Here are the admixture results using Reference 3 for Harappa participants HRP0181 to HRP0190.
You can see the participant results in a spreadsheet as well as their ethnic breakdowns and the reference population results.
Here's our bar chart and table. Remember you can click on the legend or the table headers to sort.
If the above interactive charts are not working, here's a static bar graph.
HRP0181 is half-Punjabi Jatt and half-English and the admixture results are not too different from the average of the reference British and our other Punjabi Jatt participants.
HRP0183, a Khatri, has fairly high European component, less than the Jatts but higher than most other South Asians.
HRP0186 is the most West Asian (and thus least European) of all our Georgian participants.
HRP0188,a Haryana Jatt, has the highest European component (29%) of all South Asians I think. I am surprised at the results for the two Haryana Jatts. I would not have expected their results to be much different from the Punjabi Jatts. If anything, I thought the Haryanavis would be less European than the Punjabis. Now I want to get a few non-Jatt Haryanavi participants. Anybody know someone?
Hi Zack, would it be possible to do the same thing except on a ref4C background? Like at k = 9 for example?
Zack, would it be possible to run an MDS or PCA plot with all the South-Asian participants + Caucasus-West Asian/Northeastern European/SE Asian reference populations? Also, I'm curious - the admixture proportions of the Kalash are a subject of great debate nowadays in genome blogosphere. What do they turn out like when based around your old K=12 components? Also, the last paragraph of your post, I think you mean - "HRP0188, a Haryana Jatt, has the highest European component (29%) of all South Asians I think."
Yes. The newest Haryanvi Jatt indeed has the highest European score of all the South-Asian participants. The following is a list of the South-Asians with the highest European score (top 25), in descending order;
HRP0188 (Haryanvi Jatt) - 29%
HRP0131 ([3/4] Haryanvi Tomar Jatt [1/4] U.P Jatt) - 27%
HRP0170 (Haryanvi Jatt) - 26%
HRP0093 [=HRP006 FTDNA data] (Punjabi Sandhu Jatt) - 22%
HRP0187 (Punjabi Lali Jatt) - 22%
HRP0033 (Rajasthani Brahmin) - 21%
HRP0021 (Kashmiri) - 21%
HRP0005 (Punjabi Gill Jatt) - 21%
HRP0008 (Punjabi Jhaj Jatt) - 20%
HRP0178 (Punjabi Khatri) - 20%
HRP0129 (U.P Brahmin) - 20%
HRP0162 (Punjabi/Gujarati) - 20%
HRP006 (Punjabi Sandhu Jatt) - 19%
HRP0085 (Thathai Bhatia) - 19%
HRP0108 (Halai Bhatia) - 18%
HRP0183 (UP Khatri (1/2), MP Khatri (1/4), Punjabi Khatri (1/4)) - 18%
HRP0062 (Shikaruri Sindhi Vaish) - 18%
HRP0136 (Punjabi Ramgharia) - 18%
HRP00163 (U.P Brahmin) - 18%
HRP0004 (Punjabi Brahmin) - 18%
HRP0003 (Bihari Brahmin) - 17%
HRP0077 (Bengali Brahmin) - 17%
HRP0125 (Punjabi - unspecified) - 16%
HRP0132 (Balochi/Punjabi) - 16%
HRP0073 (Punjabi Tarkhan) - 16%
They (the Jatts) exhibit the component in question at rather elevated amounts for a South-Asian populations, even relative to Punjabi speakers, and as much as the Pashtun reference samples do, and sometimes even more than the Pathans, which is interesting given their geographical location. So yes, as observers have commented before, absorption of some Saka/Scythian subtrate to an extent doesn't seem implausible at all in the case of the Jatts. The problem with the K=11, is, the European scores are not fully representative of actual North/Northeastern European admixture given the lack of a West-Asian centered component. Is there any way we can incorporate that in a run just to investigate the main European influence among South-Asians? K=11 is very indicative, but it doesn't peg the NEU score down to the actual percentage. It overestimates the North Euro score, compared to, say, the initial K=10 Dodecad admixture analysis or the Eurogenes Eurasia K=10 analysis; which both have West-Asian modal components.
you assume too much with scythian and jatts. Show me more then four random samples that could represent the vast amount of jatts that do exist. The idea is still implausible especially with such a low data count.
