PDA

View Full Version : Spatial Ancestry Analysis (SPA) results



Dr_McNinja
09-18-2013, 06:09 PM
I learned about it from the Eurogenes blog:

http://polishgenes.blogspot.com.au/2012/11/spa-for-23andme-clients.html

http://bga101.blogspot.com/2012/11/spatial-ancestry-analysis-spa-model.html

And so here are my results:

(world.model) http://i.imgur.com/s7wryg9.png

Yeah... that didn't really help at all.

(europe.model) http://i.imgur.com/xGaQmpp.png

And the Eurasian model seemed about right I think: http://i.imgur.com/MHUPvHPh.jpg Halfway between Europe and South India?

Any others care to post their results?

parasar
09-18-2013, 08:52 PM
...
(world.model) http://i.imgur.com/s7wryg9.png

Yeah... that didn't really help at all.

(europe.model) http://i.imgur.com/xGaQmpp.png

...


23andme would be much more accurate for placement with their huge database and feedback correlations to get clearly defined AIMs (http://en.wikipedia.org/wiki/Ancestry-informative_marker). They should place you reliably in South Asia.

Dr_McNinja
09-19-2013, 12:00 AM
23andme would be much more accurate for placement with their huge database and feedback correlations to get clearly defined AIMs (http://en.wikipedia.org/wiki/Ancestry-informative_marker). They should place you reliably in South Asia.Yeah that's useless for determining more precise relationships to other individuals though, they just highlighted all of Afghanistan, Pakistan, India, Nepal, Bangladesh, and Sri Lanka. 23andMe customers who used SPA could compare their exact relative location results. I might have gotten middle of Iraq and someone else might have gotten Turkey, or Iran or something. The map underneath doesn't matter as much as relative distance between individuals. So just wondering what results other South Asians were getting? It would be similar idea to PCA.

geebee
09-19-2013, 08:48 AM
Hmm. Maybe I did something wrong, or maybe my background is just too weird. I tried it, ended up with Romania for one ancestry, and somewhere between Ireland and Greenland for the other. (Closer to Ireland, but still out in the Atlantic.)

geebee
09-19-2013, 09:13 AM
Okay, I tried it again with my dad's file, first treating him as unadmixed. The location it came up with was Belgium. Given that my father's supposedly half German and half British, seems like a fair splitting of the difference.

When I tried again with the -n 2 switch, got (1) a location somewhat to the west of Ireland, though not as far as I'd gotten for myself; and (2) a location on the border of Romania and Ukraine. Fun, but still not sure I fully understand it. But there are a great many things I don't understand. Less and less all the time, I think. Somebody keeps jacking up the amount of stuff yet to know, darn it.

geebee
09-19-2013, 09:45 AM
"World" is pretty weird, though. I seem to be an West Atlantean. :)

AJL
09-19-2013, 10:41 PM
You can go directly to the site here:

http://genetics.cs.ucla.edu/spa/23andme.html

I ran my mother's results on mixed mode, and got the Bay of Biscay and the North Sea. The average of these, though, is around Dieppe, which makes sense given her predominant Isles/France/Germany/Netherlands mix. I suspect even on mixed mode it is Population Finder–like, in that the spot is really still an average arrived at by guesswork.

Sein
09-29-2013, 04:54 AM
Hi Dr_McNinja,

What is your precise ethnic background? My spot on the map is in Turkey for the Eurasian Cline model. Since your spot is in Iraq, are you Punjabi? I'm Pashtun.

Dr_McNinja
09-30-2013, 05:52 AM
Hi Dr_McNinja,

What is your precise ethnic background? My spot on the map is in Turkey for the Eurasian Cline model. Since your spot is in Iraq, are you Punjabi? I'm Pashtun.Yes. Well, my admixture results suggest an origin deeper in India than even Punjab but otherwise yeah, I'm Punjabi.

Dr_McNinja
09-30-2013, 06:00 AM
Hi Dr_McNinja,

What is your precise ethnic background? My spot on the map is in Turkey for the Eurasian Cline model. Since your spot is in Iraq, are you Punjabi? I'm Pashtun.
By the way if you've used Geno 2.0, you can do an admixture analysis using Dodecad: http://dodecad.blogspot.com/2012/11/geno-20-patch-for-diydodecad.html

Sein
09-30-2013, 06:56 AM
That's really interesting. If my location in central Turkey is construed as Eastern Afghanistan-Northwestern Pakistan, than northwestern Iraq is around the Pakistan-India border in Punjab, maybe a bit east of Lahore. I tried the Eurasian Cline model on Razib Khan's raw-data (he's made it public, and he's encouraged people to use it), and his location is in far southeastern Iran, almost in Pakistani Balochistan. Taking our locations into account, that is basically Bangladesh. So, this tool really does work, but it's placing South Asians far too west of where we should be. If a lot of South Asians use this tool, and record their actual regions of origin, the distances will make perfect sense, as will the general distribution/shape of the locations.

I also have my 23andMe raw-data. I've utilized those tools before. I'm HRP0281 on Harappa Ancestry Project.

Dr_McNinja
09-30-2013, 07:00 PM
That's really interesting. If my location in central Turkey is construed as Eastern Afghanistan-Northwestern Pakistan, than northwestern Iraq is around the Pakistan-India border in Punjab, maybe a bit east of Lahore. I tried the Eurasian Cline model on Razib Khan's raw-data (he's made it public, and he's encouraged people to use it), and his location is in far southeastern Iran, almost in Pakistani Balochistan. Taking our locations into account, that is basically Bangladesh. So, this tool really does work, but it's placing South Asians far too west of where we should be. If a lot of South Asians use this tool, and record their actual regions of origin, the distances will make perfect sense, as will the general distribution/shape of the locations.

I also have my 23andMe raw-data. I've utilized those tools before. I'm HRP0281 on Harappa Ancestry Project.Ah, cool. I'm HRP0349. Do you know what tribe of Pashtun your family is?

And yeah, that's what the posts about SPA said, their locations will be way off but the relative plots between individuals would be highly relevant.

Sein
09-30-2013, 11:16 PM
The relatives always change the story, but it seems this is, for all intents and purposes, the family background. My mother is 50% Afridi, 25% Marwat, and 25% Khalil. My father is 50% Khalil, and 50% of his tribal background is unknown. Both sides of the family look exceedingly different.

Dr_McNinja
10-01-2013, 12:16 AM
The relatives always change the story, but it seems this is, for all intents and purposes, the family background. My mother is 50% Afridi, 25% Marwat, and 25% Khalil. My father is 50% Khalil, and 50% of his tribal background is unknown. Both sides of the family look exceedingly different.Since most of those tribes are in or near Pakistan, that would put me somewhere between Jaipur and Lucknow which is what I figured, admixture-wise.

Dr_McNinja
10-01-2013, 12:29 AM
Btw, you mind posting screenshots of your results and Razib's results on this?

Sein
10-01-2013, 01:18 AM
I think your placing yourself far too south and east. Your admixture percentages are very similar to the Punjabi Brahmin, Punjabi Ramagarhia, and "generic" Punjabi (probably mostly Pakistani) participants at HAP. By contrast, the Meghawal samples and Rajasthani participants are rather different, and the UP samples even more so. The only UP samples that are similar to you are the HAP UP Muslims. But, this group clearly has substantial West Asian admixture. I think most of these participants claim cosmopolitan origins outside South Asia proper (Arab, Persian, or Pashtun ancestry) and this seems correct, because they're very different from the Metspalu UP Muslim samples. That's why their group average resembles Punjabis rather than other North Indians, because Punjabis are mostly North Indian, but with substantial West Asian-European admixture. Although you're more South Asian shifted than most Punjabi Jatt participants at HAP, you're results are perfectly within Punjabi variation, and very unusual for North Indians.

To put your location in context, I think we must remember a few things. The HGDP Pashtun samples were acquired from Parachinar (or it's environs). In other words, the samples are almost from Afghanistan (Parachinar juts into that country). I am sure the samples are representative of most Pashtuns in Eastern Afghanistan. I'm slightly less South Asian than the HGDP Pashtun average (much less "Baloch" on HAP, which is a South Asian-specific West Eurasian component, and slightly less "South Indian"). So, even if we ignore the fact that the HGDP Pashtun samples are probably representative of most Afghan Pashtuns, my location on the world map for the Eurasian cline is probably in Paktia-Ghazni. Dr. Douglas McDonald's "world averages" tend to place me in the Zabul-Kabul axis (I've had 4). Also, the Indus River is responsible for some genetic discontinuity. In that case, you end up somewhere in the greater Punjab. I am sure most non-Jatt Punjabis will cluster close to you.

Dr_McNinja
10-01-2013, 05:01 PM
I think your placing yourself far too south and east. Your admixture percentages are very similar to the Punjabi Brahmin, Punjabi Ramagarhia, and "generic" Punjabi (probably mostly Pakistani) participants at HAP. By contrast, the Meghawal samples and Rajasthani participants are rather different, and the UP samples even more so. The only UP samples that are similar to you are the HAP UP Muslims. But, this group clearly has substantial West Asian admixture. I think most of these participants claim cosmopolitan origins outside South Asia proper (Arab, Persian, or Pashtun ancestry) and this seems correct, because they're very different from the Metspalu UP Muslim samples. That's why their group average resembles Punjabis rather than other North Indians, because Punjabis are mostly North Indian, but with substantial West Asian-European admixture. Although you're more South Asian shifted than most Punjabi Jatt participants at HAP, you're results are perfectly within Punjabi variation, and very unusual for North Indians.

To put your location in context, I think we must remember a few things. The HGDP Pashtun samples were acquired from Parachinar (or it's environs). In other words, the samples are almost from Afghanistan (Parachinar juts into that country). I am sure the samples are representative of most Pashtuns in Eastern Afghanistan. I'm slightly less South Asian than the HGDP Pashtun average (much less "Baloch" on HAP, which is a South Asian-specific West Eurasian component, and slightly less "South Indian"). So, even if we ignore the fact that the HGDP Pashtun samples are probably representative of most Afghan Pashtuns, my location on the world map for the Eurasian cline is probably in Paktia-Ghazni. Dr. Douglas McDonald's "world averages" tend to place me in the Zabul-Kabul axis (I've had 4). Also, the Indus River is responsible for some genetic discontinuity. In that case, you end up somewhere in the greater Punjab. I am sure most non-Jatt Punjabis will cluster close to you.That's interesting, what do you mean by the indus river being responsible for some genetic discontinuity?

The thing is the non-Jatt Punjabis usually have significantly lower NE-Euro. Even the other Pakistani/Muslim Jatt Punjabis have lower NE-Euro and higher West Asian which clusters them around Arain Punjabi, which makes sense since the Muslim Jatts and Arain have probably been intermarrying for some time. Mine are inbetween them and the Sikh Jatts who still remain relatively endogamous. Plus there's the high %s of North/East Asian which seems a mix (I've seen the Jatt Muslim group with high East Asian and the Sikh Jatts with high North Asian (Arctic)).

The family history is that we're just Punjabi Jatts (mostly from Hoshiarpur) who had some Sikh branches back in the 1800s. That would explain all of the above. But not the low West Asian. The 35-37% South Indian could even pass for maybe Uttar Pradesh or Rajasthan or even more south, all areas where there are plenty of Jatts. But the low Baloch/West Asian is the confusing bit. It's lower even than the Rajasthanis/Meghawals. That's why I surmised an origin outside of Punjab, but among Jatt populations. I had a pretty decent one-to-one match with a Sikh Punjabi Jatt on 23andMe (75.95% if I remember correctly). Admixture on the other hand represents geographic signals within ancestry more than actual ancestry (especially along the paternal line).

Now consider Jatts from Haryana and just over the border in Uttar Pradesh who also look very different from both me and other Punjabis. Which is why I think we're from somewhere further East. DNA Tribes did peg my closest match with Tharus of Nepal/India and then Uttar Pradesh populations. The alternative is perhaps Gujarat, but their West Asian isn't usually that low, though there is greater variance of that component there.

In the Harappa spreadsheet, the up-muslim group averaged 35-34-10-8 (S-Indian, Baloch, Caucasian, NE-Euro) and the metspalu up-muslim group averaged 53-33-3-5... either way, the Baloch/West Asian seems about right between all UP groups. Maybe not in UP, maybe north or south of it, but in terms of east-west, to me those numbers seem like the closest fit. Otherwise the other areas with 33-35% Baloch are all in East/Southeast India, aside from the Pashtun (Nepal, Andhra Pradesh, Karnatka, Bihar, Bangladesh).

I think the S-Indian and Baloch are useful for geographic placing, but not so much the Caucasian or NE-Euro which vary by group or caste (Muslims in India generally having higher Caucasian and Brahmins/Kshatriya generally having higher NE-Euro).

The issue is there's a lack of data from populations in inner India as I was discussing with another user in my thread in the Member Results forum. There's virtually no information about populations in Madhya Pradesh or Chattisgarh (there was a significant Jatt presence in Madhya Pradesh at least).

I'm waiting on my mom's results from 23andMe to see if the numbers are the same from both sides of the family or if I can trace it specifically from my father's side. My paternal grandfather did come most recently from Hoshiarpur to Faisalabad in his own lifetime whereas the rest of my family (mom's side) emigrated from Hoshiarpur much earlier (19th century sometime). It could be I got high S-Indian, low Baloch and NE-Euro from him but high Caucasian from mom's side. We'll see.

EDIT: My Harappa Oracle results:


[1,] "up_harappa_5" "5.1599"
[2,] "punjabi_harappa_10" "5.9113"
[3,] "kashmiri-pandit_reich_5" "6.1242"
[4,] "singapore-indian-c_sgvp_10" "6.6743"
[5,] "bihari-muslim_harappa_4" "6.6761"
[6,] "nepalese-a_xing_12" "7.8431"
[7,] "brahmin-uttar-pradesh_metspalu_8" "7.8661"
[8,] "bengali-brahmin_harappa_6" "8.3394"
[9,] "vaish_reich_4" "9.5397"
[10,] "gujarati-b_hapmap_34" "10.0522"


[1,] "16% irula_xing_22 + 84% punjabi-jatt_harappa_8"
[2,] "15.8% pulliyar_metspalu_5 + 84.2% punjabi-jatt_harappa_8"
[3,] "81.5% meghawal_reich_5 + 18.5% stalskoe_xing_5"
[4,] "81.9% meghawal_reich_5 + 18.1% urkarah_xing_18"
[5,] "79.1% iyengar-brahmin_harappa_7 + 20.9% urkarah_xing_18"
[6,] "17.3% lezgin_behar_18 + 82.7% up-kshatriya_metspalu_7"
[7,] "78.7% iyengar-brahmin_harappa_7 + 21.3% stalskoe_xing_5"
[8,] "29.5% north-kannadi_chaubey_8 + 70.5% pathan_hgdp_23"
[9,] "82.3% up-kshatriya_metspalu_7 + 17.7% urkarah_xing_18"
[10,] "81.2% maharashtrian_harappa_4 + 18.8% urkarah_xing_18"

Sein
10-01-2013, 07:54 PM
You know, I think you've pointed out something very interesting. Most participants seem to be getting lower Baloch percentages, and higher East Eurasian percentages. I think we are seeing the "calculator effect" at work here, at least to some extent.

All of this information about Jatts is very fascinating. How exactly do these geographically widespread populations relate to each other on a socio-cultural basis?

I completely share your feelings on this, we really do need numerically robust samples that are collected on the basis of geographic contiguity. Central India is really missing in public datasets, and I don't think that Zack has a substantial amount of participants from the region (if he even has any participants from this region). That is the one downside of HAP, it has to rely on customers of genotyping companies to contribute their data. If one day we do see Jatt samples from deeper in India, I think we might find that you are absolutely correct, your family really does originate amongst these communities. Nevertheless, since your family does not have any memory of origins outside the greater Punjab, I'm betting you're results might just be a reflection of Punjabi Jatt variation.

That is very interesting, how exactly do the Jatts of Haryana and Uttar Pradesh differ from phenotypically from their Punjabi compatriots?

The Indus River is unique since it has always constituted the natural boundary for greater India (I mean, "India" is derived from "Sindhu"). In the past, rivers seem to have been moderately effective at slowing gene flow between neighboring populations. Although, I would qualify this by saying that Pashtuns who live in the Indus riverain probably have substantial admixture with Punjabi populations.

Dr_McNinja
10-01-2013, 09:31 PM
You know, I think you've pointed out something very interesting. Most participants seem to be getting lower Baloch percentages, and higher East Eurasian percentages. I think we are seeing the "calculator effect" at work here, at least to some extent.This is also a possibility. Harappa blog has been silent for a while, I wonder if Zack is making a new calculator or running everyone through it again like some were asking.


All of this information about Jatts is very fascinating. How exactly do these geographically widespread populations relate to each other on a socio-cultural basis?Jatts are more or less the indigenous populations of the area descended from the Indo-European waves from before 2000 years ago. As far as I know the word, 'Jat', literally just means caste or sub-caste. It was like an agricultural caste, but alienated from the Hindu caste system until much later (around the time Islam showed up, when many were pushed further into India and then started making their own states and marrying into local royal families) when they became permanently associated with the kshatriya caste. So they remained more endogamous than the other Indo-European groups.

The Jatt tribes had varying origins (from which their surnames came) but married into one another and formed their own distinct group based on their common identity of agricultural lifestyles and social status. Jatts are also probably in the ancestry of most groups in Pakistan/Northwest India.

Typical origins for most Jatts range from the first Indo-European visitors from the Central Asian Andronovo culture ( http://en.wikipedia.org/wiki/File:Indo-Iranian_origins.png ) to the late Indo-Scythians over a millenium later. There are probably Jatts who hail from the other various residents of the region too, maybe the Yuezhi/Tocharians, maybe even Indo-Greek/Greco-Bactrian (for some reason my paternal tribe, Gills, were said to have some kind of mixed Greek/Central Asian ancestry). Basically whenever a people invaded and settled into the region along the Indus to take up an agricultural lifestyle, they became Jatts/Jats.

The original residents of Sindh when the Arabs invaded were Jatts, Bhattis (identified by the Arabs as a separate tribe), Medes, etc. It's said the Arain Punjabi are the descendants of the Jatts who converted to Islam and mixed with those Arabs who stayed there. I've found more J2 (no sub-clades) among Arain Punjabis on 23andMe than any other. Whatever the case, I doubt there's any trace left of those few Arabs aside from that haplogroup (mostly found in Iraq, Syria, etc). In a similar fashion, Bhattis became Rajputs and Jatts. My maternal grandmother's mother was a rare Bhatti Jatt (most Bhattis are Rajputs now).

One of the theories of origins for Rajputs is the Indo-Hephthalites (Huna, or White Huns) who invaded India in the 4th-5th centuries if I remember correctly. They were finally stopped by apparently a Jatt kingdom in the northwest of Madhya Pradesh, almost exactly between Jaipur and Lucknow (so there were Jatts that far in India).

This "non-servile", "martial" (as the British put it) agricultural caste persisted until modern times where they alternatively fought for or against the Mughals, Marathas, and British. Eventually forming their own kingdom under the Sikh Emperor, Ranjeet Singh, in the Punjab. A lot of Jatts moved back to Punjab at this time as the rest of India fell to the British thus the association with Jatts as a traditionally Punjabi people though there's a ton of them all along the border with India all the way into Sindh (around 27 million in Pakistan, mostly Muslim, and another similar number in India split between Hindus and Sikhs). I would not be surprised if there were Jatt groups all across northern India and even in East India more than 2000 years ago.

Based on my haplogroup (J2b2*) and its heatmap (http://www.flickr.com/photos/ydna-haplogroup-j2-m172/8675629245/), I figure my paternal ancestry is probably from one of the earliest Jatt groups in India. It seems J2b2* (according to Eupedia) rode in along with the R1a Indo-Europeans (most Jatts are R1a like others of the region) but apparently only in one small early wave, and they spread the furthest into India (later groups didn't seem to penetrate as far into the subcontinent).

If you read any history of Jatts from their own people, the further back in time you go, every Indo-European group becomes one Jatt/Jat clan or another. Jat is also 'Got' in Sanskrit or something (since another word for sub-caste or caste is 'got', like in the word 'gotra' or clan... so people might say what 'jat/zat/got' are you). So they made up stories about how Gotland/Jutland up in Scandinavia are all ancestral Jats. It's fairly ridiculous but you see how they're associated with the early Indo-European groups in south Asia/South Central Asia from a really long while back. My late grandfather used to say Jatts hail from Haryana and it's an important place for them.

The furthest Jatt groups are in Ukraine with some in Russia and Afghanistan (Jati/Jakata who speak a Western Punjabi language, Jakati) and are Muslim gypsies. No idea how they got there (if they were there from the beginning, they probably got left behind as the Turkic tribes displaced the old Indo-European tribes).

