my 23 and Me and McDonald results are 100% consistent (67% Iranian, 33% South Asian--probably Sindhi according to Dr. McDonald). This individual's S-Indian component is even lower than ours, so I highly doubt he would read as South Asian, even with the elevated Baloch component. The Caucasian component is also just too high for a typical South Asian. I am not comparing this person to individuals, of which there are always gong to be plenty of outliers and overlaps, but to the averages which represent the overall population.

      To whoever mentioned Iranian Zoroastrians: Iranian Zoroastrians are not South Asian influenced. They are slightly darker than the majority of Iranians because the majority of Iranians inhabit the Northwest part of the country, closer to Armenia, Azerbaijan, and Turkey. Yazd is significantly hotter, drier, and further south than where the majority of Iranians reside. Their features remain distinctly Iranian though. I have met many Iranian Zoroastrians and there is no mistaking them for South Asian. Not all of them are dark, either.

      I'm open to being corrected, but I am not sure when or how a large scale South Asian influence in Iran would even happen. Historically, most of the influence between South and West Asia moved in a West-East direction, so if anything, Northeast India/present-day Pakistan are influenced by Iranians...not so much the other way around. Otherwise the average S-Indian component would be much higher for Iranians.

      • I'd say it becomes complicated in the case of your brother. His "Baloch" and "Caucasian" are equal. About Iranians, I can't really say anything, but with the Bandari participant, he/she is clearly a liminal South Asian, rather than a West Asian. On the north-south Eurasian axis in PCA-MDS plots, the Bandari participant will be further South Asian shifted in comparison to your brother. That dimension of variance isn't cleanly tied to the proportion of West Eurasian ancestry an individual has. For example, my mother is significantly more West Eurasian than myself, based on chromosome paintings, supervised Admixture runs, and PCA plots. She is just more closer to Europeans and West Asians than myself. Yet, she is significantly more South Asian shifted. In fact, she is almost a northwestern South Asian. Many northwestern South Asians will be slightly-moderately more West Eurasian than myself, but all of them will be significantly more South Asian.

        I guess we might begin to see much more heterogeneity among Iranians with more sampling.

      • I don't really consider the overall profile of the Iranian to be South Asian. Deviating from a typical Iranian, yes, but South Asian is a stretch. My brother's sample is the Parsi submission and we have similar numbers.

        Parsis have significant recent (=the last 1300 or so years since their arrival to the Subcontinent) South Asian admixture. This is clear from their mtDNA haplogroups (see Quintana-Murci et al. 2004) and now, it seems, also from their autosomal genetics.

        • My brother's results speak to our mixed heritage and history. Using 3rd party utilities, we generally get similar profiles...47% Armenian, 53% Punjabi, 75% Iranian Jew, 25% Gujarati, etc. Always X% West Asian, Y% South Asian. Interestingly, according to Dr. Mcdonald, who I think provided us with the most accurate results, we also have a strong influence from the Levant.

          Here is the map of our results:


          Sindh represents the farthest reach of the Umayyad caliphate, which expanded eastward to present-day Pakistan at roughly the same point in time that the Parsis migrated from Iran to India. That, I believe, would explain the pull from Syria.

          As to where the Parsis historically came from--it's not entirely clear. Most likely the north east, where present-day Mashhad is, as well as parts of Turkmenistan and Afghanistan. Who knows? There are no real historical records, only an oral recounting of history that was authored hundreds of years after the supposed migration. Not exactly reliable! My parents look as though they are from two different countries entirely. My mother (and her family) look 100% Iranian--west Iranian, even. Not even a little bit Indian or South Asian. My father looks very South Asian. Part of me wonders if there might have been multiple migrations occurring over a period of time, as opposed to the typical story of just two that occurred centuries apart.

          As for our mTDNA, I am not sure what to make of it to be honest. Our haplogroup peaks in Rajesthan, but is one of few subclades of M that can be found in insignificant numbers in Iran: http://link.springer.com/static-content/images/613/art%253A10.1186%252F1471-2156-5-26/MediaObjects/12863_2004_Article_198_Fig1_HTML.jpg

          I'm not sure how to interpret that. I'm inclined to say it is, in our case, South Asian in origin.