The ANI values for HRP0185 (Tamil Iyengar) and HRP0190 (Hebbar Iyengar) are both 57%. For HRP0189, it s 55%. For HRP 0184 (Kannada Christian), it is 52%. I have tabulated below the ANI, ethniciy and presumed religion of Punjabis with no recent foreign admixture. If there are mistakes in my attribution of religion, please correct them.
Ethnicity Religion ANI
HRP0188 Haryana Jatt Hindu 77%
HRP0170 Haryana Jatt Hindu 76%
HRP0131 UP/Haryana Jatt Hindu 76%
HRP0006 Punjabi Jatt Sikh 75%
HRP0008 Punjabi Jatt Sikh 74%
HRP0187 Punjabi Jatt Sikh 74%
HRP0005 Punjabi Jatt Sikh 73%
HRP0126 Punjabi Jatt Sikh 73%
HRP0178 Punjabi Khatri Sikh 74%
HRP0183 UP,MP,Pun. Khatri Sikh 73%
HRP0004 Punjabi Brahmin Hindu 70%
HRP0019 Punjabi Brahmin Hindu 68%
HRP0136 Punjabi Ramgarhia Sikh 68%
HRP0162 Punjabi,Gujarati Muslim 72%
HRP0086 Punjabi,Sindhi Muslim 72%
HRP0107 Punjabi Muslim 69%
HRP0125 Punjabi Muslim 69%
HRP0073 Punjabi Muslim 69%
HRP0012 Punjabi Muslim 68%
HRP0135 Punjabi Rajput Muslim 67%
HRP0111 Punjabi Muslim 67%
HRP0106 Punjabi Muslim 66%
HRP0064 Punjabi Muslim 61%
The data suggest that for West Eurasian ancestry:
Hindu Jatt > Sikh Jatt > Muslim Jatt
I would explain this in terms of the strict observance of endogamy by Hindu Jatts. Sikh Jatt men have probably married non-Jatt women to some extent and Muslim Jatts have done so more extensively.
HRP0086 is half Punjabi and half Sindhi, both sides being Hindu. The Punjabi side is Khatri (Zack might want to account for that in the spreadsheet?). I am the one who referred them/solicited them into the project, hence..
I don't think it's appropriate for you to assume what a participant's religion is.
^ Being a touchy issue, I would most certainly agree.
@Balaji, I don't think it's accurate to automatically assume that the religious affiliation of Panjabi Khatris is Sikhism. HRP0178 happens to be my wife and her family is definitely Sanatani Hindu.
There is nothing "sensitive" about it apart from making spurious conclusions based on very incomplete data and an almost fundamental misunderstanding of the complex geo-demography of the Punjab (and its two halves).
From a Pakistani perspective it seems virtually all the samples are stemming from the Majha region when there are several more sub-regions with different histories.
At any rate what I find interesting about this project is how quickly counter-intuitive conclusions are being grasped at. This really does remind me of the studies in the last decade where the Palestinians were treated as Bedouin foreigners whereas all Jewish people were genetically homogeneous. If that were the case then why does Israel have the second highest skin cancer rates in the world after Australia?
Much in the same place we can ask the same questions the cliches about Muslims and Pakistanis have a "certain look" stems from somewhere. Somehow denying any sort of exogenous link is not only counter-intuitive but absolutely ahistorical (where did the swarms of immigrants during a 1200yrs of Muslim presence exactly go then).
People like Zack Ajmal (quarter Egyptian) or even myself (half Iranian) are not a fractional minority; its a recurring phenomenon in the Islamic world.
I honestly and geniunely do not give a damn what the ANI/ASI proportions are but what I find somewhat irking is the conclusions being reached so quickly (the subliminal focus on upper-castes and the Euro component). Its great that Zack embarked on this project but I find the casteist/communalist mindset of some emerging; this is the sort of time when I realise just how different Pakistanis & Indians really are whatever our "genomes maybe". For Pakistanis the sliver of a nation doesn't really matter (no one cares about ancestral proportions or even race, despite a focus on colour and features) but it seems in India caste remains so important.
You are overreacting. People are merely interested in the source and history behind these ancestral components. No one is denigrating the validity of your ethnic identity as Pakistanis and Muslims, etc; nor are people being supposedly being casteist by discussing the dynamics behind the statistics produced by the project.