I found this bit about Jatts on a Christian missionary website which contains all the local stereotypes: http://joshuaproject.net/people-profile.php?peo3=19369&rog3=AF (The Arabs said the Jatts were the fighters sent with war elephants to back up the Persian army by the Raja of Sindh, those who were enslaved converted to Islam and wound up working for the Arabs... Raja Dahir was a Brahmin Hindu who usurped the throne from the previous rulers who were Buddhist and since most of the Jatts and other similar tribes were Buddhist (and mostly had been until modern times), they wound up siding with the Muslims when they came in)


I completely share your feelings on this, we really do need numerically robust samples that are collected on the basis of geographic contiguity. Central India is really missing in public datasets, and I don't think that Zack has a substantial amount of participants from the region (if he even has any participants from this region). That is the one downside of HAP, it has to rely on customers of genotyping companies to contribute their data. If one day we do see Jatt samples from deeper in India, I think we might find that you are absolutely correct, your family really does originate amongst these communities. Nevertheless, since your family does not have any memory of origins outside the greater Punjab, I'm betting you're results might just be a reflection of Punjabi Jatt variation.Yep, I'm hoping my parent's results shed some light. If they're both similar to me, that might say something about the general community in Hoshiarpur or Faisalabad, where they moved.


That is very interesting, how exactly do the Jatts of Haryana and Uttar Pradesh differ from phenotypically from their Punjabi compatriots?I don't have much experience with them but I heard they had a reputation for being tall. I'd be curious to see what that Haryana Jatt in HAP with 20% NE-Euro looked like.


The Indus River is unique since it has always constituted the natural boundary for greater India (I mean, "India" is derived from "Sindhu"). In the past, rivers seem to have been moderately effective at slowing gene flow between neighboring populations. Although, I would qualify this by saying that Pashtuns who live in the Indus riverain probably have substantial admixture with Punjabi populations.Yeah, a lot of people probably flocked from both East and West of it to there, my ancestors almost certainly did.

Sein
10-01-2013, 10:13 PM
Lol, I forgot my own HAP ID. I'm HRP0282.

Dr_McNinja
10-02-2013, 03:14 AM
Oh yeah and regarding the calculator effect, I read some articles/studies posted on Dienekes' blog that Amerindian is present in Northern Europeans. It's possible that Siberian and Beringian admixtures are also relics from the ancestral NE-Euro component (since we know Indo-Europeans related to Northern Europeans entered India, but obviously Native Americans did not). Siberian is also high in Kazakhs and higher in Pashtun than in other South Asian groups so it could indicate Central Asian ancestry or links in Indians/Pakistanis while Beringian I'd group with Amerindian. Maybe higher numbers of one over the other could indicate the different waves or dates of ancestral NE-Euro admixture.

Hard to tell though because it seems there was certainly some noise/calculator effect going on in the HAP entries (can't tell any pattern with Papuan which I suppose should represent aboriginal tribal affinity).

Sein
10-02-2013, 03:41 AM
I really find the Amerindian-related admixture among Europeans exceedingly cool. Based on this preprint from the Reich lab, it seems all Europeans have some really substantial East Eurasian admixture, 20% amongst the Basques and Sardinians, and 40% amongst Russians http://arxiv.org/abs/1212.2555. And Russians aren't even the most East Eurasian shifted of Europeans, so I think the share of East Eurasian ancestry in Europe might be even higher than that. The paper was somewhat ambiguous about whether this admixture is pan-West Eurasian, or just specific to European populations. So I asked David Reich. He is a very busy man, so I was exceedingly pleased that he took the time to answer my questions. All I can really say, he's an awesome man. His response:

"We don’t think it is present in some West Eurasian populations from the Middle East (e.g. Bedouins) but we do think it is probably present in West Eurasians from Pakistan as well as in people from the Caucasus and people from Europe."

This basically confirms your intuition. Also, this means a substantial amount of the "ASI" percentages found in northwestern South Asians, and South-Central Asians, actually involve Amerindian ancestry shared with Europeans (this will basically track "Northeast Euro" percentages at HAP). On the other hand, it seems Southwest Asians-North Africans probably lack any substantial East Eurasian ancestry, but West Asians do have Amerindian admixture. The cool thing though, Mozabites are as West Eurasian as Russians. Also, it seems populations from the Northwestern fringe of South Asia, like Pashtuns, Baloch, Sindhis, and probably Punjabis, are as West Eurasian as many European populations, and actually more West Eurasian than some Northeast European populations.

Dr_McNinja
10-02-2013, 09:11 PM
I really find the Amerindian-related admixture among Europeans exceedingly cool. Based on this preprint from the Reich lab, it seems all Europeans have some really substantial East Eurasian admixture, 20% amongst the Basques and Sardinians, and 40% amongst Russians http://arxiv.org/abs/1212.2555. And Russians aren't even the most East Eurasian shifted of Europeans, so I think the share of East Eurasian ancestry in Europe might be even higher than that. The paper was somewhat ambiguous about whether this admixture is pan-West Eurasian, or just specific to European populations. So I asked David Reich. He is a very busy man, so I was exceedingly pleased that he took the time to answer my questions. All I can really say, he's an awesome man. His response:

"We don’t think it is present in some West Eurasian populations from the Middle East (e.g. Bedouins) but we do think it is probably present in West Eurasians from Pakistan as well as in people from the Caucasus and people from Europe."

This basically confirms your intuition. Also, this means a substantial amount of the "ASI" percentages found in northwestern South Asians, and South-Central Asians, actually involve Amerindian ancestry shared with Europeans (this will basically track "Northeast Euro" percentages at HAP). On the other hand, it seems Southwest Asians-North Africans probably lack any substantial East Eurasian ancestry, but West Asians do have Amerindian admixture. The cool thing though, Mozabites are as West Eurasian as Russians. Also, it seems populations from the Northwestern fringe of South Asia, like Pashtuns, Baloch, Sindhis, and probably Punjabis, are as West Eurasian as many European populations, and actually more West Eurasian than some Northeast European populations.That's fascinating. Now I just wonder what it means that my HAP numbers had both high Siberian and Beringian (total of 4.42% and if you count Amerindian, 6.01%). I think 1% of my NE-Euro went into Siberian in the HAP algorithm (9.51% on DIY-HAP to 8.34% on HAP and Siberian went up a percent). FTDNA's Population Finder said 9.13% Northern European with plus or minus 1.85%. I'm not sure if that's noise or the HAP algorithm becoming more proficient since our NE-Euro component should be pretty old (3000 years is my guess).|

EDIT: Also, my Dodecad DV3 numbers:


South_Asian 47.56
West_Asian 17.64
West_European 12.02
Mediterranean 6.52
East_European 6.21
Northeast_Asian 4.26
Southeast_Asian 3.16
Southwest_Asian 2.6
I wonder if the higher East Asian/East Eurasian is throwing some of these calculators for a loop (thus the Northeast Indian matches on DNA-Tribes). There's another Punjabi Jatt whose results had higher East European and lower Mediterranean so I wonder if that shift has any relevance. My West European was close to his, about a percent off. But my Southeast Asian was about 1.7% higher whereas Northeast Asian was only 1% higher. It could be what would otherwise be Northeastern European is showing up as East Asian?

There isn't much development of the South Asian J2b2* front at FTDNA but it's possible that since that has spread across northeast India (and the earliest Indo-Europeans and the Jatts descended from them were out that far) my ancestors originally came from there and picked up some East Asian along the way which somehow persisted in the West (which is only possible if a community moved together and remained endogamous... which isn't terribly likely but not impossible).

Sein
10-02-2013, 11:29 PM
All of this does make one wonder. Do you have any more information on your paternal haplogroup?

Dr_McNinja
10-03-2013, 02:28 AM
http://www.eupedia.com/europe/Haplogroup_J2_Y-DNA.shtml#J2b

Branches: http://www.yfull.com/tree/J2b2/ , http://i.imgur.com/w3djDRp.png , http://i.imgur.com/3hHJmxg.png (GIH = Gujarati in Houston, STU = Sri Lankan Tamil in the UK, PJL = Punjabi in Lahore, ITU = Indian Telegu in the UK, BEB = Bengali in Bangladesh)

Heatmap: http://farm9.staticflickr.com/8388/8675629245_0f99cf87c3_o.jpg

I just sent in my FTDNA sample today. I only ordered the YDNA-12 and will see from there. I'll have to order the M241 and L283 SNPs at some point. The former being J2b2 and the latter being the European sub-groups of J2b2. FTDNA doesn't offer testing of the South Asian J2b2 SNPs yet.

Here's a Punjabi in Lahore with J2b2: ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/data/HG02690/ I wonder if it's possible to get the raw data from this and run it in an admixture calculator

Post I made about it:


I think the Eupedia theory of J2b2 coming into India with R1a has a lot of merit now. Compare the heatmap from the first page to these:

http://en.wikipedia.org/wiki/File:IE_expansion.png

http://upload.wikimedia.org/wikipedia/commons/7/77/Chariot_spread.png

http://en.wikipedia.org/wiki/File:Indo-Iranian_origins.png

The J2b2 corresponds to the furthest reaches of the R1a migration/invasion from the North. The only question is why is it at the furthest reaches or fringes? Even that little dot in Swat. Perhaps J2b2 was in the earliest wave and thus propagated to these fringes whereas the R1a movement continued and filled in the rest of Northern and Central India?

As for when J2b2 got caught up with the R1a, the Kurgan hypothesis indicates that the Indo-Europeans as early as the 3rd Millennium BC went into Neolithic Europe and transformed that farming culture. Farming probably got to Neolithic Europe with the J2b2. Then the blank being filled in is that somehow some of these people from Old Europe came around through Central Asia in the latter half of the 2nd Millennium BC and wound up in northwest Punjab, Northern and Eastern India, and Sindh/Gujarat.

My ancestry is Jatt (i.e, indigenous peoples of northwestern India commonly reputed to be the most direct or earliest ancestors of the Central Asian people who came into India) from within India, further Southeast, and my genetic admixture shows higher South Indian and low Baloch compared to the other Northwestern populations (meaning either a late trip back up to Punjab and/or they were living in Punjab and not mixing too much with the others). The J2b2 could mean they were very old Jatts who had been in India for over a thousand years. Most of the J2b2 disseminated into the local ethnicities but some might have remained "Jat" and then moved back to Punjab.

This doesn't rule out the crazy Alexander theories because he happened to conquer precisely these areas and probably had J2b2 people with him which is a real headache.So based on the heatmap overlaid on the map of early Indo-European settlements in India and the spread of J2b2 throughout North and East/Southeast India among higher castes as well as near Sindh and a little bit in northern Pakistan (apparently the Burusho have a lot of J2b2) makes me think they hitched a ride with the earliest Indo-Europeans in India and the later waves were R1a and dominated the area inbetween (the bulk of northwest India).

Dr_McNinja
10-06-2013, 12:07 AM
Sein, have you done the Eurogenes K36 admixture calculator? I'd be interested to see how the breakdown in European admixture proportion looks for a Pashtun and how it differs from Punjabis if it's not too much trouble. Here was mine: http://i.imgur.com/bUFxiNE.png

Sein
10-06-2013, 02:46 AM
Sure thing. Also, you're results are very interesting:

South_Central_Asian=41.98%
South_Asian=21.12%
North_Caucasian=14.40%
Armenian=6.56%
Eastern_Euro=4.49%
East_Med=3.87%
Fennoscandian=1.85%
Siberian=1.76%
Amerindian=1.14%
West_Caucasian=0.89%
North_Sea=0.63%
Volga-Ural=0.58%
Oceanian=0.52%
East_African=0.22%

Sapporo
10-06-2013, 04:43 AM
Here are my results for comparison as well.

Population:
South_Central_Asian = 44.43%
South_Asian = 30.25%
North_Caucasian = 12.74%
Eastern_Euro = 4.11%
North_Sea = 2.51%
East_Central_Euro = 1.97%
Fennoscandian = 1.07%
Armenian = 0.97%
Volga-Ural = 0.92%
East_Balkan = 0.50%
Amerindian = 0.38%
West_Caucasian = 0.13%

According to this test, my Siberian/Amerindian like ancestry is within noise level of 0% while other tests have it at around 2-3%. My Northern Euro component is partially being captured as North Caucasian plus other Euro components and my Gedrosia/Baloch as South Central Asian.

Sapporo
10-06-2013, 04:51 AM
Sure thing. Also, you're results are very interesting:

South_Central_Asian=41.98%
South_Asian=21.12%
North_Caucasian=14.40%
Armenian=6.56%
Eastern_Euro=4.49%
East_Med=3.87%
Fennoscandian=1.85%
Siberian=1.76%
Amerindian=1.14%
West_Caucasian=0.89%
North_Sea=0.63%
Volga-Ural=0.58%
Oceanian=0.52%
East_African=0.22%

Not surprising that you have substantial Caucasus like ancestry that we are lacking in. Especially, the South Caucasus Armenian component.

Sein
10-06-2013, 05:57 AM
Hey Sapporo,

Very interesting results. I think it's fascinating that you are somewhat intermediate between me and Dr_McNinja.

Sapporo
10-06-2013, 06:08 AM
Hey Sapporo,

Very interesting results. I think it's fascinating that you are somewhat intermediate between me and Dr_McNinja.

Yeah. Not sure what could really explain that considering we are of the same ethnicity and even share a clan name I believe.

Dr_McNinja
10-06-2013, 02:59 PM
I think the haplogroups might be the key here. J2b2* probably was one of the earliest waves of Indo-Europeans into the subcontinent and one of the earliest groups of Jats. One of the explanations for my South Asian-heavy admixture is that my ancestors were from somewhere in North India further to the east (J2b2* spread is heavier there and in Gujarat, and R1a filled in the area behind its spread as well as in the same areas).

So my thinking is the earlier Indo-Europeans who moved into India, their cousins were also the earliest ones in Europe and they wound up being the Germanic/Nordic group that had the time to move the furthest West in Europe. I keep getting Orcadian/Orkney Islands in some admixture Oracle results and FTDNA's Population Finder picked it up as well. I assume that's showing up here as North Sea (which is taken from populations in Kent, whereas North Atlantic is from populations in Cornwall and Southwestern Scots... the population in Kent is supposed to be more Germanic in heritage according to discussions on it I've found at the Eurogenes blog and other such forums). Here's a map of it: http://www.worldatlas.com/aatlas/infopage/northsea.gif Kent would be the southeastern tip of England.


North sea is an area in the very South East.. It represents the Germanic England, with a shared ancestry to it's East in Norway, Netherlands & Denmark..

North Atlantic is Cornwall, with a shared ancestry along the North Atlantic. It represents Pre-Germanic Britain, with a shared ancestry in Ireland & Western Britain.

But looking at the R1a heatmap:

http://upload.wikimedia.org/wikipedia/commons/9/94/R1a1a_distribution.png

These would be the general areas I'd expect for cousins of Indo-European descendants in India who came with one of the later waves, like perhaps the Scythians or the Hephthalites. More Northeastern European.

Some of my Caucasian percentage is missing and picked up as Gedrosian and some European I think (European overall should be almost 1% lower and Caucasian should be 2% higher to match the Harappa admixture results). Or perhaps this is just more accurate than Harappa since FTDNA's population finder did find higher margin of error in my Caucasian results. I think those are the most volatile for everyone. Not sure where my extra 5-7% South Indian is coming from.

Here are the Eurogenes K36 reference populations, btw: https://docs.google.com/file/d/0B9o3EYTdM8lQUlVTRmxQdGpZdW8/edit?pli=1

Another interesting thing is that the high Siberian/Beringian/Amerindian I'm getting in other calculators is reduced to less than 2% total here. I do believe those components in Indians also indicate a more ancient Northern European relative (and in Pashtun the Siberian probably indicates the opposite, since if they are descended from Hephthalites they were in Central Asia until more recently and in position to pick up more of it before they came south... Kazakhs have really high Siberian).

Dr_McNinja
10-06-2013, 04:23 PM
Here are my Eurogenes K13 and K12b results:

K13:


South_Asian 45.22
West_Central_Asian 25.27
Caucasus 8.93
North_European 12.65
Mediterranean 1.59
North_Eurasian 1.66
Southwest_Asian 0.7
East_Asian 0
East_Siberian 2.39
Amerindian 1.02
Northeast_African 0
Pygmy 0
West African 0.56

K12b:


South_Asian 43.74
West_Central_Asian 27.17
Caucasus 10.5
West_European 5.55
North_European 7.63
Finnish 0.92
Mediterranean 0.11
Siberian 3.69
Southwest_Asian 0.24
East_Asian 0
West_African 0.44
East_African 0

Reference populations:


K12b
Western European - Cornwall
Siberian - Central Siberia
East African - Masaai
West Central Asian - Balochistan
South Asian - South India
West African - Nigeria
Caucasus - Georgia
Finnish - Eastern Finland
Mediterranean - Sardinia
Southwest Asian - Bedouin
North European - Lithuania
East Asian - Eastern China


K13
North European - Lithuania
West African - Nigeria
Mediterranean - Sardinia
Northeast African - Ethiopia Oromo
North Eurasian - Central Siberia
South Asian - South India
Southwest Asian - Bedouin
Pygmy - Mbuti Pygmy
Caucasus - Georgia
East Siberian - Koryaks
East Asian - Eastern China
Amerindian - South America
West Central Asian - BalochistanAnd from K36, North Sea = Kent, North Atlantic = Cornwall, East Central Euro = Lithuanian and Ukrainian, East Euro = Kargopol Russian and Mokshas and Russian, and Central Euro = Hungarian.

Interesting how my results are Lithuanian-heavy in K13 and K12b but a lot less so in K36. Western Europe (first Cornwall, then more specifically North Sea) stayed the same, around 5-6%. It looks like the missing European/Siberian/etc went into Gedrosian which jumped almost 10% from K12b/K13 to K36.

EDIT: My overall European totals went from 14.1% (West+North+Finnish) in K12b to 12.65% in K13 with 1.66% of North Eurasian (a type of Siberian) coming out. And then down to 10.37% in K36, within the error of margin of FTDNA and DIYHarappa.

It seems the lower resolution Eurogenes calculators were capturing some of the Gedrosian/Baloch as European.

South_Central_Asian in K36 includes not just Balochistan but Brahui, Burusho, Pathan and Sindhi from HGDP (so basically Pakistan). I'm not sure then if K36 is more accurate because these might be "stealing" some of the extra European or Caucasus numbers from the K12b/K13 calculators. These aren't indigenous groups except perhaps the Burusho and Brahui. So my missing Siberian might be in the Pathan or Burusho or something being captured as South_Central_Asian.

But it doesn't affect a comparison between us since we're all from the same area (and K36 shows a west-east gradient possibly correlating to age of ancestral migration).

My Siberian/Amerindian went from ~2% in K36 to 3.69% in K12b and 5.07% in K13.

It is kind of weird that less is captured by a sole Baloch component in the Eurogenes calculators than with the Harappa calculator. Eurogenes had to include all of HGDP Pakistan to get the same as just Balochistan in Harappa. I guess Zack's selection was pretty good? Probably the most accurate, except for the lack of a Germanic European component.

EDIT #2: Gedrosian in Europe from Eupedia: http://www.eupedia.com/images/content/Gedrosian-admixture.gif and more numbers: https://docs.google.com/spreadsheet/ccc?key=0ArJDEoCgzRKedGdRbkxKMDdlZkJWc21tdkpldWxwV %20mc#gid=0

It looks like Gedrosian is highest in Western Europe (Orcadian/Orkney and Cornwall/Kent). Eastern Europe has almost no Gedrosian. So I'm not sure if Western European is being captured as Gedrosian or vice-versa, but when the calculator includes it, it definitely shows up in my results and the other Eastern European numbers drop. I wonder if Zack added a Western Europe component to represent Germanic influence, if my Baloch component would drop even lower than it already is, and so would my NE-Euro...

EDIT #3: Looking over the Harappa spreadsheet again, I think my Mediterranean is disappearing into Eurogenes' Western European. Since there's no Germanic European component in Harappa, the Western Europeans only have about half NE-Euro and then have over 30% Mediterranean and a good chunk of Baloch. I've been getting slightly higher Mediterranean in Dodecad/Globe13/etc but I have none whatsoever in Eurogenes K36 and almost 0% in K12b. Sapporo's Mediterranean also disappeared. But Sein's hasn't so it probably represents something legit there. I'm really curious now what would happen to my Baloch if a Western European component was included in Harappa.

Dr_McNinja
10-06-2013, 05:52 PM
I emailed Zack asking if in his upcoming post looking at Punjabis, if he could use a Western Euro/Germanic component too. I have a feeling some of the NE-Euro will branch off into that, and all the Mediterranean for Jatts will go into that, and maybe some of the Baloch too. If I wind up with even less than 33% Baloch, that would really pin down our origins somewhere well to the East of Punjab.

Also would be curious if the Siberian/Beringian/Amerindian would persist or disappear into something else.

Also a North/West Caucasian split too. Not sure why I have West Caucasian and you guys barely have any, even though your North Caucasian is so high. (Just checked, Harappa's Caucasus appears to be West Caucasus, which is Abhkasian and Georgian, which might be why I'm at numbers close to Sapporo there (both around 11%)... North Caucasians like Adygei are only registering ~50% Caucasian on Harappa... Sapporo might have substantially higher Caucasian if both North/West are picked out)

And Sapporo, I think some of your Siberian went into Volga-Ural.

