I have added the HarappaWorld Admixture results for HRP0273-HRP0283 to the individual spreadsheet.
I got two participants from the Geno 2.0 Project. While I have calculated their HarappaWorld Admixture results, please note that Geno2 has only about 14,000 SNPs in common with HarappaWorld. Thus these results are very noisy.
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.
We got our first Pashtun participants, one Afghan and one Pakistani. Both have very similar results and are not much different than the HGDP Pathan sample average in their South Indian component.
HRP0278, a Bengali (mostly), is more East Asian components than any other Bengali participants (including my friend Razib.)
If I did the 2.0 and they list me as K1 and I did a mtDNA full sequence test and it lists me as K1e how will I allow for the difference or will it matter?
The full sequence test gives you your detailed Haplogroup while other tests base their results only on the SNPs tested.
I understand that. I ran the admixture test and it say that I am 12.194% Romany. Is that the same as Romanian? I have seen the tygray on your page of admixtures and that is pretty close to mine. I just do not see where the Southwest Asian comes from. My family is from Europe.
Are you talking about the Oracle? It does not tell you that you are 12% Romany. Instead it uses your and reference groups' admixture results to figure out which is the combination with the least error. Sometimes that can be very wrong. What are your HarappaWorld admixture results? And where in Europe are you from?
Southwest Asian can possibly be due to Middle Eastern (for example Jewish) or North African ancestry. Southern Europeans also seem to have a small amount.
I do not know. I am not sure how to load those up and I did not get them written down before the site went down. I am Northern European and my family tree does trace to the UK and Northern European. I do not know if you can look at them but my kit number is F176798. I am not sure if we used the Oracle or not it said. I do not know which one I used to get these results from. I am trying to learn all of this. I wish I would have gotten more information from my parents before they passed away.
1 utahn-white @ 3.368
2 N-European @ 3.911
3 utahn-white @4.353
4 british @5.442
5 orcadian @ 8.189
6 French @ 9.143
7 Hungarian @10.759
8 Slovenian @11.919
9 romany @12.194
10 ukranian @ 17.804
350 iterations
These are not percentages, they are distances. 12 is a large distance. It shows that you are closest to the reference samples from Utah, but not quite the same.
I see. What is Utahn exactly? When you say distance, does that mean that I am 3.368 away from Utahn? How do I upload my results to your database? Thanks for your help.
Residents of the state of Utah in the United States. These samples have Northern and Western European ancestry.
The distance means that your admixture results are that far away from the average admixture results of the Utahn samples.
Unfortunately, you don't have ancestry from the regions I am interested in.
the ratio is different from both my parents re: southeast vs. northeast.
Razib, what do you think would explain the higher East Asian admixture in HRP 0278? Zack, you listed the participant as Bengali (mostly)? Do you know if he/she has some recent East Asian type ancestry?
HRP0278 has one great-grandparent from Hyderabad.
Yes. Your parents are from SE Bangladesh, right? HRP0278 is probably from somewhere with more NE Asian influence.
comilla. near the border with tripura.
That seems unlikely, considering that both my parents are from northwest Bangladesh, at least as far as I know up to my great-grandparents. As for you, Razib, might I know how the admixture percentages with respect to southeast: northeast differs?
Also, I am curious to know if it is possible to date the north east asian admixture from the fragmentation level, since it is so far off from the other Bangalis in the sample.
i'm in the database. hrp002 or something? you can look, but it's mostly southeast. looking at your 23andMe chromosome painting i'm wondering if you've had recent admixture from tibet or something. the 'asian' segments are quite long.
As I said earlier, I was going to ask my parents concerning their grandparents, and I made a cursory comparison of google images for the various indigenous and/ or tribal groups from the West Bengal and Northwest Bangladesh region since it seemed reasonable that it would explain the NE Asian and SE Asian admixture percentages, and the ratio between the two. In the case of my mother's paternal grandfather, I was told that he looked 'completely foreign' and based on pictures, showed similarity with the Bhutia, Limbu people (mother's relatives), and the same goes for some of my maternal aunts and uncles, and especially my maternal grandmother. In the case of my father's paternal grandfather, my father described his maternal grandfather as 'somewhat foreign looking' and based on pictures, show similarity to the Rabha tribe. However, none of my grandparents have any memory of non-Bengali origins, or migrations of their ancestors, just as I suspected.