I see what I see & there is a fundamental difference between mere interest and morbid fascination. On limited analysis to go against the grain of commonly understood knowledge is always a fallacy.
Perhaps this can be treated as a useful exercise in confidence?
Anecdotal opinions based on forums seems to indicate that Pakistanis are the most concerned about skin color and denigrating South Indians based on physical appearance. You can go to Anthroscapes Forum or Biodiversity forum, and there is always a Paki trolling about this.
There may be an individual or two morbidly fascinated by this, but to make sweeping statement on generalities is idiotic. The population admixture differs here in european or Onge content, and its human nature to be inquisitive about where the admixture came from. There is no superiority or inferiority associated with admixture results. It is what it is.
Dear Friend,
Your work is interesting, if the data is segregated into groups, based on geography it will be more interesting. Data should be grouped into country wise and further state wise, then analysis will be easy. All the best. Jeyakumar Ramasami.
I have been thinking of posting group average results for the ethnicities represented in the project participants. Thanks for the suggestion.
I agree with Zachary Latif. I thought the focus of this project was South Asia, not Europe. Isn't there a Eurogenes project for that?
I had no idea that so many people would be so fascinated with their European scores. All of the commentary seems to be dominated by this subject. Boring.
I would like to sincerely apologize to anyone who has been offended by my speculation as to the religion of their birth. Religion is of course inextricably tied up with ethnicity and such individuals as HRP0026 (Goan Catholic Brahmin) and HRP0053 (Kerala Muslim Rawther) have specified their religion. Others may not want to disclose such information and I am sorry if I have offended them.
I have tabulated below the SW Asian admixture of HRP participants from West Punjab, East Punjab and Haryana.
HRP0035 Punjabi (3/4), Egyptian (1/4) 24%
HRP0158 Punjabi (3/4), Egyptian (1/4) 24%
HRP0001 Punjabi (3/4), Egyptian (1/4) 24%
HRP0161 Punjabi (3/4), Egyptian (1/4) 21%
HRP0006 Punjabi Jatt 15%
HRP0086 Punjabi (1/2), Sindhi (1/2) 15%
HRP0126 Punjabi Jatt 14%
HRP0183 UP, MP, Punjabi Khatri (1/4) 14%
HRP0008 Punjabi Jatt 13%
HRP0012 Punjabi 13%
HRP0135 Punjabi Pahari Rajput 13%
HRP0178 Punjabi Khatri 13%
HRP0107 Punjabi 12%
HRP0125 Punjabi 12%
HRP0111 Punjabi 12%
HRP0187 Punjabi Jatt 11%
HRP0005 Punjabi Jatt 11%
HRP0162 Punjabi (1/2), Gujarati (1/2) 11%
HRP0073 Punjabi 11%
HRP0106 Punjabi 11%
HRP0170 Haryana Jatt 10%
HRP0004 Punjabi Brahmin 10%
HRP0019 Punjabi Brahmin 10%
HRP0188 Haryana Jatt 8%
HRP0131 UP/Haryana Jatt 8%
HRP0136 Punjabi Ramgarhia 8%
HRP0064 Punjabi 6%
Zack Ajmal and his family have elevated SW Asian percentages because of their Egyptian ancestry. However the HRP participants from West Punjab do not appear to have higher SW Asian numbers than people from East Punjab.
Ramgarhia Punjabi has one side of his family from west punjab and the other is in India. You are assuming too much these calculations. SNI may be a proxy for onge but its not for certain.
I have tabulated below the Onge numbers in descending order for the Punjab HRP participants. Onge is a proxy for ASI.