EDIT: And also there's a lack of Rajput representation in Harappa. I think some Rajputs can be traced to ancestral Kshatriya caste Indians (like a few of the ones in Harappa now), or to Jatts who changed social stratus, so they would have results similar to Brahmins but the Indo-Hephthalite descendant Rajputs should have high Caucasian and are showing up in some Jatts, since many Jatts have mixed Rajput/Jatt ancestry in Punjab. That could explain the influx of Caucasian into an otherwise severely endogamous group (why some Haryana Jatts have very low and others have substantial Caucasian). Might also explain Sapporo's high Caucasian if a Harappa run with a North Caucasian component is done. Among Muslims it could also represent mixture with other Indo-Iranian groups who have repeatedly moved into the area, not just Hephthalites (possible ancestors of Pathans).

EDIT #2: Also in DNA-Tribes my Western Europe result is more specific to the Netherlands rather than England or the Orkney Islands. Not sure how accurate they are though. I don't see any Dutch/Netherlands in Harappa but they are right behind the British (Cornwall/Kent/Orcadians) in Gedrosian %.

EDIT #3: FTDNA groups European like this:


Western European
Basque
French
Orcadian (Orkney Islands)
Spanish

Northeast European
Finnish
Russian

Southeast European
Romanian

Southern European
Italian
Sardinian
TuscanAnd in DNA-Tribes, Basque and 'Greek Mixed' are higher up on my list, but were not picked up by FTDNA's Population Finder which only registered Orcadian.

Sein
10-06-2013, 08:49 PM
My results for K12b, and K13:

K12b
West Central Asian=31.42%
South Asian=26.89%
Caucasus=18.23%
North European=9.52%
Siberian=4.22%
Finnish=3.12%
Southwest Asian=3.09%
Western European=2.30%
East Asian=0.80%
West African=0.40%

K13
South Asian=30.28%
West Central Asian=28.37%
Caucasus=15.64%
North European=13.98%
Southwest Asian=3.04%
East Siberian=2.74%
Amerindian=1.93%
Mediterranean=1.33%
North Eurasian=1.53%
East Asian=0.52%
West African=0.32%
Pygmy=0.31%

Dr_McNinja
10-06-2013, 10:59 PM
^ There's a surprising difference between your South Asian and West Central Asian results from K12b to K13. And your Mediterranean also turned into Western European.

Sein
10-07-2013, 04:58 AM
It does seem that they are rather different calculators. Have you tried the Eurogenes Jtest-EU test? Not really meant for non-Europeans, but for what it's worth, my results (the EU test is nearly identical):

Jtest
SOUTH_ASIAN=38.17%
WEST_ASIAN=28.21%
EAST_EURO=15.2%
EAST_MED=10.7%
SIBERIAN=3.22%
NORTH-CENTRAL_EURO=2.08%
SOUTH_BALTIC=1.48%
WEST_AFRICAN=0.39%
EAST_ASIAN=0.3%
EAST_AFRICAN=0.22%
ASHKENAZI=0.02%

1 Kalash 8.56
2 Burusho 8.88
3 Balochi 10.51
4 Brahui 11.32
5 Sindhi 13.17
6 IN 27.16
7 Gujarati 30.48
8 Bangladeshi 32.06
9 IR 33.91
10 Lezgin 34.88
11 Kurdish 36.47
12 TR 39.98
13 GE 43.26
14 Armenian 43.96
15 IQ 44.68
16 Mandean 44.76
17 Assyrian 45.41
18 RO 47.58
19 Serbian 47.9
20 GR 48.15

Using 2 populations approximation:
1 50% Balochi +50% Burusho @ 4.676
2 50% Brahui +50% Burusho @ 4.985
3 50% Balochi +50% Kalash @ 6.561
4 50% Brahui +50% Kalash @ 6.733
5 50% Gujarati +50% Lezgin @ 6.901
6 50% Balochi +50% Sindhi @ 7.016
7 50% Brahui +50% Sindhi @ 7.187
8 50% Burusho +50% Kalash @ 7.209
9 50% Kalash +50% Sindhi @ 7.688
10 50% IN +50% Lezgin @ 7.927
3081 iterations.

Also, if it's not too much trouble, the Eurogenes Hunter_Gatherer vs. Farmer test is very interesting, if you'd like to compare. Although, the "South Asian Hunter Gatherer" component is really a hybrid cluster, and is significantly much more West Eurasian than "ASI". I'd say about 70% West Eurasian.

Anatolian Farmer=39.65%
South Asian Hunter Gatherer=39.51%
Baltic Hunter Gatherer=11.83%
North Eurasian Hunter Gatherer=3.38%
Middle Eastern Herder=2.58%
South American Hunter Gatherer=2.02%
Oceanian Hunter Gatherer=0.38%
Bantu Farmer=0.36%
Pygmy Hunter Gatherer=0.18%
East Asian Farmer=0.11%

The Eurogenes calculators are the only ones on which I score some Sub-Saharan African admixture. McDonald, 23andMe, Geno 2.0, the Dodecad calculators, and HarappaWorld don't really detect any. I think this means that if the parameters are European variation, I am somewhat African shifted in comparison to Europeans.

Dr_McNinja
10-07-2013, 05:12 PM
I see a test there I hadn't before, 'EUTest V2 K15'. I think he just released this test today. My results from that:


Population
North_Sea 4.37%
Atlantic 4.09%
Baltic 2.95%
Eastern_Euro 7.15%
West_Med -
West_Asian 21.61%
East_Med 1.96%
Red_Sea 0.14%
South_Asian 53.26%
Southeast_Asian -
Siberian 1.27%
Amerindian 1.83%
Oceanian 0.46%
Northeast_African -
Sub-Saharan 0.91%^ The Oracle matches were useless but Uttar Pradesh Brahmin was my highest match @ 8.019.

My Hunter Gatherer vs. Farmer results:


Population
Anatolian Farmer 26.97%
Baltic Hunter Gatherer 10.09%
Middle Eastern Herder 0.13%
East Asian Farmer -
South American Hunter Gatherer 1.13%
South Asian Hunter Gatherer 56.54%
North Eurasian Hunter Gatherer 2.92%
East African Pastoralist -
Oceanian Hunter Gatherer 0.30%
Mediterranean Farmer 1.33%
Pygmy Hunter Gatherer -
Bantu Farmer 0.58%

My Jtest results:


Population
SOUTH_BALTIC 1.23%
EAST_EURO 10.11%
NORTH-CENTRAL_EURO 3.94%
ATLANTIC 3.28%
WEST_MED -
ASHKENAZI -
EAST_MED 4.91%
WEST_ASIAN 18.39%
MIDDLE_EASTERN -
SOUTH_ASIAN 55.41%
EAST_AFRICAN -
EAST_ASIAN -
SIBERIAN 2.69%
WEST_AFRICAN -

And my results with the old EUTest:


Population
SOUTH_BALTIC 1.23%
EAST_EURO 10.11%
NORTH-CENTRAL_EURO 3.94%
ATLANTIC 3.28%
WEST_MED -
EAST_MED 4.91%
WEST_ASIAN 18.39%
MIDDLE_EASTERN -
SOUTH_ASIAN 55.41%
EAST_AFRICAN -
EAST_ASIAN -
SIBERIAN 2.69%
WEST_AFRICAN -

Least-squares method.

Using 1 population approximation:
1 IN @ 8.975
2 Sindhi @ 10.068
3 Gujarati @ 12.608
4 Bangladeshi @ 15.511
5 Burusho @ 18.234
6 Kalash @ 25.453
7 Balochi @ 30.381
8 Brahui @ 31.110
9 South_Indian @ 34.681
10 IR @ 55.413

Using 4 populations approximation:
1 Gujarati + Gujarati + Lezgin + South_Indian @ 3.958
2 Gujarati + IN + Lezgin + South_Indian @ 4.127
3 IN + IN + Lezgin + South_Indian @ 4.545

It's interesting that the Caucasian I'd normally get is being picked up under South Asian and not West Asian.

The new EUTest V2 K15 might be the most accurate yet for European admixture. It has Atlantic (Cornwall/Southwestern England I assume), North Sea, Baltic (Lithuania?), and Eastern Euro all together in one test. In the old EUTest my Mediterranean numbers jump up again without a North Sea component. Not sure if it's better than K36, but K36 might have issues with all the components. On the other hand, K36 has about the right amount of total European admixture while this one almost doubles it. So...

Scarlet Ibis
10-07-2013, 07:28 PM
I see a test there I hadn't before, 'EUTest V2 K15'. I think he just released this test today. My results from that:


Population
North_Sea 4.37%
Atlantic 4.09%
Baltic 2.95%
Eastern_Euro 7.15%
West_Med -
West_Asian 21.61%
East_Med 1.96%
Red_Sea 0.14%
South_Asian 53.26%
Southeast_Asian -
Siberian 1.27%
Amerindian 1.83%
Oceanian 0.46%
Northeast_African -
Sub-Saharan 0.91%^ The Oracle matches were useless but Uttar Pradesh Brahmin was my highest match @ 8.019.

My Hunter Gatherer vs. Farmer results:


Population
Anatolian Farmer 26.97%
Baltic Hunter Gatherer 10.09%
Middle Eastern Herder 0.13%
East Asian Farmer -
South American Hunter Gatherer 1.13%
South Asian Hunter Gatherer 56.54%
North Eurasian Hunter Gatherer 2.92%
East African Pastoralist -
Oceanian Hunter Gatherer 0.30%
Mediterranean Farmer 1.33%
Pygmy Hunter Gatherer -
Bantu Farmer 0.58%

My Jtest results:


Population
SOUTH_BALTIC 1.23%
EAST_EURO 10.11%
NORTH-CENTRAL_EURO 3.94%
ATLANTIC 3.28%
WEST_MED -
ASHKENAZI -
EAST_MED 4.91%
WEST_ASIAN 18.39%
MIDDLE_EASTERN -
SOUTH_ASIAN 55.41%
EAST_AFRICAN -
EAST_ASIAN -
SIBERIAN 2.69%
WEST_AFRICAN -

And my results with the old EUTest:


Population
SOUTH_BALTIC 1.23%
EAST_EURO 10.11%
NORTH-CENTRAL_EURO 3.94%
ATLANTIC 3.28%
WEST_MED -
EAST_MED 4.91%
WEST_ASIAN 18.39%
MIDDLE_EASTERN -
SOUTH_ASIAN 55.41%
EAST_AFRICAN -
EAST_ASIAN -
SIBERIAN 2.69%
WEST_AFRICAN -

Least-squares method.

Using 1 population approximation:
1 IN @ 8.975
2 Sindhi @ 10.068
3 Gujarati @ 12.608
4 Bangladeshi @ 15.511
5 Burusho @ 18.234
6 Kalash @ 25.453
7 Balochi @ 30.381
8 Brahui @ 31.110
9 South_Indian @ 34.681
10 IR @ 55.413

Using 4 populations approximation:
1 Gujarati + Gujarati + Lezgin + South_Indian @ 3.958
2 Gujarati + IN + Lezgin + South_Indian @ 4.127
3 IN + IN + Lezgin + South_Indian @ 4.545

It's interesting that the Caucasian I'd normally get is being picked up under South Asian and not West Asian.

The new EUTest V2 K15 might be the most accurate yet for European admixture. It has Atlantic (Cornwall/Southwestern England I assume), North Sea, Baltic (Lithuania?), and Eastern Euro all together in one test. In the old EUTest my Mediterranean numbers jump up again without a North Sea component. Not sure if it's better than K36, but K36 might have issues with all the components. On the other hand, K36 has about the right amount of total European admixture while this one almost doubles it. So...

Excellent. Thanks for the heads up about it.

Sein
10-08-2013, 02:13 AM
Very interesting results my friend. For the new Eurogenes EUtestV2 K15 calculator, my results:

South_Asian=36.90%
West_Asian=30.09%
Eastern_Euro=12.50%
East_Med=7.07%
North_Sea=6.84%
Siberian=1.95%
Oceanian=0.76%
Sub-Saharan=0.47%
Southeast_Asian=0.15%

Single Population Sharing
1. Pathan 5.84
2. Burusho 9.5
3. Balochi 10.16
4. Brahui 10.59
5. Punjabi_Jat 10.91
6. Sindhi 11.62
7. Makrani 14.05
8. Tadjik 18.63
9. Brahmin_UP 24.09
10. Gujarati 24.87
11. Kshatriya 26.33
12. Bangladeshi 29.45
13. Lezgin 32.38
14. Iranian 32.68
15. Kumyk 32.8
16. Chechen 34.03
17. Dharkar 34.28
18. Kurdish 34.65
19. Nogay 35.21
20. Kanjar 35.48

Mixed Mode Population Sharing:
1. 87.9% Pathan + 12.1% Lezgin @ 3.83
2. 88.6% Pathan + 11.4% Kumyk @ 4.07
3. 89% Pathan + 11% Chechen @ 4.08
4. 90.8% Pathan + 9.2% Georgian @ 4.19
5. 91.1% Pathan + 8.9% Abhkasian @ 4.22
6. 89.8% Pathan + 10.2% Balkar @ 4.23
7. 89.9% Pathan + 10.1% Adygei @ 4.24
8. 90.7% Pathan + 9.3% North_Ossetian @ 4.38
9. 82.5% Pathan + 17.5% Tadjik @ 4.41
10. 70.7% Pathan + 29.3% Balochi @ 4.44
11. 91.7% Pathan + 8.3% Armenian @ 4.51
12. 72.4% Pathan + 27.6% Brahui @ 4.56
13. 90% Pathan + 10% Iranian @ 4.6
14. 90.5% Pathan + 9.5% Kurdish @ 4.61
15. 90.8% Pathan + 9.2% Nogay @ 4.64
16. 91.5% Pathan + 8.5% Turkish @ 4.72
17. 79.5% Pathan + 20.5% Makrani @ 4.74
18. 92.8% Pathan + 7.2% Assyrian @ 4.79
19. 52.3% Balochi + 47.7% Punjabi_Jat @ 4.91
20. 75.2% Sindhi + 24.8% Lezgin @ 4.92

Sapporo
10-08-2013, 04:54 AM
Posting more results for comparison purposes:

Eurogenes:

K12b

South Asian = 34.86%
West Central Asian = 34.04%
North European = 12.86%
Caucasus = 11.94%
Siberian = 2.20%
Western European = 1.95%
Finnish = 1.82%
Southwest Asian= 0.33%

K13

South Asian = 38.72%
West Central Asian = 29.23%
North European= 16.72%
Caucasus = 10.66%
North Eurasian = 1.60%
Southwest Asian = 1.21%
East Siberian = 1.16%
Amerindian = 0.70%

Jtest

SOUTH_ASIAN = 48.06%
WEST_ASIAN = 24.68%
EAST_EURO = 17.54%
EAST_MED = 4.57%
SOUTH_BALTIC = 2.21%
NORTH-CENTRAL_EURO = 1.69%
ATLANTIC = 0.95%
SIBERIAN = 0.30%

Single Population Sharing:

# Population (source) Distance
1 Sindhi 6.91
2 Burusho 11.39
3 Kalash 13.37
4 IN 17.56
5 Gujarati 20.44
6 Balochi 20.48
7 Brahui 21.36
8 Bangladeshi 23.7
9 South_Indian 42.14
10 IR 44.71
11 Lezgin 45.33
12 Kurdish 47.19
13 TR 50.68
14 IQ 53.8
15 GE 54.09
16 Armenian 54.51
17 RO 54.66
18 Serbian 54.73
19 Mandean 54.77
20 Udmurt 55.12

Eurogenes Hunter_Gatherer vs. Farmer Admixture Proportions

South Asian Hunter Gatherer = 49.65%
Anatolian Farmer = 33.88%
Baltic Hunter Gatherer = 14.06%
North Eurasian Hunter Gatherer = 1.73%
South American Hunter Gatherer = 0.69%


Also, Sein calculated my Spatial Analysis for me. I clustered in Western Syria.

Sapporo
10-08-2013, 05:28 AM
Eurogenes EUtest V2 K15:

Admix Results (sorted):

# Population Percent
1 South_Asian 45.72
2 West_Asian 25.28
3 Eastern_Euro 14.57
4 East_Med 4.76
5 North_Sea 4.17
6 Baltic 3.57
7 Amerindian 1.59
8 Siberian 0.33

Single Population Sharing:

# Population (source) Distance
1 Punjabi_Jat 4.88
2 Pathan 5.59
3 Sindhi 6.77
4 Burusho 10.38
5 Brahmin_UP 15.58
6 Gujarati 16.3
7 Kshatriya 17.62
8 Balochi 17.64
9 Brahui 17.88
10 Makrani 21.19
11 Bangladeshi 21.93
12 Dharkar 25.84
13 Tadjik 26.54
14 Kanjar 26.7
15 Velamas 29.52
16 Uttar_Pradesh 29.7
17 Kurumba 30.58
18 Kol 30.87
19 North_Kannadi 31.99
20 Dusadh 32

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 80.3% Pathan + 19.7% Kshatriya @ 3.66
2 79% Pathan + 21% Gujarati @ 3.66
3 86.3% Pathan + 13.7% Kanjar @ 3.68
4 78.4% Pathan + 21.6% Brahmin_UP @ 3.72
5 88.5% Pathan + 11.5% North_Kannadi @ 3.75
6 88.3% Pathan + 11.7% Kol @ 3.84
7 87.9% Pathan + 12.1% Uttar_Pradesh @ 3.86
8 90.2% Pathan + 9.8% Chamar @ 3.88
9 89.7% Pathan + 10.3% Piramalai @ 3.96
10 86.6% Pathan + 13.4% Dharkar @ 3.96
11 89% Pathan + 11% Dusadh @ 3.99
12 88.9% Pathan + 11.1% Kurumba @ 4.11
13 59.2% Punjabi_Jat + 40.8% Pathan @ 4.13
14 88.7% Pathan + 11.3% Velamas @ 4.15
15 89.9% Pathan + 10.1% Chenchu @ 4.2
16 85.8% Pathan + 14.2% Bangladeshi @ 4.31
17 71.9% Punjabi_Jat + 28.1% Sindhi @ 4.45
18 82.8% Punjabi_Jat + 17.2% Burusho @ 4.47
19 91.2% Punjabi_Jat + 8.8% Brahui @ 4.58
20 91.4% Punjabi_Jat + 8.6% Balochi @ 4.6

Sein, it's within margin of error, but I'm slightly closer than you to the HGDP Pakistani Pashtuns than yourself on this Oracle. I think it's mostly down to me having more Gedrosia and South Asian ancestry with lower Siberian/East Eurasian ancestry. Also, most oracles (including those with Punjabi Jatts) seem to think my admixture is closer to a South Asian shifted HGDP Pashtun rather than a Northwestern South Asian with affinity to the HGDP Pashtuns. Also, I'm curious which Houston Gujarati group is on this oracle. Finally, your oracle indicates you are like one of the HGDP Pakistani Pashtun with a shift to North Caucasian groups.

Sein
10-08-2013, 05:40 AM
Now those are some really interesting results! To be honest, I think you do fit within Pashtun variation. And based on what we see in open genome projects, I think Punjabi Jatts are the closest contemporary population to Pashtuns. I'm willing to bet that if we had Tajik autosomal data from Afghanistan, we'd find that Pashtuns are almost perfectly equidistant (and very close) to both Afghan Tajiks and Punjabi Jatts.

Sapporo
10-08-2013, 05:47 AM
Now those are some really interesting results! To be honest, I think you do fit within Pashtun variation. And based on what we see in open genome projects, I think Punjabi Jatts are the closest contemporary population to Pashtuns. I'm willing to bet that if we had Tajik autosomal data from Afghanistan, we'd find that Pashtuns are almost perfectly equidistant (and very close) to both Afghan Tajiks and Punjabi Jatts.


Well, I'm not too sure about that. I think we determined the two most South Asian shifted HGDP Pashtun samples were outliers since they were even ever so slightly more South Asian shifted than some of the Punjabi groups. Then there was 1-2 HGDP Pashtuns with numbers similar to the Punjabi Jatts. However, the rest of the HGDP Pashtuns seemed related in terms of genetic distance but clearly distinct and clearly much closer to you and HRP0281. Anyways, I agree about your notions on Afghan Tajiks. They are mostly likely more South Asian and Southwest Asian shifted and less North Euro and East Asian shifted than Tajikistan Tajiks which would put them closer to Afghan Pashtuns than their Tajikistan counterparts. So, most Pashtun variation should lie in between Afghan Tajiks and some Northwestern South Asian groups with some outliers who are more Iranian shfited and South Asian shifted. Then, there are likely extreme outliers like the single HGDP Pashtun who seems to have significant East Eurasian ancestry and maybe even an extreme outlier with more Northern Euro like ancestry due to the Indo-Iranian tribes like Scythians and Hephthalites?

Sein
10-08-2013, 06:49 AM
It is very interesting that you bring this up, since many anthropologists and historians have claimed that Pashtuns are descendants of the Hephthalites (and as you well know, Jatts are always associated with Scythians). It is thought this would explain the organization of Pashtun tribes into confederacies that have no geographic coherency, but are rather united by supposedly shared patrilineal genealogy. Also, the Hephthalites spoke an East Iranian language, just like Pashto.