I was incorrectly informed that my mother's paternal greatgrandfather was from Hyderabad. Actually, their origins are unclear, but from my maternal side, it would appear that my mother's paternal and maternal grandfathers had lived/ moved in Darjeeling, Murshidabad (not Hyderabad, which was in error) and Jalpaiguri. Her maternal and paternal grandmothers have always been in northwest Bengal (Nilphamari). As for my father, they all have lived in northwest Bengal (Dinajpur) at least upto four generations before me.
Thanks Zack, for this great undertaking. It correctly pointed out that I was in error. 🙂
So according to your knowledge, you are 100% Bengali? Should I correct that in my ethnicity spreadsheet?
Yes, that would be correct. Also, I can't help but notice that in the map, the other Bengali participants are from the Southern areas, and I am the only one from the northwest. Given that northwest Bangladesh is separated from the rest of Bangladesh by the Padma and Jamuna rivers, I have to wonder if similar percentages would be found in Bengalis from that region, making me representative of the mean, and not an outlier. I would imagine that travel over flat land from northwest bengal all the way to India and Tibet might have been easier than across rivers in earlier times. And yes, I guess that would make me at least 100% Bengali, if not 100% Bangladeshi.
Ajay, your stereotype of an afghan is just that. Afghan tend to have more robust features than the people from the plains (Punjab). Most people will also agree that on average an ethnic pashtun from afghanistan will have lighter skin than punjabis, but when you start talking about facial features , that a bit hard to measure unlike skin color, there are plenty of pashtuns who could pass for being punjabi or vice versa.
I personally feel what is more interesting is the high North Euro component in the Greater Punjab area manifested clearly in some of the populations in that area. It peaks there and drops off the further north and west you go. Pashtuns have a higher SW Asia component but lower North Euro component than some of the Indian groups.
Who seriously cares if Pashtuns are lighter skinned than other north western South Asians? I just cannot for the life of me understand the odd assumptions and obsessions that are being manifested here. Such a discussion is only possible amongst people of South Asian descent with some real leisure time, and a reliable internet connection. Besides, since when do real human populations sort out into “types� Biology is inherently messy, and variation/diversity is the central reality. Recognition of this is what made Darwinian thought so revolutionary. Also, genetic variation in our species is very slight. We humans are a rather homogenous bunch. If you want to contest this, just look at the patterns of genetic differentiation found in our closest ape relatives. The reason I note this is because Pashtuns and Punjabis are neighboring populations. If sub-Saharan Africans and Native Americans/Australasians (the most genetically differentiated humans) are so genetically close as to belong to a single subspecies, I do not think Pashtuns and Punjabis are any big deal. From a purely phenotypic perspective, Pashtuns and Punjabis overlap to a great extent. Many Pashtuns could pass for Punjabis, and this goes in the other direction as well. I am not denying the existence of some consistent differences. Most Pashtuns have a heavy West Asian vibe, and most Punjabis have a heavy South Asian vibe. This has a lot to do with small differences in nasal form, skin pigmentation, and craniofacial gracility. Also, light haired/eyed Pashtuns are a much more substantial minority than light haired/eyed Punjabis (but do note the word minority, as the vast majority of Pashtuns are as dark haired/eyed as Punjabis). I am well aware of the fact that “vibe†isn't exactly a scientific term, but it works here. We are describing differences that are very minor, blurry, ephemeral, and subjective.
I am familiar with Mardan, Rawalpindi and AJK and in all honesty people do generally look much alike. There is a fair mix of lighter, darker eyes and skins amongst the traditional farming classes. Jats, Rajputs or Indus Pathans. Mostly the flatter/wider nosed or darker people are from the earliest settlers in this region. Those who have married from places like Swat or Dir have offspring who resemble the ancestry of their maternal line quite often . I guess it's to do with what genes people inherit from both parents that determine looks. It's that simple. Both by parents are Jat from and so were my grand/grandparents maternal and paternal. This I guess is reflected in my admixture.