HRP0064 Punjabi 23%
HRP0019 Punjabi Brahmin 19%
HRP0106 Punjabi 18%
HRP0136 Punjabi Ramgarhia 18%
HRP0012 Punjabi 17%
HRP0135 Punjabi Pahari Rajput 17%
HRP0111 Punjabi 17%
HRP0125 Punjabi 16%
HRP0073 Punjabi 16%
HRP0004 Punjabi Brahmin 16%
HRP0008 Punjabi Jatt 15%
HRP0178 Punjabi Khatri 15%
HRP0107 Punjabi 15%
HRP0005 Punjabi Jatt 15%
HRP0162 Punjabi (1/2), Gujarati (1/2) 15%
HRP0006 Punjabi Jatt 14%
HRP0086 Punjabi (1/2), Sindhi (1/2) 14%
HRP0183 UP, MP, Punjabi Khatri 14%
HRP0188 Haryana Jatt 14%
HRP0161 Punjabi (3/4), Egyptian (1/4) 13%
HRP0126 Punjabi Jatt 13%
HRP0187 Punjabi Jatt 13%
HRP0170 Haryana Jatt 13%
HRP0131 UP/Haryana Jatt 13%
HRP0035 Punjabi (3/4), Egyptian (1/4) 12%
HRP0158 Punjabi (3/4), Egyptian (1/4) 12%
HRP0001 Punjabi (3/4), Egyptian (1/4) 12%
Owing to their Egyptian ancestry Zack and his family are at the bottom of this list. Other than them, it appears that participants from West Punjab have higher ASI than those from East Punjab.
Do we know the geographical origins of all the Punjabis?
The criterion that I used to determine whether an individual is from West Punjab or East Punjab is to see if they have specified their ethnicity in some detail (Jatt, Khatri, Brahmin, Ramgarhia etc.) These I assumed are either Sikh or Hindu and from East Punjab. The other individuals I assumed are Muslim and from West Punjab. This criterion will not always lead to correct results and AV has posted a couple of cases of this. Also with partition there was an exchange of populations. In addition Muhammad Usman Butt and others have correctly pointed out that it is not valid to generalize from the few self-selected HRP participants to the genetics of Punjab.
That's a fairly poor way to figure out ancestral location.
Also, I am not sure if it makes sense to divide Punjab into East & West for our purposes. Religion/community is one factor, but there are also the environmental and cultural differences between the northern plateau/mountainous regions, western border region, southern arid areas and the river valleys etc.
There are a number of possibilities - east west with the Ravi (which bisects the middle-manjha) region as the border. North-south with the northern doabas and the southern multani. Even in the north there are slight dialect differences among the malwais, majhails, and the eastern doabis. We could also divide using Lahnda (western, Sindhu region) and Punjabi (Panchnad region). http://en.wikipedia.org/wiki/File:Dialects_Of_Punjabi.jpg
The question remains whether these distinctions are meaningful and whether we can resolve the genetics using admixture analysis.
So few unknown people calling themselves pakistani punjabis will now define pakistani punjab genetics, do you even know that there are atleast four cultural and hence racial zones in pakistani punjab, central punjab is similar to indian punjab in terms of language and culture but northern , western and southern parts of pakistani punjab are quite distinct in terms of culture and linguistics from central punjab and these regions are around 80 % of pakistani punjab.
Thank you Rani for your kind support.
Perhaps being an "outsider" (of Pakistani origin with some Iranian heritage but now a British subject) I am very discerning to the "casteist" mindset, which perhaps may not be as easily manifested to others.
The fact is that nothing is touched on our commonalities we all have a huge South Asian mix.
Instead the focus is how "European" or "Onge" or "ANI" or "ASI" one is; which has clear implications. Start with the assumption of unity and end with the presumption of unity; the trivial differences really do not matter and it should not matter in the virtual sphere where in the subcontinent lower castes & dark skins are so heavily reviled.
Also please don't get me wrong I geniunely enjoy good analysis like the sort Zack A. & Razib do even Dienekes has some great stuff up. Its scientific & rigorous nevertheless statements like "West Punjab is more ASI than East Punjab" "Muslims because their ancestry is more cosmopolitan are therefore more ASI".