          This is also a good read pertaining to several South Asian populations with origins (claimed or verified) in other regions: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC447589/

          • The Parsi paternal lineages are overwhelmingly Iranian, while their maternal lineages are overwhelmingly South Asian. Autosomally they are somewhere in between, as expected.

      • Although it begs the question as to whether the elevated South Indian component is due to the supposed origin of Zoroastrianism from further east such as Afghanistan or Tajikistan.

        • Before the spread of Islam almost all Persians and other Iranic ethnic groups of Persia (=ancient Iran) were Zoroastrian. Your theory may be correct if Parsis descend from the southeastern parts of Greater Iran (such as what are today Pashtun and Balochi lands). Do you have any historical info about where Parsis came from?

        • By "your theory" I mean explaining the relatively high South Asian ancestry of Parsis with descent from an Iranic people who had already relatively high South Asian ancestry. As I said, the only candidate Iranic populations for that seem to be in what are today Pashtun and Balochi lands.

  2. I submitted HRP0370 and HRP0374.

    Just wanted to point out that HRP0374 gained around ~0.25% each in San/E African/Pygmy going from the GEDmatch DIY Harappa calculator to this official one and HRP0370 gained almost half a percent in each category.

  3. The Nepalese results are consistent with the Xing et al results. I wonder if this guy's theory is correct even though he has an agenda:

    Start from chapter 2.
    I am however not sure about the origin of the Khasas.

    • The Khasas are supposed to be Eastern Iranic, who moved Eastward along the shadow of the Himalayas to Western and Central Nepal giving their names to various places like Kash-gadh (Kashgar), Kashi in Xinjiang?, even Kash-mir?. They are mentioned in the same breath as the Kamboj people in the puranas.The first admixture occurred when the Khasas encountered the Kiranta people along the Himalayas.

      Now, the Kiranta people were Sino-Tibeto-Burman tribes who had moved Westward from Southern China, Tibet, Burma region towards the Eastern Himalayas. According to historians tribes encountered each other around the Gandak region in Central Nepal. These two tribes, sometime fought each other, sometimes had alliances. Even in present day Nepal, Sino-Tibeto-Burman and Indo-Aryan ethnic groups sometime share the same last name--Thapa, Budha, Bista, Rana etc which is interesting because their marriages are mostly endogamous ie within the same ethnic group. They harken back to a time when the two tribes intermarried, perhaps to form alliances? You will find considerable East Asian admixture in the Khasas, particularly the Chhetri (Kshyatriya) Khasas but also in Brahmins.

      The Khasas were outside the influence of the Indic Hinduism of the plains until as recent as a 1000 years ago. These people were mostly Shamanistic.

      The ruling Rajputs and Brahmins from the plains of India started migrating to Nepal from 1000 AD onward fleeing Muslim persecution. When they brought Hinduism with them into Nepal, they conferred Brahmin and Kshetriya titles to their Khasa allies. This period is known as a period of Sankritization of Nepal when Hinduism started to take hold and caste structures became entrenched.

      The ruling monarchy of the House of Gorkha which ruled Nepal until recently actually trace their lineage back a petty Rajput clan of Chittor, The famous Gurkhas are said to have originated after Bappa Rawal, a Rajput, was given divine inspiration to protect the cow. Hence: Gau (Cow)+ Rakh (Protector)= Gorkha. Ironically, most Gorkhas today that fight in Indian and British armies are neither Rajputs of the plains nor the Khasas of the hills, but hail from the Sino-Tibeto-Burman tribes! Strange are quirks of history.

      Nepalis of the hills are, on one end Iranic-Khasas, and on the other end Sino-Tibeto-Burmese Kiratas, or a whole range of spectrum in between. Then, there are people of the plains--the Madhesis-- who are identical to the people of the Indo-Gangetic plains and relatively a recent addition to the tapestry of the population.

      So Nepalis are basically

      1. Either Iranic Khasas (about~30%)
      2. Sino-Tibeto-Burman (about~30%)
      3. Madhesis (about~30%)


      4. A bewildering array of combinations in between.

      This is without mentioning that the land below the Himalayas was inhabited by Australo-Asiatic people before either the Khasas, Kirantas or Madhesis arrived.