Dr_McNinja
10-09-2013, 01:31 AM
An Indo-Hephthalite origin of Rajputs would explain the genetic proximity (including the influx of recent Caucasian admixture in a seemingly random pattern in Punjab, the only place it doesn't correlate to religion). The Rajputs mixed into the Jatts thoroughly so many Jatts today would have actual Hephthalite ancestry, it would make sense they'd be the closest match to the Pashtun. Even my own closest match in Northwestern India on DNA-Tribes was Pashtun... I was as close to Lithuanians as I was to Arain Punjabi even though they live in the same area.

Also, Sindh was ruled by a Rajput dynasty (usurped by a Brahmin shortly before the Arabs invaded) not long after the Hephthalite invasion of India (the Rai Dynasty, established in 489 CE). But before the Hephthalites, it was Shakas/Sakas (a.k.a. Indo-Scythians) who ruled it. When the Arabs got there, they said the inhabitants were Jats, Rajputs, Bhattis, Medes, etc.

Dr_McNinja
10-09-2013, 02:35 AM
We need to find some Brahmins and Punjabi Rajputs to compare with...

Dr_McNinja
10-09-2013, 02:41 AM
My Oracle results from EUTest V2 K15:


Single Population Sharing:

# Population (source) Distance
1 Brahmin_UP 7.09
2 Gujarati 8.1
3 Sindhi 8.66
4 Kshatriya 9.02
5 Punjabi_Jat 9.17
6 Pathan 13.37
7 Bangladeshi 14.81
8 Burusho 16.87
9 Dharkar 16.9
10 Kanjar 18.16
11 Velamas 20.82
12 Uttar_Pradesh 21.04
13 Kurumba 21.73
14 Kol 22.22
15 Dusadh 23.35
16 North_Kannadi 23.55
17 Chenchu 24.48
18 Brahui 24.96
19 Balochi 25.06
20 Piramalai 26.05

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 85.1% Kshatriya + 14.9% Chechen @ 2.08
2 85.9% Kshatriya + 14.1% North_Ossetian @ 2.15
3 85.6% Kshatriya + 14.4% Adygei @ 2.36
4 85.5% Kshatriya + 14.5% Balkar @ 2.43
5 89.6% Brahmin_UP + 10.4% Abhkasian @ 2.55
6 85% Kshatriya + 15% Lezgin @ 2.62
7 55.6% Brahmin_UP + 44.4% Sindhi @ 2.66
8 73.6% Kanjar + 26.4% Chechen @ 2.72
9 85% Kshatriya + 15% Kumyk @ 2.72
10 86.7% Kshatriya + 13.3% Georgian @ 2.77
11 86.9% Kshatriya + 13.1% Abhkasian @ 2.81
12 89.6% Brahmin_UP + 10.4% Georgian @ 2.85
13 89.1% Brahmin_UP + 10.9% North_Ossetian @ 2.9
14 87.5% Gujarati + 12.5% North_Ossetian @ 2.92
15 86.9% Gujarati + 13.1% Chechen @ 3.01
16 87.2% Gujarati + 12.8% Balkar @ 3.03
17 87.4% Gujarati + 12.6% Adygei @ 3.07
18 89.1% Brahmin_UP + 10.9% Adygei @ 3.13
19 60.6% Kshatriya + 39.4% Pathan @ 3.14
20 67.2% Brahmin_UP + 32.8% Pathan @ 3.21



Least-squares method.

Using 1 population approximation:
1 Brahmin_UP @ 8.019
2 Gujarati @ 8.797
3 Kshatriya @ 9.988
4 Sindhi @ 10.078
5 Punjabi_Jat @ 10.525
6 Pathan @ 15.352
7 Bangladeshi @ 15.736
8 Burusho @ 18.878
9 Dharkar @ 18.884
10 Kanjar @ 20.251
165 iterations.



Using 2 populations approximation:
1 50% Brahmin_UP +50% Sindhi @ 3.208
2 50% Kshatriya +50% Punjabi_Jat @ 3.660
3 50% Kshatriya +50% Sindhi @ 3.966
4 50% Gujarati +50% Punjabi_Jat @ 4.226
5 50% Dharkar +50% Pathan @ 4.315
6 50% Gujarati +50% Sindhi @ 4.393
7 50% Kshatriya +50% Pathan @ 4.565
8 50% Brahmin_UP +50% Punjabi_Jat @ 4.605
9 50% Kanjar +50% Pathan @ 4.856
10 50% Gujarati +50% Pathan @ 5.325
13695 iterations.



Using 3 populations approximation:
1 50% Kshatriya +25% Chamar +25% Chechen @ 3.001
2 50% Kanjar +25% Chechen +25% Kanjar @ 3.019
3 50% Kanjar +25% Chechen +25% Dharkar @ 3.045
4 50% Kanjar +25% Chechen +25% Kshatriya @ 3.138
5 50% Kshatriya +25% Punjabi_Jat +25% Sindhi @ 3.205
6 50% Brahmin_UP +25% Sindhi +25% Sindhi @ 3.208
7 50% Sindhi +25% Brahmin_UP +25% Brahmin_UP @ 3.208
8 50% Kol +25% Chechen +25% Kshatriya @ 3.239
9 50% Dharkar +25% Chechen +25% Kanjar @ 3.254
10 50% Kol +25% Brahmin_UP +25% Chechen @ 3.277
210790 iterations.



Using 4 populations approximation:
1 Brahmin_UP + Chechen + Kanjar + Piramalai @ 2.745
2 Chechen + Kanjar + Kol + Kshatriya @ 2.842
3 Brahmin_UP + Chamar + Chechen + Kanjar @ 2.963
4 Chechen + Gujarati + Kanjar + Kol @ 2.983
5 Chechen + Gujarati + Kanjar + North_Kannadi @ 2.986
6 Chamar + Chechen + Gujarati + Kanjar @ 2.988
7 Chechen + Kanjar + Kshatriya + North_Kannadi @ 2.990
8 Chechen + Kanjar + Kshatriya + Piramalai @ 2.996
9 Chamar + Chechen + Kshatriya + Kshatriya @ 3.001
10 Chechen + Kanjar + Kanjar + Kanjar @ 3.019
11 Chechen + Gujarati + Kanjar + Piramalai @ 3.023
12 Chechen + Dharkar + Kanjar + Kanjar @ 3.045
13 Brahmin_UP + Chechen + Dharkar + Piramalai @ 3.064
14 Brahmin_UP + Kshatriya + Punjabi_Jat + Sindhi @ 3.087
15 Chechen + Kanjar + Kshatriya + Kurumba @ 3.106
16 Chechen + Kanjar + Kshatriya + Uttar_Pradesh @ 3.115
17 Brahmin_UP + Chechen + Kanjar + North_Kannadi @ 3.131
18 Chechen + Kanjar + Kanjar + Kshatriya @ 3.138
19 Brahmin_UP + Chechen + Kanjar + Kol @ 3.141
20 Brahmin_UP + Chechen + Kurumba + North_Kannadi @ 3.149

219970 iterations.

Done.

Btw, Sapporo, did you find it as funny as I did when you first heard about the Kanjar? My friend refused to believe the Wikipedia article on them was real, he only knew the word as an insult.

parasar
10-09-2013, 02:42 AM
An Indo-Hephthalite origin of Rajputs would explain the genetic proximity (including the influx of recent Caucasian admixture in a seemingly random pattern in Punjab, the only place it doesn't correlate to religion). The Rajputs mixed into the Jatts thoroughly so many Jatts today would have actual Hephthalite ancestry, it would make sense they'd be the closest match to the Pashtun. Even my own closest match in Northwestern India on DNA-Tribes was Pashtun... I was as close to Lithuanians as I was to Arain Punjabi even though they live in the same area.

Also, Sindh was ruled by a Rajput dynasty (usurped by a Brahmin shortly before the Arabs invaded) not long after the Hephthalite invasion of India (the Rai Dynasty, established in 489 CE). But before the Hephthalites, it was Shakas/Sakas (a.k.a. Indo-Scythians) who ruled it. When the Arabs got there, they said the inhabitants were Jats, Rajputs, Bhattis, Medes, etc.

Unfortunately this Hepthalite-Rajput connection is all made up and there little or no proof at all. The Shak are considered to be an entity spread from Gujarat, Malwa, Mathura, Khotan areas, while the evidence is quite robust, there are some holes in this formulation too.

As far as Rajputs ruling Sindh, that cannot be possible as the Rajput identity had not developed at that time. Rajputs could be from any ruling clan, and if I am not mistaken, initially the term referred to son of rulers (ie sons of those rulers who had been dispossessed by Mahmud).

The Arab accounts give us a reasonable account for the period.
If I had to guess who the ruler of Sindh was it would be a one of remnant Mauryas of Magadha, but this is only speculation. The reason I think it was a Maurya is because the Chachnama says: "After that period, the news of Rai Sáhasi's death coming to the ear of his brother Maharat, the king of Chitór, the latter prepared an expedition, and with a large army and followers, and furious elephants, and brave warriors, marched out to fight with Chach. He encamped within a league of Alor, and sent a number of his private servants and favourite domestics to Chach with the following message:—“ I am the rightful heir to this kingdom, and this country is the property of my fathers and grandfathers. It is but right that I should have for my own my brother's heritage: to you that same post of chamberlain and lieutenant will be awarded, and every endeavour will be made to treat you liberally.”"

Now from Mewar accounts we know that just prior to the period the Brahmans (later Sesodia Rajputs) came to rule Chittor the incumbent was a Maan Mori (= Maurya, likely later Pramar Rajputs).

Plus we also have some supporting inscriptional evidence from Nausari which says the Arabs [Tajiks] overan the Mauryas on their way to attacking Dakshinapatha. "Tajika ... attacked the kingdoms of Sindhu, Saurashtra, Cavataka, Maurya, and Gurjara, and had advanced as far as Navasari, where the prince was ruling at the time. The Arabs penetrated through the Baroda gap and plain near the Rann of Cutch." http://books.google.com/books?id=bCVyhH5VDjAC&pg=PA208

Sapporo
10-09-2013, 03:51 AM
We need to find some Brahmins and Punjabi Rajputs to compare with...

I'm sharing with a Punjabi Brahmin on 23andme but he's not that active on it. He is one of the two Punjabi Brahmins in HAP. I also believe HAP has two Pakistani Pahari Rajputs who are more or less Punjabis. I actually know one of them a bit but he's not active online anymore.

Sapporo
10-09-2013, 04:08 AM
Btw, Sapporo, did you find it as funny as I did when you first heard about the Kanjar? My friend refused to believe the Wikipedia article on them was real, he only knew the word as an insult.

Yeah, I find find it humorous even though I don't use much South Asian profanity (since I stick to English profanity when using it). It's even more funny to see them as a population on my oracle mixed mode. :p

Dr_McNinja
10-09-2013, 09:01 AM
Unfortunately this Hepthalite-Rajput connection is all made up and there little or no proof at all.That is true of most history but this theory has currency in major academic circles outside and within India.


As far as Rajputs ruling Sindh, that cannot be possible as the Rajput identity had not developed at that time.It was about a century before the earliest mention of them, it's as close as we're going to get. And it stands to reason they existed before their identity, so their own existence, presumably in an influential role, had to precede the formation of and the spread of the knowledge of their identity.


Rajputs could be from any ruling clan, and if I am not mistaken, initially the term referred to son of rulers (ie sons of those rulers who had been dispossessed by Mahmud).Of course, that's how the identity started. I don't think the Rajputs outside of Punjab or neighboring areas have any connection to the Hephthalites. I also didn't mean that the Rai of Sindh were the earliest or only Rajputs, just that they are usually identified as a Rajput dynasty.


If I had to guess who the ruler of Sindh was it would be a one of remnant Mauryas of Magadha, but this is only speculation.The Mauryas were out of power five centuries before this though. And the Rai came after the Shakas/Sakas apparently lost their rule in Sindh, so the continuity isn't there either. How would they have suddenly gotten there? The timing is what makes the Hephthalite origin theory attractive.


The reason I think it was a Maurya is because the Chachnama says: "After that period, the news of Rai Sáhasi's death coming to the ear of his brother Maharat, the king of Chitór, the latter prepared an expedition, and with a large army and followers, and furious elephants, and brave warriors, marched out to fight with Chach. He encamped within a league of Alor, and sent a number of his private servants and favourite domestics to Chach with the following message:—“ I am the rightful heir to this kingdom, and this country is the property of my fathers and grandfathers. It is but right that I should have for my own my brother's heritage: to you that same post of chamberlain and lieutenant will be awarded, and every endeavour will be made to treat you liberally.”"

Now from Mewar accounts we know that just prior to the period the Brahmans (later Sesodia Rajputs) came to rule Chittor the incumbent was a Maan Mori (= Maurya, likely later Pramar Rajputs).This says the Mori were a Jat clan who took the surname from the Mauryan period:

http://en.wikipedia.org/wiki/Mor_clan

They might or might not have an actual link to the Mauryan dynasty. Rajput, like the Jat originally, originated as a class with varying origins. The timing seems to suggest the Hephthalites integrated into the Kshatriya as Rajputs and Gujjar probably, and mixed with the Jats and others who were in Punjab. They had actually conquered Northwestern India so they were the rulers there, which gives them the foundation for establishing their own "Rajput" identity within the Kshatriya, as rulers of other backgrounds did in other areas.

parasar
10-09-2013, 03:53 PM
That is true of most history but this theory has currency in major academic circles outside and within India.

I very much doubt that is the case at present.

The theory was put forth by V.A. Smith: "who were these Rajputs ... The Mauryas and Guptas belong to a dead and buried past, remembered only through books, inscriptions, and coins, while the clans whose ruling families came into notice during the mediaeval period are still very much alive, and in many cases form numerous and influential sections of the existing population ... The break in tradition seems to be due in large measure .... Hun invasion ... "

And then repeated by A.L. Basham in his The Wonder That Was India.

I have looked carefully at most Rajput lines and only in a couple of instances is any Hun or Kushan connection warranted.





It was about a century before the earliest mention of them, it's as close as we're going to get. And it stands to reason they existed before their identity, so their own existence, presumably in an influential role, had to precede the formation of and the spread of the knowledge of their identity.

Where did you see Rajputs mentioned just a century after?
I agree that they existed before their identity. And if we look carefully we can establish the links going much before any Shak or Hun. The Mori - Pramar connection would be an example.



The Mauryas were out of power five centuries before this though. And the Rai came after the Shakas/Sakas apparently lost their rule in Sindh, so the continuity isn't there either. How would they have suddenly gotten there? The timing is what makes the Hephthalite origin theory attractive.

This says the Mori were a Jat clan who took the surname from the Mauryan period:

http://en.wikipedia.org/wiki/Mor_clan

They might or might not have an actual link to the Mauryan dynasty. Rajput, like the Jat originally, originated as a class with varying origins. The timing seems to suggest the Hephthalites integrated into the Kshatriya as Rajputs and Gujjar probably, and mixed with the Jats and others who were in Punjab. They had actually conquered Northwestern India so they were the rulers there, which gives them the foundation for establishing their own "Rajput" identity within the Kshatriya, as rulers of other backgrounds did in other areas.

I had provided inscriptional evidence from Nausari (~738AD) above that the Mauryas were indeed existing in a reduced form at the time of Arab invasions.
Other notices:

1. A Maurya chief Suketu Varma is mentioned in a 6th century inscription.

2. The Chalukyas of Badami battled the Mauryas. http://books.google.com/books?id=RzX2oerL7BcC&pg=PA326

3. Guttal: "These Mauryas were perhaps descendants of the Maurya dynasty of Pataliputra,2 which was founded by Chandragupta, the Sandrocottus of the Greeks, in the fourth century B.C., and of descendants of which we seem to have some still more recent traces in Western India in the twelfth and thirteenth centuries A.D. in the Mahamandalesvaras or great feudatory nobles of the Gutta family, or the lineage of Chandragupta, whose inscriptions are found at and in the neighbourhood of Chaudadampur in the Dharwad District, and at Halebid in Maisur, and who were feudatories of the Western Chalukya Kings and their successors." (Mr. Fleet's " Dynasties of the Kanarese Districts" pp. 6, 7.)
http://books.google.com/books?id=S8wOAAAAQAAJ&pg=PA198

4. Tradition reported by Tod: "another tradition often repeated by the same learned inquirer [Tod], both in his Annals of Rajasthan and his contributions to the R. A. S. Transactions ... One tribe of this Rajput race, the Mori, is in the habit at this day of claiming for their own the celebrated Chandragupta Maurya, founder of the dynasty so called at Palibothra, in the days of Seleucus Nicator."

The whole history of northern India has been made problematic because of lack of clear understanding of the Tajik (Arab) invasion. It is around the period of Arab invasions that the clans (Chalukya-Solanki, Rastrakuta-Rathore, Chahamana-Chauhan, Parmara-Puar, Prathihara-Parihar, etc), later to be coined Rajput, came into prominence.

(Though we are not Rajputs, we have a branch - Bisen http://www.joshuaproject.net/people-profile.php?peo3=20382&rog3=IN - now under the Rajput umbrella. Some nuggets on the non-Rajput branches.
http://www.thehindu.com/multimedia/dynamic/00636/22_sm_sinha_1_jpg_636595g.jpg
http://www.thehindu.com/features/magazine/vignettes-from-an-age-of-war/article2037700.ece )

Dr_McNinja
10-09-2013, 10:27 PM
Found a Nepalese Brahmin user at Anthroscape, his numbers:

K15:

North_Sea 9.02%
Atlantic 2.66%
Baltic-
Eastern_Euro 10.00%
West_Med-
West_Asian 21.43%
East_Med-
Red_Sea-
South_Asian 49.69%
Southeast_Asian 4.03%
Siberian 1.68%
Amerindian 0.81%
Oceanian 0.66%
Northeast_African-
Sub-Saharan-

K36:


Amerindian 0.17%
Arabian-
Armenian-
Basque-
Central_African-
Central_Euro 2.72%
East_African-
East_Asian 1.46%
East_Balkan-
East_Central_Asian-
East_Central_Euro-
East_Med-
Eastern_Euro-
Fennoscandian 3.66%
French-
Iberian-
Indo-Chinese 0.37%
Italian-
Malayan 0.65%
Near_Eastern-
North_African-
North_Atlantic 4.21%
North_Caucasian 6.93%
North_Sea 2.62%
Northeast_African-
Oceanian0.62%
Omotic-
Pygmy-
Siberian 0.69%
South_Asian 37.83%
South_Central_Asian 34.10%
South_Chinese 2.06%
Volga-Ural 0.91%
West_African-
West_Caucasian 0.99%
West_Med-

Even more West shifted than me.

Sapporo
10-10-2013, 12:29 AM
Found a Nepalese Brahmin user at Anthroscape, his numbers:

K15:


K36:



Even more West shifted than me.


Well, I thought you were specifically looking for a Punjabi Brahmin. That Nepalese Brahmin is my friend. I'm also sharing with a Kashmiri Pandit on 23andme. I can always try to ask her to provide her calculator results or join anthrogenica.

I'm curious which of the European components is more "Northern shifted" out of Baltic, North Sea and Eastern European. I think Atlantic would be more Southwestern European shifted since it is also very high in Iberians. The North Sea is a region shared between the British Isles, Germany, Northern France, Belgium and Western Scandinavia. I presume Baltic is simply Northeastern European like Lithuania but what about Eastern European? Would it peak in Russians and be high in Baltic populations and Finns too? Or is it high in Slavic speaking Balkan populations too?

Dr_McNinja
10-23-2013, 03:37 AM
Here are the results of an Afghan Pashtun, Popalzai from both sides, direct descendant of Ahmad Shah Durrani on both sides (from Kandahar):


Harappa DIY

S-Indian 15.04%
Baloch 35.35%
Caucasian 22.88%
NE-Euro 13.60%
SE-Asian -
Siberian 2.29%
NE-Asian 0.93%
Papuan -
American 1.06%
Beringian 0.49%
Mediterranean 3.37%
SW-Asian 4.85%
San 0.13%
E-African -
Pygmy -
W-African -

Eurogenes K36

Amerindian 0.45%
Near_Eastern 5.32%
Arabian 1.12%
Central_African 0.30%
East_Med 0.07%

North_Atlantic 0.50%
North_Sea 0.09%
Volga-Ural 0.71%
Iberian 2.51%
Fennoscandian 4.94%
Eastern_Euro 2.35%
East_Balkan 1.72%
East_Central_Euro 1.44%


North_Caucasian 16.80%
West_Caucasian 2.93%
Armenian 4.53%


South_Asian 9.90%
South_Central_Asian 44.31%

Eurogenes K13

North European 17.11%
West African 0.50%
Mediterranean 5.87%
Northeast African 0.95%
North Eurasian 2.57%
South Asian 20.85%
Southwest Asian 2.83%
Pygmy 0.06%
Caucasus 19.78%
East Siberian 0.37%
East Asian -
Amerindian 1.05%
West Central Asian 28.07%

Eurogenes K12b

Western European 6.03%
Siberian 3.07%
East African 1.32%
West Central Asian 32.37%
South Asian 15.73%
West African -
Caucasus 22.09%
Finnish 2.79%
Mediterranean 3.50%
Southwest Asian 3.07%
North European 10.01%
East Asian -
EUTest V2 K15

North_Sea 2.76%
Atlantic 5.68%
Baltic 4.41%
Eastern_Euro 13.01%
West_Med -
West_Asian 32.26%
East_Med 9.06%
Red_Sea 1.27%
South_Asian 27.52%
Southeast_Asian -
Siberian 1.95%
Amerindian 1.12%
Oceanian -
Northeast_African 0.93%
Sub-Saharan -
Very interesting. I don't know what 'Iberian' in K36 is but there's this seemingly ancient West European showing up here that I haven't seen in non-Brahmin populations. Could be the result of royal intermarrying when the Abdali clan (direct ancestors of the Durranis) came to Afghanistan with the previous high/royal class. I wonder if other Afghan Pashtun from the same area but of different lineages will also have it.