I think everyone casts findings into their biases. It has been discussed before that NW Indians (and Pakistanis) are not easily classified by caste at all. These are tribal groupings. This is clearly tough for people from South or East India to completely grok because that is not the way those societies are organized. People marry within tribes, not within castes in the Northwest. Secondly, the sample sizes are so miniscule here that genericizing based on them is nutty. Let's wait until we have 15 at least of each group before we do it. And by group, I mean tribes for the NW subcontinent. If you have 4 Haryana Jatts and 4 East Punjabi Jatts and 4 Muslim Jatts total randomness will result if you attempt to extrapolate. Also, remember that different Jatt groups will have varying histories. We cannot say *anything* about Muslims or Hindus or Sikhs because tribal groups in the Northwest do not operate like that. If you want to do this, you need to compare Muley Jat (http://en.wikipedia.org/wiki/Muley_Jat) with UP/Haryana Hindu Jatts. Prediction: you will find that Hindus and Muslim Jatts are very alike. The "biradari" effect is extremely powerful in the Northwest. The "caste" effect is weak. Rajputs and Khatris do not intermarry by and large (regardless of religion) - it doesn't matter one whit whether they're the same official caste or not. They don't care. To get a glimmer into the mindset here, read this: Hindu Rajasthani Gujjars identify much more closely with and would rather vote for a Kashmiri Muslim Gujjar in elections than a Meena Rajasthani Hindu. Because "a Gujjar is a Gujjar." What caste are Gujjars? No one knows and no one cares, you can classify them whichever way you want.
[Disclaimer: I've become irregular about reading this blog and may not respond in a timely manner]
Also I would like to know about family backgound and geographical affliation of the people referring to themselves as pakistani punjabi here in this admixture results. I really cannot grasp how pakistani punjabis could be less west eurasians than indian punjabis when all people in my city gujranwala look far west eurasians than any indian punjabi on any given day, I am in germany since many years and have seen and interacted countless indian punjabis and for them I am always a turk,arab or afghan.
I think Xing et al. have used limited pakistani punjabi samples in their admixture study , these guys are arains and arains are one of the oldest muslim punjabi population in pakistan, their admixture results are more credible than some unknown punjabis from pakistan on harappa dna project.
Dienekies has posted the admixture results for the Arain from Xing et al. on his Dodecad website.
https://docs.google.com/spreadsheet/ccc?authkey=COCa89AJ&key=0ArAJcY18g2GadDUyeEtjNnBmY09EbnowN3M3UWRyNnc&hl=en_US&authkey=COCa89AJ#gid=5
From this I calculated ANI of 73% for the Arain
What about the other Xing Indian populations?
I think ANI is pretty vague term, the most important is west-asian component that is pretty high in arains that might explain why we pakistanis look more similar to west asian populations in our phenotype than indians including indian sikhs or other indians. Thanks.
Here are the ANI for the other Xing et al. people.
AP_Brahmin 59
AP_Madiga 43
AP_Mala 41
TN_Brahmin 60
TN_Dalit 42
Irula 37
Nepalese 53
What's the formula you use to calculate total ANI/West-Eurasian scores based on Dodecad v3 K=12? Are you participating in the Dodecad project yourself, btw?
ANI=East_European + West_European + Mediterranean + West_Asian + Southwest_Asian + 14.7 + 0.19*South_Asian
I am participating in the Dodecad project and the ANI that I calculated from it for myself agreed with the value that I calculated from HRP within 1%.
You should probably leave a message in the ancestry thread. That'd help viewers put your results into context.
The 73 % is average for 25 participants, it does not tell about the individual scores of the participants which must range above and below 73 %.
And another thing no one in pakistani punjab refers to himself as punjabi in terms of racial classification , people always refer to themselves with their cities and their particular tribal or biradari affiliation, so why people here are referring to themselves and pakistani punjabi as no such thing in terms of racial classification exist in pakistani punjab, they should write their city and surname. Thanks.
Zachary...I give you Mohammad Usman Butt to counter your point about, "this is the sort of time when I realise just how different Pakistanis & Indians really are whatever our "genomes maybe".
I counter that we are exactly the same in our "narrow mindedness" about these issues.
The Reference 3 data on the Harappa website is worth looking at in detail.
https://docs.google.com/spreadsheet/ccc?key=0AuW3R0Ys-P4HdGI3V2Z0SEs5WmRPcVoybDJXNzRIWXc&hl=en#gid=9
The Bene Israel and the Cochin Jews are quite interesting. Both communities have clearly mixed with the local Indian people as shown by their Onge values of 19% and 24% respectively. But their Jewish origin is evident in higher values of SW Asian. Also their ANI (or Total West Eurasian) is 74% and 66% respectively which is higher than that of the other communities among whom they lived.
http://en.wikipedia.org/wiki/Cochin_Jews
http://en.wikipedia.org/wiki/Bene_Israel
In addition the Bene Israel have 1% E African ancestry. Ashkenazy Jews have 1% E African admixture while Sepharadic Jews have 2% E African and 1% W African.