  4. We appreciate the work that you are doing.

    I would like to point out here that it might be incorrect to club all the Keralite Christians as a monolithic entity, because they are not. From my minimal knowledge, I would suggest that the following grouping would (more or less) be appropriate in the context of genetic analyses.

    1. Syrian Christian (Northist) - Origins uncertain (as per Britannica Encyclopaedia). These are Christians who existed before contact with Europeans and who claim that they are natives converted by St. Thomas the Apostle. They form the majority.

    2. Syian Christian (Southist) - Origins disputed (or uncertain). They claim that they are migrants from the Middle East. This is a very small community.

    3. Latin Catholics - This community originated during European contact, through conversion of natives by the Portuguese. They are sizable in number.

    Other smaller communities include Ango-Indians (mixed European and Indian ancestry) and Christian Dalits.

    I fear that if you group Keralite Christians as a single entity, they might average out to the general Keralite population. We might miss the subtle (or possibly pronounced, if the folklore is indeed correct) differences between the aforementioned sub-groups.


  5. @ Raz - I'm quite surprised, I didn't know that the Baloch had "hamsayah" groups as well. I thought this was an exclusively Pashtun social formation. I guess this must be a common phenomena in the eastern Iranian plateau.

    Nevertheless, what I meant was that all Pashtun tribes claim descent from a single progenitor, while Baloch tribes are more heterogeneous in terms of self-conception. All Pashtun tribes, from Herat in the west to Attock in the east, will tell you that they are descendants of a single man, "Qais Abdur Rashid". A Pashtun tribe can only claim to be Pashtun if it is somehow linked by descent to this individual. I thought the Baloch do not conceptualize their broader tribal dynamic in this manner. It was my understanding that lineage grafting has taken place very frequently among the Baloch (this is rather difficult in a Pashtun context, even though it has definitely occurred). Then there are people like the Brahui, who I'd assume are considered Baloch, but obviously have different historical origins than the Baloch proper. Also, there are the Dehwari, who correct me if I'm wrong, are Baloch, but can also be construed as the Tajiks of this region.

    You're obviously very knowledgeable about Baloch anthropology and history, so I'd love it if you could answer a few questions I've always had about the Baloch (I've found that the scholarly literature is very bare). Do the Baloch have tribal confederations? If so, are these groupings tied by common descent? Do the Baloch have any specific oral traditions concerning their pre-Islamic history and origins? Since Balochi is a Northwestern Iranian language, is it reasonable to say that the Baloch proper originate from the western end of the plateau? Also, do the Baloch have a tribal code of conduct similar to "Pashtunwali"? If so, are revenge and hospitality as important values in this tribal system as they are in "Pashtunwali"? I'm sorry for asking so many questions, but I've been interested in answering them for a long time. High quality material on Baloch ethnography and history is virtually non-existent (or perhaps I'm just not having any luck at finding such material). If you could also point out some nice scholarly material, I'd really appreciate it.

    I have to agree with you about the Khetran Baloch individual. The HGDP Baloch don't seem to display Turkic admixture. I guess Turkic contact was very minimal in this region. In the case of the Baloch, I'd say contact with the Arabian peninsula was as strong as Pashtun contact with Turkic Central Asia. I think a fairly robust connection with the Arabian coast and peninsula explains the slight-minor African percentages found in the HGDP Baloch. But besides the East Eurasian percentages, the best support for Pashtun admixture comes from this individual's "Caucasian" component. His percentage is much higher than either the Baloch, or other Northwestern South Asian populations.

  6. What's the point in running blog if you are going to delete comments that you don't not like?

    • The point of the blog is to get an insight into South Asian genetics, not to provide a platform to every internet asshole to spout off.

  7. So mentioning that 31% of Indian Punjab is Dalit (official Indian government stats), to add to other lower/middle castes on top of that, makes me an asshole? I am just saying the Punjabi sample you have on Harappa is not truly reflective of the Punjabi population as a whole. All I am saying is, we need more people from Punjab from all sections of society to submit their DNA before we can get a clearer picture of the region as a whole.