Sapporo
10-23-2013, 06:42 AM
@Dr_McNinja

Where did you get his results from? Quite interesting how he is much less South Asian than the HGDP Pashtuns overall (although pretty similar to HRP0286 without as much East Eurasian ancestry; I'll note that some of the individual HGDP Pashtuns are only slightly more South Asian than him) and much more Southwest Asian/Med influenced. I wonder if his results are typical for all Popalzai or his results are influenced from his direct ancestry from Ahmad Shah Durrani. He also seems to have a higher Caucasus component in addition to a more consistent Southwestern/Western Euro component that is probably Med-like in addition to his Northern Euro which is about the same as the HGDP Pashtuns.

I believe Sein showed me the results of a Kandahari Pashtun with very similar results to this individual. At least his DIY Harappaworld results.

DIY HarappaWorld:

Baloch-34.13%
Caucasian-21.92%
S-Indian-17.22%
NE-Euro-13.10%
SW-Asian-5.68%
NE-Asian-2.73%
Siberian-1.82%
Mediterranean-1.21%
Beringian-1.12%
American-0.65%
SE-Asian-0.42%

Sein
10-23-2013, 08:19 AM
Thank you very much Dr_McNinja! Is this individual going to participate in HAP? I tried to get the other Popalzai individual, whose results Sapporo presented, to participate, but they just weren't interested. I really hope you can get your Popalzai friend to participate. You once mentioned that you know a few Popalzai individuals, and that you would encourage them to get genotyped via 23andMe? This all makes me very anxious to see more Afghan Pashtun results. Also, is it possible for you to tell us this individual's paternal and maternal haplogroups?

Sein
10-23-2013, 11:57 AM
@ Sapporo - I think your idea about Ahmad Shah Durrani is interesting, but I would note that Ahmad Shah Durrani was fully Pashtun. Also, you are definitely right, at least three (I think) HGDP Pashtuns are only slightly more South Asian than this individual. Also, one HGDP Pashtun is only 12% "South Indian". This individual clusters with all other HGDP Pashtuns in ChromoPainter/FineStructure analyses, and also in Mclust analyses, so despite being much less South Asian, this individual HGDP Pashtun seems to be very close to the other samples. But this leaves one question. If HGDP00214 is not unusual in comparison to the HGDP Pashtuns, why is there such a gap between this individual and other HGDP Pashtuns? It seems the gap is beginning to be bridged by Afghan Pashtuns. If the HGDP sample was much bigger, say 100+ individuals, I wonder how much the "South Indian" component would vary? Are there any individuals from the HGDP sampling location who would bridge the gap between HGDP0214 and other individuals in the HGDP dataset? I think so. I find it difficult to believe that HGDP0214 is a lone individual. If the HGDP sample was 100+, I'd guess the range of the "South Indian" component would be between 8% and 40%.

Dr_McNinja
10-23-2013, 02:05 PM
Thank you very much Dr_McNinja! Is this individual going to participate in HAP? I tried to get the other Popalzai individual, whose results Sapporo presented, to participate, but they just weren't interested. I really hope you can get your Popalzai friend to participate. You once mentioned that you know a few Popalzai individuals, and that you would encourage them to get genotyped via 23andMe? This all makes me very anxious to see more Afghan Pashtun results. Also, is it possible for you to tell us this individual's paternal and maternal haplogroups?It's a she, her maternal haplogroup is J1b. I can get her brother to do a FTDNA YDNA-12 test but the results for that will take some time since FTDNA is slower.

Sein
10-23-2013, 02:12 PM
Oh, interesting. Can you get her to participate in HAP?

Dr_McNinja
10-23-2013, 02:13 PM
@ Sapporo - I think your idea about Ahmad Shah Durrani is interesting, but I would note that Ahmad Shah Durrani was fully Pashtun. Also, you are definitely right, at least three (I think) HGDP Pashtuns are only slightly more South Asian than this individual. Also, one HGDP Pashtun is only 12% "South Indian". This individual clusters with all other HGDP Pashtuns in ChromoPainter/FineStructure analyses, and also in Mclust analyses, so despite being much less South Asian, this individual HGDP Pashtun seems to be very close to the other samples. But this leaves one question. If HGDP00214 is not unusual in comparison to the HGDP Pashtuns, why is there such a gap between this individual and other HGDP Pashtuns? It seems the gap is beginning to be bridged by Afghan Pashtuns. If the HGDP sample was much bigger, say 100+ individuals, I wonder how much the "South Indian" component would vary? Are there any individuals from the HGDP sampling location who would bridge the gap between HGDP0214 and other individuals in the HGDP dataset? I think so. I find it difficult to believe that HGDP0214 is a lone individual. If the HGDP sample was 100+, I'd guess the range of the "South Indian" component would be between 8% and 40%.
I think there might be more South Indian in people whose ancestries trace to cities, since this person's family has been in Kandahar for many generations and there's always been some migration to and from cities.

The other thing is the West Euro/Med component showing up here which I think is possibly a holdover from the Abdalis/Durranis marrying into the high caste former rulers of Afghanistan since they came down into the area as Hephthalites or whatever they were. The previous dynasties of Afghanistan were Buddhist Shahis, some Indian, some Turkic, etc and I'm sure one segued into the next via intermarriage (apparently Jatts say these Shahis were also Jatts, they basically call all the indigenous non-Brahmin Indo-European endogamous groups of Asia, "Jat"). It's too "Western" to be modern Jatt or Kshatriya (mine only went to North Sea), it looks like really old/indigenous Brahmin, which in the Afghanistan area could be the old ruling classes. This could possibly correlate to the relative lack of NE-Asian... perhaps the higher classes didn't mix into the incoming Mongolians and Central Asians. The Abdali family was mostly set by the 1100s.

Dr_McNinja
10-23-2013, 02:16 PM
Oh, interesting. Can you get her to participate in HAP?Yes, I'm going to send her data to Zack

Sein
10-23-2013, 02:50 PM
Yes, I'm going to send her data to Zack

Great!

I think the tribal history of the Abdali is relevant here. The Abdali didn't really start out as an independent tribal unit. They are actually just a Tarin lineage, Bor to be exact. The other Tarin groups are the Tor, Spin, and Zharh Tarin (literal translation of these terms is "Black", "White", and "Yellow" Tarin). Most of these close relatives of the Abdali are found in Pakistani Balochistan, where they constitute a large percentage of that province's Pashtun population. The only reason the Abdali are categorized separately from these groups is because they have been very prolific (the Abdali are a huge tribe), and have been on an independent trajectory since the 18th century. Also, I think Kandahar did not have a Pashtun population in the 11th century. Babur was a very observant, and exceedingly intelligent, writer. Whenever he encountered a Pashtun tribe, he noted their existence and location in the Baburnama (along with all the trouble they caused him; reading the stuff he wrote about Waziri Pashtuns is pretty entertaining, because they still behave the same way, you just can't get any taxes out of them). Almost every well known Pashtun tribe the length and breadth of AF-PAK is noted by him, and he finds/fights most of them where they still live. Yet, when he travels in the present Abdali territory, he finds no Pashtuns. Instead, the region was inhabited by Nukdari people, completely unrelated to Pashtuns. That was the early 16th century.

Dr_McNinja
10-23-2013, 07:51 PM
Great!

I think the tribal history of the Abdali is relevant here. The Abdali didn't really start out as an independent tribal unit. They are actually just a Tarin lineage, Bor to be exact. The other Tarin groups are the Tor, Spin, and Zharh Tarin (literal translation of these terms is "Black", "White", and "Yellow" Tarin). Most of these close relatives of the Abdali are found in Pakistani Balochistan, where they constitute a large percentage of that province's Pashtun population. The only reason the Abdali are categorized separately from these groups is because they have been very prolific (the Abdali are a huge tribe), and have been on an independent trajectory since the 18th century. Also, I think Kandahar did not have a Pashtun population in the 11th century. Babur was a very observant, and exceedingly intelligent, writer. Whenever he encountered a Pashtun tribe, he noted their existence and location in the Baburnama (along with all the trouble they caused him; reading the stuff he wrote about Waziri Pashtuns is pretty entertaining, because they still behave the same way, you just can't get any taxes out of them). Almost every well known Pashtun tribe the length and breadth of AF-PAK is noted by him, and he finds/fights most of them where they still live. Yet, when he travels in the present Abdali territory, he finds no Pashtuns. Instead, the region was inhabited by Nukdari people, completely unrelated to Pashtuns. That was the early 16th century.
That's fascinating. So where were they in the meantime? Outside of South Asia or in Pakistan/India?

Dr_McNinja
10-23-2013, 08:59 PM
What were the Oracle results for the other Pashtun? This was hers in Harappa I believe:


Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 61.1% haryana-jatt (harappa) + 38.9% kurd (xing) @ 3.08
2 65.2% bhatia (harappa) + 34.8% turk-istanbul (hodoglugil) @ 3.21
3 53.7% sindhi (hgdp) + 46.3% stalskoe (xing) @ 3.32
4 72.5% pathan (hgdp) + 27.5% turk-istanbul (hodoglugil) @ 3.35
5 57.1% haryana-jatt (harappa) + 42.9% iranian (harappa) @ 3.39
6 65.1% bhatia (harappa) + 34.9% turk-aydin (hodoglugil) @ 3.45
7 72.4% kalash (hgdp) + 27.6% turk-aydin (hodoglugil) @ 3.48
8 67.7% pathan (hgdp) + 32.3% stalskoe (xing) @ 3.48
9 72.4% pathan (hgdp) + 27.6% turk-aydin (hodoglugil) @ 3.48
10 62.3% sindhi (harappa) + 37.7% stalskoe (xing) @ 3.5
11 66% bhatia (harappa) + 34% turk (behar) @ 3.59
12 73.2% pathan (hgdp) + 26.8% turk-kayseri (hodoglugil) @ 3.62
13 73.2% pathan (hgdp) + 26.8% turk (behar) @ 3.62
14 65.9% bhatia (harappa) + 34.1% turk-kayseri (hodoglugil) @ 3.63
15 76.8% kalash (hgdp) + 23.2% ashkenazi (harappa) @ 3.67
16 76.6% kalash (hgdp) + 23.4% ashkenazy-jew (behar) @ 3.78
17 53.8% punjabi-arain (xing) + 46.2% stalskoe (xing) @ 3.85
18 71.4% pathan (hgdp) + 28.6% kumyk (yunusbayev) @ 3.86
19 71.1% pathan (hgdp) + 28.9% turkish (harappa) @ 4.01
20 59.2% haryana-jatt (harappa) + 40.8% kurd (harappa) @ 4.04

And the 4-population in the EUTest V2 K15:


Using 4 populations approximation:
1 Adygei + Balochi + Brahmin_UP + Tadjik @ 4.918
2 Brahmin_UP + Burusho + Chechen + Kurdish @ 5.104
3 Balochi + Brahmin_UP + Chechen + Tadjik @ 5.124
4 Adygei + Brahmin_UP + Brahui + Tadjik @ 5.14I think that Brahmin and Haryana Jatt reflects the West Euro/Med bit? Or is that common in the results of other Pashtun too?

EDIT: Here are the Harappa Oracle-4:


Using 4 populations approximation:
1 bhatia + haryana-jatt + kurd + tajik @ 2.213
2 haryana-jatt + kurd + pathan + tajik @ 2.324
3 haryana-jatt + kurd + pathan + tajik @ 2.325
4 bhatia + haryana-jatt + kurd + tajik @ 2.441
5 balochi + bhatia + romanian-b + urkarah @ 2.460
6 bhatia + brahui + georgian + singapore-indian-d @ 2.515
7 bhatia + brahui + romanian-b + urkarah @ 2.519
8 balochi + balochi + north-ossetian + singapore-indian-d @ 2.572
9 haryana-jatt + kalash + kurd + tajik @ 2.581
10 haryana-jatt + kalash + kurd + tajik @ 2.581
11 adygei + balochi + balochi + singapore-indian-d @ 2.593
12 haryana-jatt + iranian + kalash + tajik @ 2.641
13 haryana-jatt + kalash + kurd + tajik @ 2.685
14 brahui + punjabi-jatt-sikh + romanian-b + urkarah @ 2.691
15 brahui + romanian-b + sindhi + urkarah @ 2.697
16 haryana-jatt + kurd + sindhi + tajik @ 2.704
17 brahui + burusho + romanian-b + urkarah @ 2.723
18 haryana-jatt + kurd + sindhi + tajik @ 2.742
19 haryana-jatt + kurd + punjabi-khatri + tajik @ 2.796
20 haryana-jatt + kurd + pathan + tajik @ 2.812Consistent Haryana Jatt, I'm guessing due to the European admixture.

Sein
10-23-2013, 09:10 PM
You know, this is a very interesting question. It's a complete mystery to me. I'd assume Pakistan (especially Northern Balochistan), or maybe further north in eastern Afghanistan, probably the Paktia region, which has been solidly Pashtun territory for at least 1,000+ years. Another interesting thing about the Abdali is that after settling in the Kandahar region, they were expelled by the Ghilzai. They basically fled to Herat. After Abdali ascendancy, the region was "re-Durranified". The Ghilzai are an exceedingly tough and aggressive group of Pashtuns, even for an ethnicity that is culturally predisposed towards violence and truculence (lol). The Abdali are very peaceful in comparison to most Pashtun tribes, and are far from warlike. The only reason the Ghilzai lost Kandahar is because they are very disorganized, and like Karlanri Pashtuns, lack any reliance on centralized authority. The Abdali, by contrast, are the most hierarchical of all Pashtuns, and are very well segmented and differentiated (a lot of Pashtun tribes in the south are hierarchical in political structure. The kind of political freedom/anarchy enjoyed-suffered by the Afridi or Mohmand of the north is sorely lacking. This deserves some anthropological research, but I think agriculture plays a role, as does the terrain). They simply took advantage of Ghilzai infighting.

Sein
10-23-2013, 09:31 PM
My other Pashtun friend's oracle results for HAP oracle:
Admix Results (sorted):

# Population Percent
1 Baloch 34.13
2 Caucasian 21.92
3 S-Indian 17.22
4 NE-Euro 13.1
5 SW-Asian 5.68
6 NE-Asian 2.73
7 Siberian 1.82
8 Mediterranean 1.21
9 Beringian 1.12
10 American 0.65
11 SE-Asian 0.42

Single Population Sharing:

# Population (source) Distance
1 pashtun (harappa) 4.4
2 pathan (hgdp) 11.15
3 kalash (hgdp) 11.82
4 tajik (yunusbayev) 12.14
5 sindhi (harappa) 13.49
6 punjabi-khatri (harappa) 14.05
7 burusho (hgdp) 14.42
8 kashmiri (harappa) 14.83
9 bhatia (harappa) 15.9
10 punjabi-jatt-sikh (harappa) 15.98
11 haryana-jatt (harappa) 16.08
12 punjabi-jatt-muslim (harappa) 16.74
13 turkmen (yunusbayev) 17.02
14 gujarati-muslim (harappa) 18.02
15 kashmiri-pandit (reich) 18.05
16 punjabi (harappa) 18.74
17 up-muslim (harappa) 19.71
18 punjabi-brahmin (harappa) 19.85
19 punjabi-arain (xing) 19.97
20 sindhi (hgdp) 20.17

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 89.8% pashtun (harappa) + 10.2% urkarah (xing) @ 3.12
2 90.3% pashtun (harappa) + 9.7% lezgin (behar) @ 3.16
3 89.8% pashtun (harappa) + 10.2% stalskoe (xing) @ 3.26
4 91.8% pashtun (harappa) + 8.2% chechen (yunusbayev) @ 3.27
5 91.1% pashtun (harappa) + 8.9% kumyk (yunusbayev) @ 3.28
6 93.2% pashtun (harappa) + 6.8% adygei (hgdp) @ 3.45
7 93.1% pashtun (harappa) + 6.9% balkar (yunusbayev) @ 3.45
8 92.2% pashtun (harappa) + 7.8% turk-istanbul (hodoglugil) @ 3.47
9 93.3% pashtun (harappa) + 6.7% north-ossetian (yunusbayev) @ 3.49
10 66.5% sindhi (harappa) + 33.5% stalskoe (xing) @ 3.49
11 91.3% pashtun (harappa) + 8.7% kurd (xing) @ 3.51
12 92.6% pashtun (harappa) + 7.4% turk (behar) @ 3.53
13 90% pashtun (harappa) + 10% iranian (harappa) @ 3.55
14 91.8% pashtun (harappa) + 8.2% azeri (harappa) @ 3.55
15 86.6% pashtun (harappa) + 13.4% turkmen (yunusbayev) @ 3.56
16 92.7% pashtun (harappa) + 7.3% turk-kayseri (hodoglugil) @ 3.56
17 81.8% pashtun (harappa) + 18.2% tajik (yunusbayev) @ 3.56
18 93.9% pashtun (harappa) + 6.1% armenian (behar) @ 3.57
19 92.7% pashtun (harappa) + 7.3% uzbekistan-jew (behar) @ 3.58
20 92.6% pashtun (harappa) + 7.4% turk-aydin (hodoglugil) @ 3.6

Dr_McNinja
10-23-2013, 11:05 PM
Very interesting. So there's a noticeable difference among some Pashtun groups.

Dr_McNinja
10-23-2013, 11:07 PM
Oh and she only had 2.5% Neanderthal on 23andMe which I thought was on the low side for someone with so much Caucasian/Euro admixture.

Dr_McNinja
10-23-2013, 11:44 PM
Here's the Ancestry Composition: http://i.imgur.com/tEHsiwF.png

Sein
10-23-2013, 11:57 PM
Based on what we're seeing, I think you're right. Nevertheless, I wish we had substantial samples for Pashtuns both west and east of the Durand line. The average HGDP Pashtun seems to conform very well with the ancestry of most Pashtuns, but the "Baloch" component is the real difference. Pakistani or Afghan, all non-HGDP Pashtuns score Baloch/Gedrosia percentages significantly lower than any individual in the HGDP data-set. There is a HGDP Pashtun with a very low South Indian percentage, at 12%, and two individuals who are at 18%, so the differences in the South Indian component are more moderate. But non-HGDP Pashtuns have percentages that compensate less Baloch/Gedrosia with higher Southwest Asian-Mediterranean affinity, more Northern European affinity, and more East Eurasian affinity. It would be awesome if we had 300 samples each from every predominantly Pashtun province in Afghanistan, 300 samples each from every Federally Administered Tribal Area in Pakistan, and 300 samples each from every large district in Khyber Pakhtunkhwa. That may sound like more than necessary, but I really think this would be the only way we could state anything with confidence, and answer every possible question. This applies to every large, geographically expansive population. Although, I know this is just a nice thought I have, more of a wish. Something like this is logistically almost impossible. But we can dream :-). Anyway, huge sample sizes can be found in the uniparental haplogroup literature, if only they spent the same amount of money on autosomal genotyping.

The interesting thing is that East Eurasians are actually more Neanderthal shifted than West Eurasians. In fact, significantly more so. I think this makes sense, since West Eurasians are significantly more shifted towards Sub-Saharan Africans than East Eurasians are. A second Out-of-Africa event, but only affecting West Eurasians? Or just isolation-by-distance, leaving West Eurasians intermediate between Africans and East Eurasians? I'm 2.6% Neanderthal on 23andMe, and the other Popalzai Pashtun whose results you've just seen is 3.0%

Sein
10-24-2013, 12:01 AM
Very interesting results, would I be able to share with her on 23andMe? Also, could you post a screen shot of her results at speculative?

Dr_McNinja
10-24-2013, 01:06 AM
She won't let me make her profile public or share it, she only let me share her results here and on HAP with identifying information hidden. Here's a shot at speculative:

http://i.imgur.com/2Fadss7.png

Dr_McNinja
10-24-2013, 01:21 AM
Is there anywhere I can find more information on mtdna haplogroup J1b? 23andMe seems to have the most data on it with their heatmap:

http://i.imgur.com/fz47y9d.png

This is all I've found:


J*: The original haplogroup J originated in the Near East approximately 50,000 years ago. Within Europe, sub-lineages of haplogroup J have distinct and interesting distributions. Haplogroup J* — the root lineage of haplogroup J — is found distributed throughout Europe, but at a relatively low frequency. Haplogroup J* is generally considered one of the prominent lineages that was part of the Neolithic spread of agriculture into Europe from the Near East beginning approximately 10,000 years ago.

J1b1: The mitochondrial haplogroup J contains several sub-lineages. The original haplogroup J originated in the Near East approximately 50,000 years ago. Within Europe, sub-lineages of haplogroup J have distinct and interesting distributions. Haplogroup J1b is found distributed in the Near East and southern Iberia, and may have been part of the original colonization wave of Neolithic settlers moving around the Mediterranean 6000 years ago or perhaps a lineage of Phoenician traders. Within haplogroup J1b, a derivative lineage haplogroup J1b1 has been found in Britain and another sub-lineage detected in Italy.Which sounds just like Y-DNA haplogroup J... Is that on purpose?