So are you a Pakistani Punjabi Jat because your ethnicity isn't listed on the list of participants?
Paul, that's correct. I think I did list this. My clan is more Potohar and AJK then Punjab plains or central to east.
Your haplogroups are rather interesting, especially your Y-DNA. You seem to be on of the few Y-DNA G South-Asians here other than HRP0072 and HRP0142. Have you mailed Ray Banks or Ted Kandell from the Y-DNA G Project for further inspection of your Y-raw data file? Or have you already gone about some degree of exploration and discussion about your Y-DNA? Any implicative inferences?
I've emailed the Y results to Ted. Hopefully he will make some sense of them soon. I've read that the ftdna is more precise in determining recent ancestory. On 23and me there is one 3rd to 5th cousin match except the other participant has not responded.
How accurate are these cousin predictions ?
Interesting to see a Pakistani Punjabi Jatt participant. All the other Jatts are Indian Punjabi or Haryanvi.
Look at the ethnicity spreadsheet. Your ethnicity as HRP0283 isn't listed.
https://docs.google.com/spreadsheet/ccc?key=0AuW3R0Ys-P4HdGE4eDh6emt1dUs2U2pXTkVjS0lsV1E&hl=en#gid=0
Perhaps admin will look into it. I am not familiar with genetics, I took the test for medical reasons and this is the added bonus.
haplo group G2a1c is considered rare but I don't think many have been tested in this region.
OK, I wont troll anymore, but I will bring proof of what I am saying. It is clear that the nationial geographic numbers are different look at this page: https://genographic.nationalgeographic.com/overview-of-regions-and-closest-
populations/reference-populations/
^ look at how different north indians are from Pamiri Tajik for example, I know they might be combining all north indians, but still, it gives a clearer picture of different indians are overall from the populations west of them
Also at least one pashtun on that site matches much more with Pamiris and Iranians, than he does with north Indians. There are 4-5 pasthuns there, so not sure about the rest
They have far less reference populations, which more or less makes the analysis inferior to Harappa, or any other open-source DNA project. Reich, Harappa, Dodecad and Eurogenes all have a higher number of reference populations and individual participants, hence giving a far more accurate picture of every ethnic groups' placement on the worldwide genetic canvas. That includes Pashtuns, who seem to be the bridging gap between Northwestern India/South-Eastern Pakistan and Central and West Asia. This makes sense, given their geography. I do not see what you are objecting to here. Pseudo-scientific discussions pertaining to phenotype have no place here. Either way, the extent of genetic difference between the Punjabis/Sindhis and Pashtuns is clearly reflected even in their genetic differences, and any other phenotypic differences can be accounted for by the Pashtuns' geographic location and environmental conditions. Now, for the love of God, please, give this incessant, unscientific rambling up and get a hobby (preferably outside of the realm of the internet).
well my point is Pashtuns/Kalash will be closer to populations like Pamiri Tajiks, than they are to any indic populations, since pashtuns/kalash and pamiris seem to have high amounts of Europid/Caucasian look compared to any indic groups. Also all these three groups are culturally related as well, especially pashtuns and pamiris because their languages are very close. Also the Caucasian/Europid look must be coming from some admixture which pashtuns, kalash, pamiris have and indian groups lack. We have to wait and see what that is. I really doubt all these looks are coming just because of environment factors
There are already Tajik samples on Harappa Ancestry Project which are more representative than an isolated group like Pamiri Tajiks. Their data is more informative due to Harappa's larger reference population. Also, even then, Tajiks are different than Pashtuns and Kalash in their autosomal admixture. Also, linguistics and culture is one thing but genetics is different. Pashtuns might be East-Iranian speakers but they are geographically and genetically closer to Northwest South Asians than modern day Iranians.