  8. Any ideas why such high Baloch % for Orcadians and British?

  9. Zack,
    You need an open discussion trhead like Razib does on gnxp.... What is your take on this? Can you get the sequenced genome for MA-1 ( The 24000 Ma'lta boy R* sample) and compare them with your existing south asian database using chrompainter?


    From Page 52:
    " SI 9.1.3 Results
    At K = 9, MA-1 is composed of five genetic components of which the two major ones
    make up ca. 70% of the total. The most prominent component is shown in green and
    is otherwise prevalent in South Asia but does also appear in the Caucasus, Near East
    or even Europe. The other major genetic component (dark blue) in MA-1 is the one
    dominant in contemporary European populations, especially among northern and
    northeastern Europeans. The co-presence of the European-blue and South Asian-
    green in MA-1 can be interpreted as admixture of the two in MA-1 or, alternatively,
    MA-1 could represent a proto-western Eurasian prior to the split of Europeans and
    South Asians. This analysis cannot differentiate between these two scenarios. Most of
    the remaining nearly one third of the MA-1 genome is comprised of the two genetic
    components that make up the Native American gene pool (orange and light pink).
    Importantly, MA-1 completely lacks the genetic components prevalent in extant East
    Asians and Siberians (shown in dark and light yellow, respectively). Based on this
    result, it is likely that the current Siberian genetic landscape, dominated by the genetic
    components depicted in light and dark yellow (Figure SI 6), was formed by secondary
    wave(s) of immigrants from East Asia. "

    • The Gedrosia component breaks down into Caucasian plus minor Siberian per Dienekes. Must be an archaic admixture between the two, probably before IVC.

      Yeah, I would like to see the Harappa breakdown of this Mal'ta boy.

      • the south asian refferred to here does not refer to Gedrosia, but peaks in south Indian Sakkili.

        • SB has a good point. Based on his Admixture results, the Malta boy would definitely qualify for participation in Harappa Ancestry Project ;-). He could be a South Asian if he were alive today, as the component that is modal among low-caste South Indians is the largest component of his ancestry (37% South Asian). Still, the high European (34%), Amerindian (26%), and Oceanian (4%) are extremely exotic and unique. Perhaps the South Asian affinity-relationship isn't "real", but simply a reflection of massive genetic drift across the centuries.

          • I feel the south asian and papuan reveal a migration of his ancestors from somewhere south and east of India (haplogroup P), through India (Splitting into R and Q), through Pakistan/Afghanistan( ignoring his Amerindian, he does match Brahui/Balochi) and onto siberia, where the European and Amerindian affinities precipitated. But, unless we run a harappaworld admixture, this picture will remain fuzzy. Maybe his "south asian" will mostly besplit among Gedrosia and South Indian.

    • Zack,
      As suggested by SB, have you considered analyzing and including the Mal'ta MA1, Afontova Gora 2(and Denisova) data? I think it would be really helpful for the understanding the prehistory of South Asia.

      "Sequence data for MA-1 and AG-2, produced in this study, are available for download through NCBI SRA accession number SRP029640. Data from the Illumina genotyping analysis generated in this study are available through GEO Series accession number GSE50727; PLINK files can be accessed from http://www.ebc.ee/free_data. In addition, the above data and alignments for the published modern genomes, Denisova genome, Tianyuan individual and the two ancient genomes are available at http://www.cbs.dtu.dk/suppl/malta. Raw reads and alignments for the four modern genomes sequenced in this study are available for demographic research under data access agreement with E.W."

  10. The data is freely available here http://www.evolutsioon.ut.ee/MAIT/public_data/malta/

    The supplements are really fascinating. I don't really know what one should make of the Malta individual's South Asian affinity.

    It seems that prehistory was far more interesting and complex than we could ever imagine. Who'd have thought that Native Americans and Europeans have a deep genetic connection, one that goes both ways (it seems European have substantial East Eurasian ancestry tied to Native Americans, and Native Americans have substantial West Eurasian ancestry).

    Zack has a such a substantial data-set now, with some really interesting populations. Once he can acquire the "Afghan Hindu Kush" samples, the Moorjani et al. samples, and all the other new participants he gets in the meantime, I hope Zack does a huge global ChromoPainter/FineStructure analysis, one which includes African populations. It'll take forever, but our patience will surely be repaid in such a case.