They mention Iberia and she had some Iberian show up on her Eurogenes K36 run which was weird. The other Eurogenes tests just showed West European.

Dr_McNinja
10-24-2013, 01:25 AM
Sein, if you share with me, I can do a one-to-one between you two. I'll PM you a link to my 23andMe profile.

Sein
10-24-2013, 01:32 AM
Thank you very much!

In regards to the haplogroup, I really don't think I can think of anything, I'm very uninformed when it comes to haplogroups.

Dr_McNinja
10-24-2013, 01:40 AM
Her One-To-Many comparison via my 23andMe profile:



Sein 75.65%
Dr_McNinja 75.64%
Sapporo 75.62%
Indian 75.38%
Rajasthani Indian 75.36%
South Indian (Karnatka) Brahmin 75.31%

Dr_McNinja
10-24-2013, 01:46 AM
This was your profile, right, Sein? You have a 75.89% match with the South Indian Brahmin on my list which I found weird...

Sapporo
10-24-2013, 01:50 AM
This was your profile, right, Sein? You have a 75.89% match with the South Indian Brahmin on my list which I found weird...

I think it's because we're comparing the genomes of individuals. Although, Sein will be much closer to the Popalzai individual and other Pashtuns on a PCA cluster plot or population distance wise, he shares slightly more with regards to his genome with the South Indian Brahmin on an individual level.

I have a few non Northwest South Asians who are slightly higher on my genome sharing list than actual Northwest South Asians but in general, people from my ethnicity or related populations are at the top.

Sein
10-24-2013, 01:52 AM
Thanks! That's pretty interesting, correct me if I'm wrong, but that's a high level of sharing, right? Must be a artifact of small sample size, since the more shares I get, the lower my similarity numbers get.

Also, I think Sapporo is absolutely correct, this approach is not really looking at population differentiating SNPs, but the whole genome.

Sein
10-24-2013, 04:26 AM
Also, I used the phasing function at GEDMatch, as I have my mother's raw-data. My mother's results for various calculators are rather different from my own, so I was interested in how my father would stack up. The percentages are in line with me being an average between my father and mother. My father's results:

Admix Results (sorted):

# Population Percent
1 Baloch 31.41
2 Caucasian 18.05
3 S-Indian 18.04
4 NE-Euro 12.35
5 SW-Asian 5.13
6 NE-Asian 3.18
7 Siberian 2.69
8 E-African 2.27
9 American 1.59
10 Papuan 1.55
11 Beringian 1.55
12 Pygmy 0.87
13 SE-Asian 0.76
14 Mediterranean 0.33
15 W-African 0.21

Single Population Sharing:

# Population (source) Distance
1 pashtun (harappa) 4.28
2 pathan (hgdp) 11.66
3 tajik (yunusbayev) 13.11
4 kalash (hgdp) 13.14
5 sindhi (harappa) 13.15
6 burusho (hgdp) 13.37
7 punjabi-khatri (harappa) 13.39
8 kashmiri (harappa) 13.65
9 haryana-jatt (harappa) 14.65
10 punjabi-jatt-sikh (harappa) 14.82
11 punjabi-jatt-muslim (harappa) 15.33
12 bhatia (harappa) 16.1
13 gujarati-muslim (harappa) 16.57
14 kashmiri-pandit (reich) 16.98
15 punjabi (harappa) 17.37
16 up-muslim (harappa) 17.65
17 turkmen (yunusbayev) 18.22
18 punjabi-brahmin (harappa) 18.8
19 singapore-indian-c (sgvp) 19.24
20 punjabi-arain (xing) 19.33

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 95.9% pashtun (harappa) + 4.1% somali (reich) @ 3.03
2 95.7% pashtun (harappa) + 4.3% afar (pagani) @ 3.08
3 96.1% pashtun (harappa) + 3.9% somali (pagani) @ 3.08
4 96.1% pashtun (harappa) + 3.9% esomali (pagani) @ 3.08
5 96% pashtun (harappa) + 4% somali (harappa) @ 3.09
6 95.8% pashtun (harappa) + 4.2% oromo (pagani) @ 3.09
7 95.6% pashtun (harappa) + 4.4% amhara (pagani) @ 3.09
8 95.7% pashtun (harappa) + 4.3% ethiopian (behar) @ 3.1
9 95.6% pashtun (harappa) + 4.4% tygray (pagani) @ 3.1
10 96% pashtun (harappa) + 4% wolayta (pagani) @ 3.1
11 95.7% pashtun (harappa) + 4.3% ethiopian-jew (behar) @ 3.11
12 96.4% pashtun (harappa) + 3.6% maasai (1000genomes) @ 3.15
13 96.6% pashtun (harappa) + 3.4% aricultivator (pagani) @ 3.19
14 95.8% pashtun (harappa) + 4.2% sandawe (henn2011) @ 3.21
15 96.7% pashtun (harappa) + 3.3% ariblacksmith (pagani) @ 3.21
16 96.1% pashtun (harappa) + 3.9% hadza (henn2011) @ 3.23
17 92.2% pashtun (harappa) + 7.8% uyghur (hgdp) @ 3.23
18 97% pashtun (harappa) + 3% sudanese (pagani) @ 3.28
19 97.1% pashtun (harappa) + 2.9% anuak (pagani) @ 3.29
20 91% pashtun (harappa) + 9% uzbek (behar) @ 3.3

My mother's results:

Single Population Sharing:

# Population (source) Distance
1 kalash (hgdp) 3.7
2 pathan (hgdp) 4.17
3 sindhi (harappa) 5.87
4 punjabi-khatri (harappa) 6.52
5 bhatia (harappa) 7.02
6 kashmiri (harappa) 8.7
7 punjabi-jatt-sikh (harappa) 9.03
8 punjabi-jatt-muslim (harappa) 9.5
9 kashmiri-pandit (reich) 9.71
10 burusho (hgdp) 10.08
11 pashtun (harappa) 10.37
12 punjabi-arain (xing) 10.48
13 sindhi (hgdp) 10.73
14 punjabi (harappa) 11.02
15 punjabi-brahmin (harappa) 11.21
16 singapore-indian-c (sgvp) 11.87
17 kashmiri-pahari (harappa) 12.07
18 haryana-jatt (harappa) 12.16
19 gujarati-muslim (harappa) 12.24
20 punjabi-ramgarhia (harappa) 13.01

Mixed Mode Population Sharing:

# Primary Population (source) Secondary Population (source) Distance
1 95.9% kalash (hgdp) + 4.1% irula (xing) @ 2.38
2 68.3% kalash (hgdp) + 31.7% punjabi-khatri (harappa) @ 2.4
3 94.6% kalash (hgdp) + 5.4% hakkipikki (metspalu) @ 2.42
4 94.6% kalash (hgdp) + 5.4% north-kannadi (chaubey) @ 2.43
5 93.4% kalash (hgdp) + 6.6% hallaki (reich) @ 2.44
6 94.8% kalash (hgdp) + 5.2% tn-dalit (xing) @ 2.45
7 96% kalash (hgdp) + 4% pulliyar (metspalu) @ 2.45
8 94.3% kalash (hgdp) + 5.7% sakilli (chaubey) @ 2.46
9 94.5% kalash (hgdp) + 5.5% ap-madiga (xing) @ 2.47
10 94.4% kalash (hgdp) + 5.6% tamil-vishwakarma (harappa) @ 2.47
11 94.7% kalash (hgdp) + 5.3% kurumba (reich) @ 2.47
12 89.5% kalash (hgdp) + 10.5% cochin-jew (behar) @ 2.48
13 94.9% kalash (hgdp) + 5.1% mala (reich) @ 2.48
14 94.9% kalash (hgdp) + 5.1% ap-mala (xing) @ 2.48
15 93.4% kalash (hgdp) + 6.6% naidu (reich) @ 2.49
16 94.6% kalash (hgdp) + 5.4% madiga (reich) @ 2.49
17 93.6% kalash (hgdp) + 6.4% piramalai-kallar (metspalu) @ 2.49
18 94.4% kalash (hgdp) + 5.6% bhil (reich) @ 2.49
19 94.7% kalash (hgdp) + 5.3% chenchu (reich) @ 2.5
20 93% kalash (hgdp) + 7% tamil-nadu-scheduled-caste (metspalu) @ 2.5

These agree with her non-phased results, which is surprising, since people say the results can differ, but not in her case.

AJL
10-24-2013, 04:40 AM
The percentages are in line with me being an average between my father and mother.

That usually happens if your parents are of fairly similar backgrounds -- the more different two parents are to each other, the less predictable the results of their children can be.

Sapporo
10-24-2013, 05:06 AM
@Sein

Interesting results for both your parents. Your father is like one of the HAP Pashtuns with a tiny East African shift while your mother is like a Kalash with a South Asian shift. Your father is also more like the HAP Pashtuns while your mother more like the HGDP Pashtuns. So, your father is closest to the HAP Pashtuns by far and then the HGDP Pashtuns, Yunusbayev Tajiks and HGDP Kalash before Northwestern South Asians like the HAP Sindhis, HAP Punjabi Khatris, HAP Kashmiris, HAP Punjabi Jatt Sikhs, etc. Although, the sample sizes are quite small for the HAP Pashtuns (although I suspect they will of course remain at the top even with larger sample sizes), and the HAP Sindhis (2), HAP Khatris (2) and HAP Kashmiris (3).

Your mother is a bit different. She is closest to to the HGDP Kalash and HGDP Pashtuns before some a few Northwestern South Asians and the HAP Pashtuns. Of course, sample sizes are very small though except for the Punjabi Jatt Sikhs at 12.

Sein
10-24-2013, 05:37 AM
@ Sapporo - The sample sizes are always tortuous in my mind. If only the minimum was 50 samples for each population. I hope as HAP grows, Zack can get some participation from many unsampled ethnic groups, like Afghan Tajiks, or Nuristanis, or maybe some of the tribal populations of southern-eastern India. It is great that we finally have at least 12 Punjabi Jatts, we can now say some substantive things with confidence.

Sapporo
10-24-2013, 06:05 AM
@ Sapporo - The sample sizes are always tortuous in my mind. If only the minimum was 50 samples for each population. I hope as HAP grows, Zack can get some participation from many unsampled ethnic groups, like Afghan Tajiks, or Nuristanis, or maybe some of the tribal populations of southern-eastern India. It is great that we finally have at least 12 Punjabi Jatts, we can now say some substantive things with confidence.

50 samples of each of the major populations would be quite incredible and so much data and information could be ascertained from such sizes. However, at the moment, we're stuck with the samples from studies/projects like HGDP, Xing, Reich, Behar, etc. I'd be happy with 15-20 samples for some very undersampled populations such as Afghan Pashtuns and Tajiks, Muslim Kashmiris, Northwestern Indian Brahmins, Punjabi Khatris, Hindu Sindhis, Rajasthani Rajputs, etc. I mean I had to convince about 10-11 individuals myself to join HAP from 23andMe after messaging probably around 30+.

Sein
10-24-2013, 06:43 AM
But that is really cool, you've made a big contribution to South Asian genetics. You also have quite a success rate, I've tried to get a lot of people into HAP, but I guess I'm not persuasive.

parasar
10-25-2013, 03:29 AM
Oh and she only had 2.5% Neanderthal on 23andMe which I thought was on the low side for someone with so much Caucasian/Euro admixture.

Just to muddy the water a bit on the Neanderthal admixture:
"Higher Levels of Neanderthal Ancestry in East Asians than in Europeans"

We find that, consistent with the recent finding of Meyer et al. (2012), Neanderthals contributed more DNA to modern East Asians than to modern Europeans. Furthermore we find that the Maasai of East Africa have a small but significant fraction of Neanderthal DNA. Because our analysis is of several genomic samples from each modern human population considered, we are able to document the extent of variation in Neanderthal ancestry within and among populations. Our results combined with those previously published show that a more complex model of admixture between Neanderthals and modern humans is necessary to account for the different levels of Neanderthal ancestry among human populations. In particular, at least some Neanderthal–modern human admixture must postdate the separation of the ancestors of modern European and modern East Asian populations.
http://www.genetics.org/content/194/1/199

Dr_McNinja
10-25-2013, 01:38 PM
Looking at the numbers of the Nepalese Brahmin and my friend, it looks like there's a correlation between Fennoscandian and the other Western European components. Would Finnish count as Western European? Since North Sea is basically descended from old Scandinavian I think.

Dr_McNinja
10-26-2013, 01:31 PM
Her old EUTest:


Population
SOUTH_BALTIC 2.48%
EAST_EURO 14.89%
NORTH-CENTRAL_EURO 2.47%
ATLANTIC 4.21%
WEST_MED -
EAST_MED 16.04%
WEST_ASIAN 30.67%
MIDDLE_EASTERN -
SOUTH_ASIAN 26.57%
EAST_AFRICAN 0.14%
EAST_ASIAN -
SIBERIAN 1.92%
WEST_AFRICAN 0.63%Also, when did the Oracle get the 3D/2D PCA charts? The 3D one is pretty cool!

Dr_McNinja
10-28-2013, 06:54 PM
Hey guys, I got the raw data from my mom's 23andMe test. Ran it through GEDmatch's admixture:


HarappaWorld

S-Indian 34.94%
Baloch 37.30%
Caucasian 11.38%
NE-Euro 10.51%
SE-Asian -
Siberian 0.97%
NE-Asian -
Papuan 0.81%
American 1.31%
Beringian 1.06%
Mediterranean 0.36%
SW-Asian 1.36%
San -
E-African -
Pygmy -
W-African -

Eurogenes K36

Amerindian 0.61%
Arabian -
Armenian -
Basque -
Central_African -
Central_Euro -
East_African -
East_Asian -
East_Balkan 1.36%
East_Central_Asian -
East_Central_Euro 1.22%
East_Med -
Eastern_Euro 0.23%
Fennoscandian 1.19%
French 1.53%
Iberian -
Indo-Chinese -
Italian -
Malayan -
Near_Eastern 0.84%
North_African -
North_Atlantic -
North_Caucasian 8.11%
North_Sea 4.07%
Northeast_African -
Oceanian 0.48%
Omotic -
Pygmy -
Siberian -
South_Asian 39.13%
South_Central_Asian 39.36%
South_Chinese -
Volga-Ural 0.76%
West_African 0.14%
West_Caucasian 0.98%
West_Med -

Eurogenes V2 K15

North_Sea 9.30%
Atlantic -
Baltic 0.67%
Eastern_Euro 9.27%
West_Med -
West_Asian 21.50%
East_Med 4.86%
Red_Sea -
South_Asian 51.68%
Southeast_Asian -
Siberian -
Amerindian 1.86%
Oceanian 0.20%
Northeast_African -
Sub-Saharan 0.64%
I'm not sure how to use or interpret GEDmatch's phasing tool. I put in our kit numbers and it generated two kits. When I ran the generated maternal kit, it was different from my mom's actual kit. This was the generated maternal kit's Harappa numbers:


GEDmatch generated maternal kit (from child and maternal kits)

Harappa

S-Indian 38.77%
Baloch 39.93%
Caucasian 9.53%
NE-Euro 9.25%
SE-Asian -
Siberian 0.45%
NE-Asian -
Papuan -
American 0.78%
Beringian 1.29%
Mediterranean -
SW-Asian -
San -
E-African -
Pygmy -
W-African -
Why is it different?

I ran the generated paternal kit through the admixture as well:


Harappa

S-Indian 34.87%
Baloch 30.57%
Caucasian 10.71%
NE-Euro 9.77%
SE-Asian 0.36%
Siberian 2.07%
NE-Asian 0.39%
Papuan 1.35%
American 1.02%
Beringian 1.87%
Mediterranean 2.68%
SW-Asian 1.42%
San -
E-African 1.00%
Pygmy 1.18%
W-African 0.71%
I know this is probably not accurate but is it possible for me to score higher or lower than them on an admixture component and not inbetween as is turning out with the phased data?

More:


Eurogenes K36

Amerindian 0.69%
Arabian -
Armenian 1.34%
Basque 1.42%
Central_African 0.32%
Central_Euro -
East_African -
East_Asian -
East_Balkan -
East_Central_Asian 1.18%
East_Central_Euro 2.82%
East_Med -
Eastern_Euro 0.05%
Fennoscandian 1.86%
French -
Iberian -
Indo-Chinese -
Italian -
Malayan -
Near_Eastern -
North_African -
North_Atlantic 0.46%
North_Caucasian 1.93%
North_Sea 4.92%
Northeast_African -
Oceanian 1.28%
Omotic 0.85%
Pygmy 0.10%
Siberian 3.07%
South_Asian 40.57%
South_Central_Asian 29.65%
South_Chinese -
Volga-Ural -
West_African 3.06%
West_Caucasian 4.42%
West_Med -




Eurogenes V2 K15

North_Sea 0.92%
Atlantic 7.01%
Baltic 3.29%
Eastern_Euro 5.88%
West_Med 0.79%
West_Asian 19.18%
East_Med 0.99%
Red_Sea 1.27%
South_Asian 49.01%
Southeast_Asian 0.61%
Siberian 2.97%
Amerindian 2.16%
Oceanian 1.11%
Northeast_African 0.28%
Sub-Saharan 4.53%
I thought the French in my mom's Eurogenes K36 and the North Sea heavy K15 was interesting (McDonald plotted France for my European component). My dad was almost completely Atlantic instead, and even got Basque on the K36. My numbers were half and half between North Sea and Atlantic. My father's father was from India (Hoshiarpur) and his mother was my mother's father's cousin so I figure he got these Brahmin-like results from his father. He also had some African, Siberian, and other tribal type results which my mom didn't have. He also had virtually no North Caucasian and just some West Caucasian. My mother's results seem more typically Pakistani Jatt but still with a North Sea heavy European admixture rather than Eastern or Central Europe.

Dr_McNinja
10-28-2013, 07:03 PM
Mom's Harappa Oracle results:


Least-squares method.

Using 1 population approximation:
1 punjabi-brahmin @ 3.636
2 punjabi @ 3.980
3 kashmiri-pandit @ 4.090
4 singapore-indian-c @ 4.727
5 up-muslim @ 5.449
6 punjabi-jatt-muslim @ 5.665
7 kashmiri-pahari @ 5.882
8 pushtikar-brahmin @ 5.937
9 punjabi-ramgarhia @ 6.018
10 rajasthani-brahmin @ 6.429
377 iterations.



Using 2 populations approximation:
1 50% punjabi-khatri +50% up-brahmin @ 2.117
2 50% brahmin-uttar-pradesh +50% punjabi-khatri @ 2.420
3 50% kashmiri +50% up-brahmin @ 2.493
4 50% kalash +50% srivastava @ 2.713
5 50% brahmin-uttar-pradesh +50% sindhi @ 2.766
6 50% punjabi-brahmin +50% up-muslim @ 2.829
7 50% pathan +50% vaish @ 3.019
8 50% bihari-brahmin +50% punjabi-khatri @ 3.030
9 50% bihari-muslim +50% punjabi-khatri @ 3.061
10 50% karnataka-brahmin +50% pathan @ 3.075
71253 iterations.



Using 3 populations approximation:
1 50% punjabi-khatri +25% punjabi-khatri +25% srivastava @ 1.299
2 50% punjabi-khatri +25% kashmiri-pandit +25% srivastava @ 1.385
3 50% punjabi-khatri +25% punjabi-brahmin +25% srivastava @ 1.487
4 50% punjabi-khatri +25% kanjar +25% punjabi-khatri @ 1.554
5 50% punjabi-khatri +25% haryana-jatt +25% velama @ 1.567
6 50% punjabi-khatri +25% punjabi-jatt-muslim +25% srivastava @ 1.571
7 50% punjabi-khatri +25% punjabi +25% srivastava @ 1.609
8 50% punjabi-khatri +25% punjabi-jatt-sikh +25% sourastrian @ 1.622
9 50% punjabi-khatri +25% haryana-jatt +25% meghawal @ 1.624
10 50% punjabi-khatri +25% punjabi-jatt-sikh +25% srivastava @ 1.626
3751850 iterations.