What NatGeo says for Southern, Western and Northern Indians and Pamiri Tajiks:
SI: "The 58% Southwest Asian component represents the first migration from Africa through Southwest Asia to the Indian subcontinent. It is likely the oldest component in Indian populations today."
WI: "The 58% Southwest Asian component represents the first migration from Africa through Southwest Asia to the Indian subcontinent. It is likely the oldest component in Indian populations today."
NI: "The 34% Southwest Asian component represents the first migration from Africa through Southwest Asia to the Indian subcontinent—it is likely the oldest component in Indian populations today."
PT: "As ancient populations migrated from Africa, they passed through Southwest Asia en route to India and Central Asia. This is the likely source of the 44% Southwest Asian in the Pamiri population"
Hi guys, I am interested in knowing how far back can autosomal testing determine one's admixture. Someone told me that after a few generations you can't really capture anything past 200 years.
Thanks for your help guys.
I think you are thinking of this:
http://www.familytreedna.com/faq/answers/default.aspx?faqid=17#628
This applies to matching segments with cousins. Admixture is a different beast altogether.
Thanks Zack, this helps a lot. Cheers.
I have noticed one important trend from harappa participants admixture results that Caucasus component among pashtuns, kashmiri and pakistani punjabis is always significantly hihger than their NE euro component. A totally opposite result is seen for punjabi jatts, haryanvi jats, indian brahmins of all type where NE euro component significantly higher than their Caucasus components. Actually Caucasus component for most of indian higher castes is less than even 5 %. Whereas for pakistani and afghan pashtuns, kashmiris and pakistani punjabis it is always well above 10%. I predict that the chiseled and beautiful features that differentiate pashtuns from other south asians are due to their higher Caucasian/SW asian component. And the common features shared by punjabi and haryanvi jatts with indian brahmins (as was noted by british anthropologists) are due to their higher NE euro component.
According to me it looks like a population with higher Caucasus component has admixed itself with pashtuns, kashmiris and pakistani punjabis , but the same population has affected the indian punjabis and jats and brahmins of indian to a much lesser extent. Also SW asian component is also higher in pakistani populations, I think significant higher values for Caucasus ans SW asian components reside in Middle east so an absorption of Muslim populations from middle east among pashtuns, kashmiris and pakistani punjabis might explain this unusual phenomenon. An anomalous result for Pakistani punjabis and pakistani kashmiris is value of south indian component which which is 4-5% higher than indian punjabi jatts, this is something that I really cannot explain given that pakistani punjabis and kashmiris geographically live on western side of indian punjab.
With regards
Also another trend that I have observed that pakistani punjabs, kashmiris and pashtuns are showing greater admixture values for east eurasian components than indian punjabis or higher castes. The explaination for this phenomenon could be explained due to the absorption of post islamic central asian turkics and moghuls among these groups.
With respect you do sound like Ajay with a different name!!
Symmetry and harmonious features is what makes someone 'pleasing to the eye.' You can find that anywhere where people are healthy, and take care of their appearance. If anything Caucus component is associated with semitic features....something that is not found in great concentration in the modeling industry worldwide.
Zack,
Your Individual Harappa Participants Ethnicity ends with HRP0267.
Admixture is upto HRP0283
Which spreadsheet are you looking at for ethnicity?
The Participant Ethnicity Spreadsheet. Looked at it again on the site and data is there. Who knows what I did.
Extract of Sri Lankan (Sinhalese and Tamil) Admixture Data.
http://sbarrkum.blogspot.com/2013/04/sinhalese-and-tamil-dna-admixture.html
The data is upto HRP0267.
Zack -I submitted my 23andMe genome data to your harappa address yesterday.
Ancestry/Ethnicity: Rajasthan Porwal Jain
Please check when you get a chance. Thanks.
I sent you an ID. Will process data over the weekend.
I was checking out the individual Iraqi arab results (HRP0042, HRP0043, HRP0277, HRP0279) and noticed that while the others are consistent, HRP0042 seems quite out of place with their Baloch, Caucasian, and particularly SW Asian percentages. They appear more like a mixture of Iraqi + Arabian Peninsula or East African.