  11. The Afghan sampes are finally here!


    I'm really excited, and very anxious to see their HarappaWorld results.

  12. What do you guys think of McDonald's Analysis?

    • What about it? It is another BGA analysis like many other. You get a different perspective from each of them.

      • I ask because his results were quite different than any other results.

        • Would you be willing to post your results here?

          I personally feel that Dr. McDonald has the most solid analysis on offer. This is due to his utilization of PCA. The population fits that constitute the flesh of his analysis are based on the 9-14 largest dimensions of variance. If you're from an ethnic group that is in his data-set, you're results will be pretty congruent with your actual background. FTDNA's "Population Finder" test is basically a modification of his algorithm, and has the same reference populations. Nevertheless, they've thoroughly messed it up. But representatives of FTDNA claim they've made it more "flexible". So "flexible", that I'm predominantly Palestinian with some substantial Siberian admixture! But I digress.

          Various results, assuming mixture. A lot is contingent on how Dr. McDonald tweaks the PCs, adding and removing populations:

          Sindhi=73% Chuvash=14% Iranian=13% or

          Sindhi=64% Adygei=19% Burusho=17% or

          Pathan=91% Iranian=6% Hazara=3% or

          Pathan=92% Armenian=5% Hazara=3% or

          Pathan=94% Chuvash=3% Bedouin=3% or

          Pathan=86% Burusho=8% Druze=4% Koryak=1%
          Bedouin_Sou=1% or

          Pathan=88% Makrani=8% Hazara=4% or

          Iranian=47% N_India=45% Uygur=8%

          My top 6 single-population fits, assuming no mixture, based on 14 dimensions. Ordered from most likely to least likely. These are understood to be very poor, since the algorithm tries to fit you as 100% of a single population.
          1. Pathan=100%
          2. Balochi=100%
          3. Sindhi=100%
          4. Brahui=100%
          5. Makrani=100%
          6. Burusho=100%

          He sends chromosome paintings as well. With three continents:


          He never gave me percentages for the standard painting, the one with South Asian, Middle Eastern, and Native American categories. But eyeballing it, something around half South Asian and half European, with some large Middle Eastern, East Asian, and Native American segments.

          • Also, when I say "ordered from most to least likely", I really mean ordered from "best" to "worst" fit. So, on the first 14 dimensions of the PCA Dr. McDonald computed, I'm closest to the HGDP Pashtuns, followed by the HGDP Baloch, followed by the HGDP Sindhi, etc. He sent me my top 16 single-population fits. If I find the rest of it, I'll post it. But I do remember that my 7th, 8th, 9th, 10th, and 11th best fits were, respectively, Iranian, North Indian, Turkish, South Indian, and Kalash. I guess this makes sense. I probably am closer to Iranians than to North Indians, and I probably am closer to Turks than to South Indians. But I'm closer to both North and South Indians than I am to the Kalash, since they behave unusually on lower dimensions of PCA. In this case, Admixture/Structure is superior to PCA-MDS, since with some parameter tuning, it becomes obvious that I'm more similar to the Kalash than to West Asians and peninsular South Asians (it's just that I don't share much genetic drift with them). Similar problems happen in regards to my mother; since she shows evidence of some sort of minor inbreeding, she has her "own" unique dimensions in PCA-MDS! And I'm not even joking here, she actually has whole dimensions specific to herself. She's like an isolated island on the PCA-MDS sea. And she even clusters with the Kalash and Brahui-Makrani (I've made my own PCA-MDS plots) in weirder dimensions! Although, perhaps something more interesting is happening in her case?

  13. I am waiting for the afghan samples to be calculated by zack. My guess is their south indian will be low, maybe around 15-17% on average, their Caucasian will be probably higher. one thing for sure is that afghans and pashtuns have been influenced by a movement of people from that area, perhaps this also shows up on their facial features. I just met a pashtun person today, my god, she could easily fit in one of those caucasian countries

  14. Where can I access the results? Has Zack calculated it yet? What's the sample size? Are tajiks included? I expect both to have low south Indian component( 11-17), high gedrosia and Caucasian component.