Using 4 populations approximation:
1 punjabi-khatri + punjabi-khatri + punjabi-khatri + srivastava @ 1.299
2 haryana-jatt + kalash + naidu + punjabi-khatri @ 1.385
3 kashmiri-pandit + punjabi-khatri + punjabi-khatri + srivastava @ 1.385
4 haryana-jatt + kalash + piramalai-kallar + punjabi-khatri @ 1.411
5 punjabi-brahmin + punjabi-khatri + punjabi-khatri + srivastava @ 1.487
6 haryana-jatt + kalash + kurumba + punjabi-khatri @ 1.552
7 kanjar + punjabi-khatri + punjabi-khatri + punjabi-khatri @ 1.554
8 haryana-jatt + kalash + punjabi-khatri + singapore-indian-a @ 1.560
9 haryana-jatt + punjabi-khatri + punjabi-khatri + velama @ 1.567
10 punjabi-jatt-muslim + punjabi-khatri + punjabi-khatri + srivastava @ 1.571
11 punjabi + punjabi-khatri + punjabi-khatri + srivastava @ 1.609
12 haryana-jatt + kalash + punjabi-khatri + vysya @ 1.615
13 punjabi-jatt-sikh + punjabi-khatri + punjabi-khatri + sourastrian @ 1.622
14 haryana-jatt + meghawal + punjabi-khatri + punjabi-khatri @ 1.624
15 punjabi-jatt-sikh + punjabi-khatri + punjabi-khatri + srivastava @ 1.626
16 punjabi-khatri + punjabi-khatri + rajasthani + rajasthani-brahmin @ 1.626
17 meghawal + punjabi-khatri + punjabi-khatri + rajasthani-brahmin @ 1.628
18 kalash + punjabi-khatri + srivastava + up-brahmin @ 1.632
19 punjabi-khatri + punjabi-khatri + punjabi-khatri + up-muslim @ 1.638
20 kerala-christian + punjabi-khatri + punjabi-khatri + rajasthani-brahmin @ 1.656

22839401 iterations.

Done.

Elapsed time 19.7523 seconds.Phased paternal kit's Harappa Oracle results:
Least-squares method.

Using 1 population approximation:
1 up-muslim @ 4.683
2 nepali @ 6.603
3 pushtikar-brahmin @ 7.498
4 punjabi @ 9.079
5 gujarati-muslim @ 9.097
6 punjabi-jatt-muslim @ 9.410
7 up-brahmin @ 9.479
8 kashmiri-pandit @ 9.540
9 kashmiri @ 9.727
10 rajasthani-brahmin @ 9.909
377 iterations.



Using 2 populations approximation:
1 50% caribbean-indian +50% pashtun @ 3.789
2 50% pashtun +50% sri-lankan @ 3.913
3 50% pashtun +50% srivastava @ 4.025
4 50% pashtun +50% sourastrian @ 4.379
5 50% pashtun +50% up-muslim @ 4.399
6 50% bihari +50% pashtun @ 4.429
7 50% cochin-jew +50% haryana-jatt @ 4.560
8 50% kanjar +50% pashtun @ 4.600
9 50% up-muslim +50% up-muslim @ 4.683
10 50% dharkar +50% pashtun @ 4.691
71253 iterations.



Using 3 populations approximation:
1 50% haryana-jatt +25% bene-israel-jew +25% south-african-indian @ 2.611
2 50% up +25% oriya +25% stalskoe @ 2.938
3 50% vaish +25% south-african-indian +25% stalskoe @ 2.950
4 50% haryana-jatt +25% bene-israel-jew +25% tharu @ 2.960
5 50% haryana-jatt +25% kurumba +25% pashtun @ 3.072
6 50% up +25% stalskoe +25% vaish @ 3.080
7 50% punjabi-jatt-sikh +25% malayan +25% tajik @ 3.155
8 50% haryana-jatt +25% bene-israel-jew +25% sourastrian @ 3.156
9 50% haryana-jatt +25% bene-israel-jew +25% caribbean-indian @ 3.190
10 50% oriya +25% north-ossetian +25% up @ 3.191
4847423 iterations.



Using 4 populations approximation:
1 bene-israel-jew + haryana-jatt + haryana-jatt + south-african-indian @ 2.611
2 haryana-jatt + north-kannadi + pashtun + pushtikar-brahmin @ 2.694
3 haryana-jatt + kurumba + pashtun + pushtikar-brahmin @ 2.738
4 hakkipikki + haryana-jatt + pashtun + pushtikar-brahmin @ 2.833
5 bihari-brahmin + oriya + south-african-indian + stalskoe @ 2.872
6 bene-israel-jew + haryana-jatt + nepali + srivastava @ 2.885
7 oriya + south-african-indian + stalskoe + vaish @ 2.897
8 oriya + south-african-indian + stalskoe + up-brahmin @ 2.911
9 oriya + stalskoe + up + up @ 2.938
10 south-african-indian + stalskoe + vaish + vaish @ 2.950
11 bhil + haryana-jatt + pashtun + pushtikar-brahmin @ 2.954
12 bene-israel-jew + haryana-jatt + haryana-jatt + tharu @ 2.960
13 chenchu + haryana-jatt + pashtun + punjabi-jatt-muslim @ 2.974
14 brahmin-uttar-pradesh + oriya + south-african-indian + stalskoe @ 2.976
15 haryana-jatt + malayan + pashtun + pathan @ 2.999
16 haryana-jatt + malayan + sindhi + tajik @ 3.013
17 haryana-jatt + kurumba + pashtun + punjabi-jatt-sikh @ 3.026
18 haryana-jatt + kalash + malayan + pashtun @ 3.045
19 south-african-indian + stalskoe + up-kshatriya + vaish @ 3.045
20 chenchu + haryana-jatt + pashtun + pushtikar-brahmin @ 3.053

31822251 iterations.

Done.

Elapsed time 26.0322 seconds.

Dr_McNinja
10-28-2013, 07:13 PM
Mom's Eurogenes K12b:


Western European 3.31%
Siberian 2.06%
East African -
West Central Asian 29.62%
South Asian 41.49%
West African 0.41%
Caucasus 11.94%
Finnish 2.99%
Mediterranean -
Southwest Asian 0.33%
North European 7.85%
East Asian - Phased paternal kit's Eurogenes K12b:

Western European 7.65%
Siberian 4.59%
East African 0.26%
West Central Asian 23.37%
South Asian 40.63%
West African 4.07%
Caucasus 9.97%
Finnish 1.04%
Mediterranean 0.69%
Southwest Asian 0.09%
North European 7.22%
East Asian 0.42% Afghan Popalzai/Durrani Eurogenes K12b:
Western European 6.03%
Siberian 3.07%
East African 1.32%
West Central Asian 32.37%
South Asian 15.73%
West African -
Caucasus 22.09%
Finnish 2.79%
Mediterranean 3.50%
Southwest Asian 3.07%
North European 10.01%
East Asian -


Mom's old EUTest:
SOUTH_BALTIC 1.44%
EAST_EURO 11.29%
NORTH-CENTRAL_EURO 5.20%
ATLANTIC -
WEST_MED -
EAST_MED 6.75%
WEST_ASIAN 19.93%
MIDDLE_EASTERN -
SOUTH_ASIAN 54.02%
EAST_AFRICAN -
EAST_ASIAN -
SIBERIAN 1.25%
WEST_AFRICAN 0.12%


Phased paternal kit's old EUTest:
SOUTH_BALTIC 1.32%
EAST_EURO 8.92%
NORTH-CENTRAL_EURO 3.26%
ATLANTIC 6.09%
WEST_MED -
EAST_MED 3.96%
WEST_ASIAN 17.10%
MIDDLE_EASTERN -
SOUTH_ASIAN 51.09%
EAST_AFRICAN 0.98%
EAST_ASIAN -
SIBERIAN 4.05%
WEST_AFRICAN 3.22%


Should I send my mom's data in to Zack?

Dr_McNinja
10-28-2013, 07:36 PM
McDonald map plots for the Afghan Pashtun Popalzai/Durrani:

http://i.imgur.com/HmbG5Uq.png

http://i.imgur.com/NCXbHOv.png


Most likely fit is 33.9% (+- 7.5%) Mideast (all Caucasus Area) and 66.1% (+- 7.5%) S. Asia (all Pakistan)




The following are possible population sets and their fractions, most likely at the top

Adygei= 0.441 Sindhi= 0.559 or

Adygei= 0.313 Pathan= 0.687 or

Georgian= 0.264 Pathan= 0.736



a custom fit gives results with other different Mideast and European populations, but gives the identical spot on the map. So it is just trying too hard for a perfect fit. This clearly has a larger “west of Afghanistan” component, but it is of unclear nature. The mideast plus Europe together appears eastern Mediterranean, BUT the cut between it and S. Asia is very murky and the exact nature changes as the cut is moved east and west.

Dr_McNinja
10-28-2013, 07:40 PM
Mom's 23andMe One-to-Many results:


Me 85.34%
Sapporo 75.58% (I had 75.95%)
Sein 75.34% (I had 75.71%)

Sein
10-28-2013, 08:29 PM
Your mother has very interesting results! Also, could you post the PCA plots for the Durrani Pashtun individual? Thank you.

Sein
10-28-2013, 08:32 PM
HRP0286's results are very similar, but slightly less South Asian.

Adygei= 0.343 Pathan= 0.657 or

Adygei= 0.465 Sindhi= 0.535 or

Jewish= 0.228 Pathan= 0.772 or

Turkish= 0.299 Pathan= 0.701

Dr_McNinja
10-29-2013, 02:22 PM
Ok, I need a day or two to get internet up at home

Dr_McNinja
10-31-2013, 12:01 AM
Here are the rest of the McDonald BGA plots: http://imgur.com/a/13e6a#0

Dr_McNinja
10-31-2013, 12:35 AM
Sapporo, did you see, Zack added some Rajasthani Jatts to the database. Do you know them or what their results were?

Also, I had Zack include my mother and not myself in the dataset. She's a better representative of Punjabi Jatt Muslims. My dad is believe it or not only half out of my mom's family and still his phased results seem completely North Indian Brahmin or like Haryana Jatt mixed with some tribal. His father must have had some really interesting data.

Sapporo
10-31-2013, 01:05 PM
Sapporo, did you see, Zack added some Rajasthani Jatts to the database. Do you know them or what their results were?

Also, I had Zack include my mother and not myself in the dataset. She's a better representative of Punjabi Jatt Muslims. My dad is believe it or not only half out of my mom's family and still his phased results seem completely North Indian Brahmin or like Haryana Jatt mixed with some tribal. His father must have had some really interesting data.

Yeah, I just noticed now in the ethnicity database. HRP0359 and HRP360. Both of them are the Rajasthani Jatts I'm sharing with on 23andMe. HRP0359 should have similar results to some of the Haryanvi Jatts but HRP0360 will have very bizarre results in comparison to all other Jatts. His position on the 23andMe world PCA plot is quite unexpected as he is clustering with HRP0286. There also seems to be a new Punjabi Jatt Sikh in the project as well as another Punjabi Brahmin. I'm curious about the two Nepali Brahmin as well. I'll also be watching out for the project's first Punjabi Gujjar (HRP0353).

Also, HRP0363 is actually half Baloch (Khetran) and half Durrani Pashtun even though he is labeled as just Baloch.

Sein
10-31-2013, 04:17 PM
Hi Mac,

HRP0363 is actually fully Khetran Balochi.

Sapporo
10-31-2013, 09:47 PM
@Sein

Really? He had told me he was half Pashtun and he was using Durrani as a surname. Anyways, thanks for the info. Seems I was incorrect.

Sein
10-31-2013, 10:11 PM
You know, this confused me as well. Apparently, Khetran Balochis claim Pashtun descent. They say they are descended from the ancestor of all Durranis. The British anthropologists who first came to the region claimed that they are a mixed people, intermediate between Pashtuns and Balochis. There is also a theory that they are a three-way mix between Baloch, Pashtun, and Seraiki people. The language is pretty interesting. Khetrani has been heavily influenced by both Balochi and Pashto. But it is debated whether it is a Indo-Aryan language, or a Dardic language. Pretty interesting, because I would never imagine a Dardic language this far south. Despite their ambiguous origins, and distinctive language, they are considered Balochi by neighboring Baloch tribes, and are a part of that people's tribal system.

Dr_McNinja
11-01-2013, 03:27 PM
Yeah, I just noticed now in the ethnicity database. HRP0359 and HRP360. Both of them are the Rajasthani Jatts I'm sharing with on 23andMe. HRP0359 should have similar results to some of the Haryanvi Jatts but HRP0360 will have very bizarre results in comparison to all other Jatts. His position on the 23andMe world PCA plot is quite unexpected as he is clustering with HRP0286. There also seems to be a new Punjabi Jatt Sikh in the project as well as another Punjabi Brahmin. I'm curious about the two Nepali Brahmin as well. I'll also be watching out for the project's first Punjabi Gujjar (HRP0353).

Also, HRP0363 is actually half Baloch (Khetran) and half Durrani Pashtun even though he is labeled as just Baloch.What percent similarity did you have with them on 23andMe?

Dr_McNinja
11-01-2013, 03:28 PM
My mom got 2.8% Neanderthal. I had 2.7%

Dr_McNinja
11-01-2013, 09:28 PM
Zack added the new members to the spreadsheet:

http://www.harappadna.org/2013/11/harappaworld-hrp0352-hrp0374/

Sein
11-01-2013, 09:45 PM
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. 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.

Dr_McNinja
11-01-2013, 10:11 PM
Does this imply that Punjabi Jats have been less endogamous?I would assume so just because the further east you go, the more possible it was to remain endogamous. Punjab was subject to numerous invasions and migrations in the past 2000 years. The eastern boundary of Punjab was where many of the Indians managed to hold back the waves.

The higher Mediterranean numbers for both Rajasthani Jatts make me wish we could see their Eurogenes K36/K15/K12b results. Their numbers are so close, I wonder if they're related somehow? Sapporo, do you know?

Dr_McNinja
11-01-2013, 11:07 PM
There's a pretty stark difference between my friend and my mother's DIY results on GEDmatch and the ones in the spreadsheet. My mom went from 0% in San/E African/Pygmy up to around ~0.25% each out of nowhere. HRP0370 gained almost half a percent in each.

Sein
11-01-2013, 11:08 PM
I would assume so just because the further east you go, the more possible it was to remain endogamous. Punjab was subject to numerous invasions and migrations in the past 2000 years. The eastern boundary of Punjab was where many of the Indians managed to hold back the waves.

Very interesting.

Sein
11-01-2013, 11:24 PM
There's a pretty stark difference between my friend and my mother's DIY results on GEDmatch and the ones in the spreadsheet. My mom went from 0% in San/E African/Pygmy up to around ~0.25% each out of nowhere. HRP0370 gained almost half a percent in each.

Yeah, official HarappaWorld results are much more noisy in comparison to the DIY results.

Sapporo
11-02-2013, 05:11 AM
What percent similarity did you have with them on 23andMe?

75.96% with HRP0359 and 75.86% with HRP360. If you want their clan names or info so you can share with them, please pm me.

Also, I have 2.9% Neanderthal ancestry which is 6th highest among my "friends" and 95th percentile for South Asian users.

Sapporo
11-02-2013, 05:37 AM
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.

Yes, he has very interesting results for an Iranian. I believe Bandari Iranians are from around the Persian Gulf coastal region. So, they are Southern Iranians but not necessarily Southeastern Iranians like Iranian Baloch.

It's surprising that is South Indian is higher than that one individual HGDP Pashtun at 12.6792%. The individual also has an elevated Baloch score and lowered Caucasian score versus the Behar and Dodecad Iranians. It definitely implicates the possibility that far Eastern Iranians may be South Asian shifted with the exception of the Northeast which I believe will be Central Asian Turkic shifted due to its significant Turkmen population.

However, this is all speculation for now without more sampling of Eastern Iranians excluding Iranian Baloch who should be fairly similar to their Pakistani counterparts.



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


As for the Rajasthani Jatts, they definitely do seem slightly closer to the Haryanvi Jatts rather than their Punjabi counterparts. Although, I think I should point out that one of the Haryanvi Jatts is actually half Haryanvi/half Western Uttar Pradesh Jatt. Not that this makes any difference but half his ancestry is Western Uttar Pradesh Jatt. Although, the sample sizes for Haryanvi and Rajasthani Jatts are far too small at the moment at 5 (well almost 4 in a way) and 2 respectively, I still think we do have some indication that Punjabi Sikh Jatts and Punjabi Muslim Jatts have been less endogamous than their Haryanvi and Rajasthani brethren. Part of this is likely due to the fact that Punjabi Jatts (whether Sikh or Muslim) have interacted substantially with non-Jatt Punjabis and then there comes the significance of religion with Hindu Jatts being more endogamous. Indian Punjabi Sikh Jatts have intermarried with non Jatts to an extent but I believe Pakistani Punjabi Muslim Jatts have done even more so and not only with other Punjabis but other non-Punjabi Muslims due to the more cosmopolitan nature of Islam in comparison to Hinduism and Sikhism.



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

I believe this is the case with 1 or 2 of the Nepali A's as well. Well most are quite similar to the Nepali Khas Brahmin, 1-2 are very East Asian shifted. I think individual family origins and admixture may account for this.



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. 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.

I think you're definitely spot on here. This individual is sort of a cross between a Pashtun and a Baloch with slightly greater affinity to the HGDP Pashtuns than Baloch based on his McDonald results.



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 think two things are at play here. Firstly, Afghan Tajiks are more South Asian shifted than Tajikistan Tajiks as well as being just as much East Eurasian. Secondly, the issue of inheritance. This individual may simply have inherited more East Eurasian ancestry from his Tajik ancestry than one would expect for only being 1/4 Tajik. The other possibility is that his Tajik grandparent was very East Eurasian shifted. However, we cannot rule out that his Durrani Pashtun ancestors had more East Eurasian than the average as well. I believe you noted that Southern Pashtuns are more East Eurasian shifted versus Northern Pashtuns who are more South Asian shifted.

Sein
11-02-2013, 07:31 AM
Thank you very much for your analysis Sapporo! I think you are on the mark with everything you've written.

Humanist
11-03-2013, 04:29 AM
Combining David's Eurogenes SPA points from his original run in 2012, and my results for FTDNA's Population Finder, to identify possible coordinates associated with my autosomal DNA:

Population Finder Results

http://i1096.photobucket.com/albums/g326/dok101/Faces/Humanist_Population_Finder_1a.jpg

http://i1096.photobucket.com/albums/g326/dok101/Faces/Humanist_Population_Finder_2.jpg



Eurogenes SPA coordinates


IRANIAN 35.6303756247 51.7904149029
DRUZE 30.1886412111 34.0404798293

After a bit of math, we get my rough point, point "C" : 32.5315672155, 42.1594496593

Point "D" is Babylon. During Sasanian times, this area was a part of the Asōrestān Province (AKA Del-e Īrānshahr).


http://i1096.photobucket.com/albums/g326/dok101/Faces/Humanist_Population_Finder_3.jpg

Dr_McNinja
11-06-2013, 09:37 PM
I've been collecting the Eurogenes and DIYHarappa results on a spreadsheet here:

https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdFMtRHVlZDBuQ3lMcjhxMDE4V3JoY lE&usp=sharing

If you guys have anyone else's results for me to add, post them here, or PM me them. I decided to include the DIY/GEDmatch Harappa calculator because its results are different from Zack's official one.

Looking for non-Europeans specifically (Asia/Mideast/Caucasus especially).

MitchellSince1893
11-08-2013, 10:55 PM
After much trial and error on the command prompt I finally got it to work.

I came out in Northern Bavaria near the Czech border between Bayreuth and Kulmbach. Based on other results it seems too far to the East.

For comparison, McDonald's analysis had me in the Netherlands. This seems more accurate based on what I know about my ancestry.

My father comes out just east of Schnaittenbach, Bavaria...about 45 miles SW of me.

My mother comes out 10 miles south of Ingolstadt, Bavaria...about 100 miles south of me.

vettor
11-09-2013, 10:05 PM
in eurogenes EUtest V2 K15 does East-Med represent the pontid area ( anatolia and rim of the black sea ) and west-asian the levant , mesopotamia and western persia?

here are my results

# Population Percent
1 North_Sea 13.36
2 Atlantic 24.78
3 Baltic 9.59
4 Eastern_Euro 7.71
5 West_Med 18.81
6 West_Asian 8.42
7 East_Med 13.91
8 Red_Sea 2.27
9 South_Asian 0.77
10 Southeast_Asian 0.00
11 Siberian 0.00
12 Amerindian 0.00
13 Oceanian 0.12
14 Northeast_African 0.00
15 Sub-Saharan 0.26


Pct. Calc. Option 2

0 Unable to determine 0.01%
1 North_Italian 49.82%
2 French_Basque 8.88%
3 Spanish_Andalucia 8.42%
4 Chuvash 7.55%
5 Adygei 6.98%
6 Ukrainian_Northeast 6.69%

Joe B
11-09-2013, 10:55 PM
in eurogenes EUtest V2 K15 does East-Med represent the pontid area ( anatolia and rim of the black sea ) and west-asian the levant , mesopotamia and western persia?
Looks like it covers that and more.
Eurogenes K15 now at GEDmatch
http://bga101.blogspot.com/2013/10/eurogenes-k15-now-at-gedmatch.html
K15 East-Med Map
http://img690.imageshack.us/img690/4471/ogk2.png
K15 West Asian
http://img822.imageshack.us/img822/5958/8awm.png

vettor
11-10-2013, 01:38 AM
Looks like it covers that and more.
Eurogenes K15 now at GEDmatch
http://bga101.blogspot.com/2013/10/eurogenes-k15-now-at-gedmatch.html
K15 East-Med Map
http://img690.imageshack.us/img690/4471/ogk2.png

i am confused, is this another eurogenes plotting system, because the one below for me is 2 months old using another eurogenes format of K9b

http://img20.imageshack.us/img20/2264/apob.png (http://imageshack.us/photo/my-images/20/apob.png/)

Uploaded with ImageShack.us (http://imageshack.us)

This is mine,

BTW , Mcdonald places me on the french and italian border near switzerland

Joe B
11-10-2013, 04:13 AM
i am confused, is this another eurogenes plotting system, because the one below for me is 2 months old using another eurogenes format of K9b

http://img20.imageshack.us/img20/2264/apob.png (http://imageshack.us/photo/my-images/20/apob.png/)

Uploaded with ImageShack.us (http://imageshack.us)

This is mine,

BTW , Mcdonald places me on the french and italian border near switzerland
I'm not sure about the plotting systems. Seems like a little voodoo is involved. Each different admixture test will move you around depending on the reference populations.
K9
South Asian - South India
Caucasus - Georgia
Southwest Asian - Bedouin
North Amerindian + Arctic - Northwest America
Siberian - Central Siberia
Mediterranean - Sardinia
East Asian - Eastern China
West African - Nigeria
North European - Lithuania
http://bga101.blogspot.com.au/2012/05/geographicethnic-peaks-in-eurogenes.html


Was Dr. McDonald close to what you expected with the Swiss-Italian-French border or should it be closer to Austria?
What admixture test works best for North Italy?