It is interesting to look at the Ref 3 African data both for its own sake and for the analogy to South Asia. At K=11, Zack has three African components, W African, E African and San/Pygmy.
. . . . . . . . W African . San/Pygmy . E African . Total African
hadza . . . . . . . 12% . . .. 26% . . . .. 61% . . . .. 99%
maasai . . . . . .. 25% . . . . 5% . . . .. 59% . . . .. 89%
somali . . . . . .. 20% . . . . 4% . . . .. 47% . . . .. 71%
ethiopians . . . .. 10% . . . . 3% . . . .. 46% . . . .. 59% EAST
sandawe . . . . . . 27% . . .. 19% . . . .. 43% . . . .. 89%
bantu kenya . . . . 66% . . .. 11% . . . .. 21% . . . .. 98%
luhya , , , , , , , 67% . . .. 12% . . . .. 21% . . . . 100%
san . . . . . . . .. 6% . . .. 80% . . . . . 1% . . . .. 87%
mbuti pygmy . . . . 26% . . .. 74% . . . . . 0% . . . . 100%
biaka pygmy . . . . 48% . . .. 52% . . . . . 0% . . . . 100% CENTRAL/SoUTH
bantu southafrica . 70% . . .. 27% . . . . . 3% . . . . 100%
yoruba . . . . . .. 92% . . . . 8% . . . . . 0% . . . . 100% WEST
mandenka . . . . .. 87% . . . . 8% . . . . . 3% . . . .. 98%
The W African component is now largely dominant throughout Africa except in Ethiopia and Somalia and is attributable to the spread of agriculture from West to East with the Bantu expansion.
http://en.wikipedia.org/wiki/Bantu_expansion
Similarly in the Subcontinent, the migration of agriculturalists from West to East must account for the cline of ANI.
Below are listed the African components of North African, Middle Eastern and South Asian populations from Ref. 3.
. . . . . . . . . . W African . San/Pygmy . E African . Total African
moroccans . . . . . 19% . . ... 1% . . . ….... 11% . . . .. 31%
mozabite . . . . . . 17% . . .. 1% . . …... .. 11% . . . .. 29%
yemenese . . . . .. 12% . . .. 2% . . ….. .. 10% . . . .. 24%
egypt . . . . . . . . . 8% . . .. 1% . . ….. .. 13% . . . .. 22%
bedouin . . . . . .. .. 3% . . .. 0% . …... . . . 9% . . . .. 12%
jordanians . . . . . . 4% . . .. 0% . . …... . . 7% . . . .. 11%
saudis . . . . . . . ,. 2% . . .. 0% . . …... . . 7% . . . ... 9%
palestinian . . . . . . 3% . . .. 0% . . .. .... . 6% . . . … 9%
lebanese . . . . . . .. 2% . . .. 0% . .. . …. . 4% . . . ... 6%
syrians . . . . . . . . 2% . . .. 0% . . . ….. . 4% . . . ... 6%
iranians . . . . . . . . 2% . . .. 0% . . . . ….. 1% . . . ... 3%
druze . . . . . . . . . . 0% . . .. 0% . . .... . . 2% . . .. ... 2%
samaritians . . . . . . 0% . . .. 0% . . ... . . 2% . . ... ... 2%
spaniards . . . . . . . 1% . . .. 0% . . …. . . 1% . . . .... 2%
cypriots . . . . , . . .. 0% . . .. 0% . . … . . 1% . . . ... 1%
ethiopian-jews . . . .9% . . .. 3% . . ... .. 45% . . . .. 57%
yemen-jews . . . . . . 0% . . .. 0% . . ... . . 8% . . . ... 8%
morocco-jews . . . .. 1% . . .. 0% . . . ... . 3% . . …... 4%
sephardic-jews . . . . 1% . . .. 0% . . . .. . 2% . . ….... 3%
georgia-jews . . . . .. 0% . . .. 0% . . .. . . 2% . .,,,. ... 2%
ashkenazy-jews . . .. 0% . . .. 0% . . .. . . 1% . . ….... 1%
bene-israel-jews . . . 0% . . .. 0% . . .. . . 1% . . .,, ... 1%
iranian-jews . . . . . . 0% . . .. 0% . . .. . . 1% . . ... ... 1%
iraq-jews . . . . . . . . 0% . . .. 0% . . . .. . 1% . . ... ... 1%
uzbekistan-jews . .. . 0%. . .. 0% . . . .. . 1% . . …..... 1%
cochin-jews . . . . , ,. 0%. . .. 0% . . . .. . 0% . . …. ... 0%
azerbaijan-jews . . . .0%. . .. 0% . . .. . .. 0% . . …..... 0%
siddi . . . . . . . . . .. 50%. . .. 9% . …..... . 1% . …...... 7%
brahui . . . . . . . . . . 2%. . .. 1% . …..... . 1% . …...... 4%
sindhi . . . . . . . . . . 2%. . .. 1% . …...... . 0% . …...... 3%
balochi . . . . . . . . . 2% . . .. 0% . …..... . 0% . …....... 2%
The North Africans (Moroccans, Mozabite and Egyptians) have 22% to 31% sub-Saharan admixture. Some admixture is to be expected since there has been trade and movement of people across the Sahara but the Islamic slave trade is surely also a large factor. in bringing the sub-Saharan components to North Africa. An analogous situation prevails in the Subcontinent where the Thar desert separates Pakistan and India and there is 20% to 30% ASI in Pakistan. Here also the arrival of Islam must have increased the ASI in Pakistan.
Another interesting fact is that Moroccans and Mozabites in North-West Africa have 11% E African. But the Africans south of these populations have 0% E African. Therefore the E. African element in Moroccans and Mozabites must have come from East African captives shipped all the way across the Mediterranean. Similarly Egypt is in North-East Africa and yet has 9% W African and San/Pygmy and this must have come from Africans captured from far to the south.
The African component in Bedouins, Jordanians, Saudis, Palestinians, Lebanese and Syrians is mostly E African and a lot of it is likely from Ethiopian slaves (Habshi). But all these populations also have some W African which must come from Bantu African slaves. It appears that ports in Yemen were where most of the Bantu slaves were landed before being sent to other parts of the Arab world. Yemenese have more W African than E African.
Iranians, Siddi, Brahur, Sindhi and Balochi all have more W. African than E African indicating that the African element in these populations is more Bantu than Ethiopian.
Spaniards have 2% African admixture and this must be a legacy of the time when Muslims ruled Spain.
Ethiopian Jews have mixed so much with other Ethiopians that they are not genetically distinguishable from them. Most Jewish groups have some E African admixture. Morocco Jews and Sepharadi Jews also have 1% W African. The Jews from Yemen have 8% E African which is not surprising considering the proximity of Yemen to Ethiopia But have 0% W African compared to 12% for Muslim Yemenese.
Like the Jews some other people in the Eastern Mediterranean also have 1% to 2% E African. The Cypriots have 1% E African. The Samaritans, a tiny endogamous sect, that lives in Palestine has 2%. The Druze are a Muslim sect but apparently have not inter-married with Muslims of other sects because they have only 2% E African.
There are four HRP participants (HRP0010, HRP0081, HRP0082 and HRP0147) who are Assyrian. These are Christians from Iraq and Iran. All have 0% total African. Muslim Arabs are HRP0118 (Egyptian), HRP0042 (Iraqi) and HRP0043 (Iraqi). They have 16%, 10% and 2% total African respectively.
An interesting nineteenth century figure was Sidi Mubarak Bombay, an African who was enslaved and brought to India and then gained his freedom and returned to Africa. He assisted the British in their exploration of Africa.
http://en.wikipedia.org/wiki/Sidi_Mubarak_Bombay
I made a mistake in tabulating the South Asian populations. Here is the corrected table.
. . . . . . . . . . W African . San/Pygmy . E African . Total African
siddi . . . . ….. . . 50% . . …..... 9% . . . . . 5% . . . ... 64%
makrani . .. . . . .. 5% . . …..... 1% . . . . . 1% . . ... ... 7%
brahui . . ... . . . . 2% . . …..... 1% . . . . . 1% . . ... ... 4%
sindhi . . … . . . . 2% . . .. …... 1% . . . . . 0% . . . ..... 3%
balochi . . .. . . .. 2% . …........ 0% . . . . . 0% . . . ..... 2%