(nothing wrong with voodoo)

MitchellSince1893
11-10-2013, 05:04 AM
Here are my results for McDonald and SPA. Also shows my parents and the North and South Limits for the 3 of us.

My McDonald in Red. My SPA in Purple. Mother's SPA in Pink, Father's SPA in Blue.

For some reason the range for my mother and father is much larger than my own. My dad's ranges from just south of Norway to Southern Italy, while my mom's is from the North Sea to the Adriatic Sea.

I'm wondering if it's because I connected them to me via 23andme's family tree tool. I say this because my British and Irish ancestry composition almost doubled a few days ago from 13.6% to 26%, right after I linked my parents to me on 23andme.

889

vettor
11-10-2013, 06:23 AM
I'm not sure about the plotting systems. Seems like a little voodoo is involved. Each different admixture test will move you around depending on the reference populations.
K9
South Asian - South India
Caucasus - Georgia
Southwest Asian - Bedouin
North Amerindian + Arctic - Northwest America
Siberian - Central Siberia
Mediterranean - Sardinia
East Asian - Eastern China
West African - Nigeria
North European - Lithuania
http://bga101.blogspot.com.au/2012/05/geographicethnic-peaks-in-eurogenes.html


Was Dr. McDonald close to what you expected with the Swiss-Italian-French border or should it be closer to Austria?
What admixture test works best for North Italy?



(nothing wrong with voodoo)


the best admixture mix for me seems to be dodecad k12b , but Eutest and JTest from eurogenes are fine...........i do not go searching for what i would like to see, but for what is the correct truth. The V2 K15b .........is the first time i got thrown into the iberian mix......usually I get , italian, french, german, ocadian and bulgarian.

McDonalds results for me was kind of weird , he sent me the map of the france-italy-swiss area with the green, red and blue dots all together........then he states I originate in the eastern alps ( maybe he meant western alps). 100% european, ...........same as 23andme

MitchellSince1893
11-10-2013, 02:54 PM
...my British and Irish ancestry composition almost doubled a few days ago from 13.6% to 26%, right after I linked my parents to me on 23andme.

889

Just saw this note on my 23andme page
...Ancestry Composition results have been computed using both parents.

Joe B
11-10-2013, 05:46 PM
the best admixture mix for me seems to be dodecad k12b , but Eutest and JTest from eurogenes are fine...........i do not go searching for what i would like to see, but for what is the correct truth. The V2 K15b .........is the first time i got thrown into the iberian mix......usually I get , italian, french, german, ocadian and bulgarian.

McDonalds results for me was kind of weird , he sent me the map of the france-italy-swiss area with the green, red and blue dots all together........then he states I originate in the eastern alps ( maybe he meant western alps). 100% european, ...........same as 23andme
Please count me in the "let the results speak for themselves" crowd. Figuring out what these different admixure tests are trying to say about the results takes an interpreter. Some of these test do better on some populations than others. I'm trying to have a better look at one set of my ggg-grandparents that comes from NE Italy.
Appreciate what you had to say.

vettor
11-10-2013, 05:47 PM
@ JoeB

this is the eurogenes plotting link .......i referred to............where can i find this last one?

http://polishgenes.blogspot.com.au/2013/09/coming-soon-geographic-population.html

apparently , eurogenes k9b is what geno 2.0 use

Joe B
11-10-2013, 10:17 PM
@ JoeB

this is the eurogenes plotting link .......i referred to............where can i find this last one?

http://polishgenes.blogspot.com.au/2013/09/coming-soon-geographic-population.html

apparently , eurogenes k9b is what geno 2.0 use
@ Vettor,
Thanks for the link to the Polishgenes Blog.
Are you referring to the broken link for Geographic Population Structure (GPS) prediction or something else?

Noticed this with Gedmatch.

NOTICE: Due to the current processing backlog, we are temporarily suspending DNA uploads. We expect to start accepting new uploads on about December 1.

ZephyrousMandaru
11-12-2013, 08:54 AM
Combining David's Eurogenes SPA points from his original run in 2012, and my results for FTDNA's Population Finder, to identify possible coordinates associated with my autosomal DNA:

Population Finder Results

http://i1096.photobucket.com/albums/g326/dok101/Faces/Humanist_Population_Finder_1a.jpg

http://i1096.photobucket.com/albums/g326/dok101/Faces/Humanist_Population_Finder_2.jpg



Eurogenes SPA coordinates


IRANIAN 35.6303756247 51.7904149029
DRUZE 30.1886412111 34.0404798293

After a bit of math, we get my rough point, point "C" : 32.5315672155, 42.1594496593

Point "D" is Babylon. During Sasanian times, this area was a part of the Asōrestān Province (AKA Del-e Īrānshahr).


http://i1096.photobucket.com/albums/g326/dok101/Faces/Humanist_Population_Finder_3.jpg

That's very interesting, and peculiar at the same time. I wonder now, if we removed all the Armenian samples from the Dodecad K12b Oracle (Not the one on GED Match), what would your results be?

Humanist
11-12-2013, 02:29 PM
That's very interesting, and peculiar at the same time. I wonder now, if we removed all the Armenian samples from the Dodecad K12b Oracle (Not the one on GED Match), what would your results be?

Apologies, but I am not following you, Zeph. Do you think it would significantly change? I am 1/16th Armenian. My maternal grandfather was 1/4th Armenian.

ZephyrousMandaru
11-12-2013, 06:04 PM
Apologies, but I am not following you, Zeph. Do you think it would significantly change? I am 1/16th Armenian. My maternal grandfather was 1/4th Armenian.

On the Dodecad v1 Oracles. You could input your admixture values and it'll give you a list of top population approximations. You also have the option of removing certain samples. What I'm wondering is, if you removed Armenians for example. Would you end up with a similar result to your FTDNA population results?

Dr_McNinja
11-19-2013, 12:02 AM
My mom's 23andMe ancestry composition:

97.4% South Asian
< 0.1% European
< 0.1% Nonspecific European
< 0.1% Middle Eastern & North African
< 0.1% Middle Eastern
2.5% Unassigned


And in speculative mode:

98.4% South Asian
0.7% Middle Eastern & North African
0.7% Middle Eastern
0.6% European
0.6% Nonspecific European
0.3% Unassigned

Dr_McNinja
11-19-2013, 02:55 AM
My mom's 23andMe ancestry composition:

97.4% South Asian
< 0.1% European
< 0.1% Nonspecific European
< 0.1% Middle Eastern & North African
< 0.1% Middle Eastern
2.5% Unassigned


And in speculative mode:

98.4% South Asian
0.7% Middle Eastern & North African
0.7% Middle Eastern
0.6% European
0.6% Nonspecific European
0.3% UnassignedThey changed my Ancestry Composition results. It used to be 98.x% South Asian and in speculative mode it was 98.8 or 98.9% South Asian and 1.1% Non-Specific European.

Now it's:

96.6% South Asian
0.1% East Asian & Native American
East Asian
0.1% Nonspecific East Asian
0.1% European
0.1% Nonspecific European
3.1% Unassigned

and in Speculative mode:

98.7% South Asian
0.5% European
Northern European
0.1% Nonspecific Northern European
0.4% Nonspecific European
0.1% East Asian & Native American
East Asian
0.1% Nonspecific East Asian
0.7% Unassigned

My Chromosome 15 used to have a bunch of European marks (and in Gedmatch's one-to-many results, I get a lot of big chunks matching Europeans or Americans of European descent on Chr 15) and now it's all South Asian even though my mom's Chr 15 is still mostly European/Mideastern. Which is weird I guess.

Here's the split view:

http://i.imgur.com/GvVFlHL.png

Dr_McNinja
11-20-2013, 06:36 PM
In case you guys are curious here is an updated 23andMe comparison list for HRP0370, the Popalzai Pashtun, with the new people I added to my sharing list:

75.73% - Haryana Hindu Jatt (HRP0170)
75.70% - Half English, Half Jatt Sandhu
75.66% - Pakistan Punjabi Jatt Muslim (Sandhu)
75.65% - Pakistan Pashtun (HRP0282) (Sein)
75.64% - Punjabi Jatt Muslim (Gill) (HRP0349) (Me)
75.62% - Punjabi Jatt Sikh (Grewal/Mann/Gill) (HRP0341) (Sapporo)
75.61% - Rajasthani Hindu Jatt (HRP0359)
75.54% - Haryana Hindu Jatt (HRP0275 or HRP0276)
75.53% - Punjabi Brahmin
75.52% - Punjabi Jatt (Athwal)
75.51% - Rajasthani Hindu Jatt (HRP0360)
75.41% - Afghan Pashtun Mohammadzai (HRP0286)
75.38% - Punjabi Chamar (possibly, definitely Indian)
75.36% - Rajasthani (not Jatt probably)
75.31% - South Indian Karnatka Brahmin
75.29% - Punjabi Jatt Muslim (HRP0374) (My mom)
75.22% - Bengali
75.21% - South Asian Muslim
73.77% - Punjabi Brahmin

She only had 4th cousins on her relative finder but in her Countries of Ancestry page, a bunch of people showed up in random spots on her genome from Pakistan and Afghanistan, including Sein and HRP0286 (I believe I recognized his picture).

In addition, this new guy I added off my relative finder, the Athwal Jatt, has the following:

76.11% - Pakistan Punjabi Jatt Muslim (Sandhu)
75.91% - Rajasthani Hindu Jatt (HRP0359)
75.89% - Punjabi Brahmin
75.85% - Punjabi Jatt Sikh (Grewal/Mann/Gill) (HRP0341) (Sapporo)
75.82% - Haryana Hindu Jatt (HRP0170)
75.80% - Punjabi Chamar (possibly, definitely Indian)
75.79% - Punjabi Jatt Muslim (Gill) (HRP0349) (Me)
75.77% - Haryana Hindu Jatt (HRP0275 or HRP0276)
75.74% - Half English, Half Jatt Sandhu
75.72% - South Asian Muslim
75.69% - Pakistan Pashtun (HRP0282) (Sein)
75.68% - Rajasthani Hindu Jatt (HRP0360)
75.62% - South Indian Karnatka Brahmin
75.60% - Bengali
75.60% - Rajasthani (probably not Jatt)
75.53% - Punjabi Jatt Muslim (HRP0374) (Mom)
75.52% - Afghan Pashtun Popalzai/Durrani (HRP0370)
75.47% - Afghan Pashtun Mohammadzai (HRP0286)
74.10% - Punjabi Brahmin

It's weird/funny how this Pakistani guy, the Sandhu Jatt, is at the top of the match list for virtually everybody.

MfA
11-20-2013, 07:05 PM
Plotting at Corduene like rest of the Kurmanji Kurds(greens)

Dodecad K12b
http://abload.de/img/desktop_2013_11_20_21l3ski.jpg

Dr_McNinja
11-20-2013, 08:53 PM
Sein/Sapporo:

http://www.nytimes.com/2013/11/21/science/two-surprises-in-dna-of-boy-found-buried-in-siberia.html?_r=1&

http://www.nature.com/nature/journal/vaop/ncurrent/full/nature12736.html

Sein
11-21-2013, 09:20 AM
A lot of interesting stuff in this paper. If you look at the Admixture experiments, the Malta-individual's West Eurasian ancestry is equally split between the component modal in South India, and the component modal in Northeastern Europe. Not sure what this could mean. This genome is really ancient. Perhaps at this period in time, intra-West Eurasian differentiation was weak, which is why this individual's West Eurasian ancestry has an affinity to South Asians. Also, in the TreeMix experiment, one sees a migration line from the Han Chinese to the Indians. I guess the Han Chinese are a good proxy for "ASI".

This paper is pretty ground breaking. Europeans are a fusion between West Eurasians from the Middle East, and a Native American-like East Eurasian population, with 20%-40% admixture of that East Eurasian admixture. On the other hand, Native Americans are a fusion between a West Eurasian and a East Eurasian population, with 15%-40% West Eurasian ancestry.

Dr_McNinja
11-21-2013, 04:07 PM
A lot of interesting stuff in this paper. If you look at the Admixture experiments, the Malta-individual's West Eurasian ancestry is equally split between the component modal in South India, and the component modal in Northeastern Europe. Not sure what this could mean. This genome is really ancient. Perhaps at this period in time, intra-West Eurasian differentiation was weak, which is why this individual's West Eurasian ancestry has an affinity to South Asians. Also, in the TreeMix experiment, one sees a migration line from the Han Chinese to the Indians. I guess the Han Chinese are a good proxy for "ASI".

This paper is pretty ground breaking. Europeans are a fusion between West Eurasians from the Middle East, and a Native American-like East Eurasian population, with 20%-40% admixture of that East Eurasian admixture. On the other hand, Native Americans are a fusion between a West Eurasian and a East Eurasian population, with 15%-40% West Eurasian ancestry.Can you get a bigger version of this picture?

http://www.nature.com/nature/journal/vaop/ncurrent/carousel/nature12736-f2.jpg

Sein
11-21-2013, 08:37 PM
952

This is the one with 2 migration lines.

Sein
11-22-2013, 12:34 AM
These are the two "simple" trees, only a basic topology.
953

954

Since Tajiks are less West Eurasian than the HGDP Sindhis and HGDP Pashtuns, both of these populations will probably cluster in between the Tajik and Avar on the first tree.

955

Note the migration edge coming from the root of the East Eurasians to the Indian.

parasar
11-23-2013, 04:51 PM
...

Since Tajiks are less West Eurasian than the HGDP Sindhis and HGDP Pashtuns, both of these populations will probably cluster in between the Tajik and Avar on the first tree.
...

If you remove W. Asian and the even lower E. Asian from the Tajik, both likely historical influences, the Tajik and the Mal'ta boy (0 E & W Asian) would be very close.

Dr_McNinja
11-24-2013, 02:34 AM
on HRP0370's 23andMe profile, another Afghan Pashtun got in touch via relative finder and had a high match. A ghilzai Pashtun. This is the updated list:

75.75% - Afghan Pashtun Ghilzai Esmailkhel
75.73% - Haryana Hindu Jatt (HRP0170)
75.70% - Half English, Half Jatt Sandhu
75.66% - Pakistan Punjabi Jatt Muslim (Sandhu)
75.65% - Pakistan Pashtun (HRP0282) (Sein)
75.64% - Punjabi Jatt Muslim (Gill) (HRP0349) (Me)
75.62% - Punjabi Jatt Sikh (Grewal/Mann/Gill) (HRP0341) (Sapporo)
75.61% - Rajasthani Hindu Jatt (HRP0359)
75.54% - Haryana Hindu Jatt (HRP0275 or HRP0276)
75.53% - Punjabi Brahmin
75.52% - Punjabi Jatt (Athwal)
75.51% - Rajasthani Hindu Jatt (HRP0360)
75.41% - Afghan Pashtun Mohammadzai (HRP0286)
75.38% - Punjabi Chamar (possibly, definitely Indian)
75.36% - Rajasthani (not Jatt probably)
75.31% - South Indian Karnatka Brahmin
75.29% - Punjabi Jatt Muslim (HRP0374) (My mom)
75.22% - Bengali
75.21% - South Asian Muslim
73.77% - Punjabi Brahmin

(I was at 75.46% to this person)

Sein
11-24-2013, 02:58 AM
Could you share more about this individual?

Dr_McNinja
11-24-2013, 04:56 AM
Don't know anything yet other than the above. Will let you know if they respond. I did recommend they tell the new person to join HAP

parasar
11-24-2013, 05:43 AM
Where can we see the Admixture proportions for the Mal'ta boy?

Please see: Figure SI 6 pg 55 http://www.nature.com/nature/journal/vaop/ncurrent/extref/nature12736-s1.pdf

Sein
11-25-2013, 11:51 AM
The cool thing is that one doesn't even need to look at prehistoric DNA for distinctive and unique results. At K=3-4 in Admixture/Structure, Papuans tend to turn out a mixture of East Eurasians, West Eurasians, with a small touch of Sub-Saharan African. When asking the algorithm to give three or four ancestral populations, it simply makes them out to be a three/four-way mix of the inferred components. People like Australasians don't really fit in a tripartite scheme of Sub-Saharan African-West Eurasian-East Eurasian.

I think you might have something else in mind, the Denisovan genome did not turn out a mixture between Northern Europeans, West Africans, and East Asians. I've never really read or heard of anything along those lines. The Denisovan genome represents an "archaic" Homo sapiens population, the product of admixture between Neanderthals and a completely unknown population. At the end of the day, when you look close enough, everyone is a product of admixture. For example, it's perfectly accurate to describe West Eurasians as intermediate between Sub-Saharan Africans and East Eurasians.

The Malta individual's results are very interesting. If you look at the PCA plot, he gravitates towards the Pakistani samples, rather than Europeans. And his Admixture results do support this. The component modal in low-caste South Indians is the largest fraction of his ancestry. Not sure if this a "real" result, some sort of ancient connection between South Asians, and the population the made a contribution of over 50% autosomal ancestry to Northern Europeans. This is possible. Northern Europeans and South Asians have some sort of deep and vague genetic link, one that appears consistently throughout various analyses. Some people attribute this to "Scythian" admixture, but perhaps this complicates things. Perhaps more interesting, and significantly more ancient, phenomena and events explain the South Asian-Northern European connection? My knowledge of haplogroups is abysmal (perhaps Parasar can chime in), but I was under the impression that R is supposed to have a South Asian origin? If so, that must mean something. If I'm wrong about this, please forgive me, I don't really follow the uniparental genetics literature (I mean, lets be honest, your paternal or maternal haplogroup represents only 0.1% of your ancestors 12 generations ago!).

Then again, maybe this not "real"? Perhaps this reflects huge amounts of genetic drift throughout the centuries. That is also completely possible and reasonable. I hope this aspect is further researched.

Dr_McNinja
11-26-2013, 04:43 AM
Updated list for HRP0370:


75.80% - Half Pakistani Punjabi, Half Pakistani Pathan/Pashtun (Female)
75.75% - Afghan Pashtun Ghilzai Esmailkhel
75.73% - Haryana Hindu Jatt (HRP0170)
75.70% - Half English, Half Jatt (Sandhu)
75.66% - Pakistan Punjabi Jatt Muslim (Sandhu)
75.65% - Pakistan Pashtun (HRP0282) (Sein)
75.64% - Punjabi Jatt Muslim (Gill) (HRP0349) (Me)
75.62% - Punjabi Jatt Sikh (Grewal/Mann/Gill) (HRP0341) (Sapporo)
75.61% - Rajasthani Hindu Jatt (HRP0359)
75.59% - Half British, Half Sikh Jatt (Male)
75.55% - Punjabi Chaudhry/Mian
75.54% - Haryana Hindu Jatt (HRP0275 or HRP0276)
75.53% - Punjabi Brahmin
75.52% - Punjabi Jatt (Athwal)
75.52% - Khan/Awan (Punjabi Awan?)
75.51% - Rajasthani Hindu Jatt (HRP0360)
75.48% - (Some kind of relative of top match @ 75.80%)
75.46% - Half Scandinavian, Half Moroccan+Algerian+South Asian
75.43% - European
75.42% - Punjabi Sikh Jatt (Gill)
75.41% - Afghan Pashtun Mohammadzai (HRP0286)
75.38% - Half British, Half Sikh Jatt (Female)
75.38% - Punjabi Chamar (possibly, definitely Indian)
75.38% - Kashmiri Butt, Muslim
75.36% - Rajasthani (not Jatt probably)
75.31% - South Indian Karnatka Brahmin
75.29% - Punjabi Jatt Muslim (HRP0374) (My mom)
75.24% - Pakistani/Indian from Bukhara, Tajikstan in 18th century
75.22% - Bengali
75.21% - South Asian Muslim
74.74% - (Relative of female @ 75.80%)
73.77% - Punjabi BrahminSein I'm pretty sure you have this new addition (the Punjabi/Pathan woman) on your list already. I got 3 profiles from them added at once. Do you know their relation to one another? You can tell me in PM on 23andMe or here. Not sure what to label the other two.

Sein
11-26-2013, 05:10 AM
I sent a PM my friend. The interesting thing about the individual in question is that she tends to match very high with most Pashtuns. She is also my highest match.

Sein
11-26-2013, 05:39 AM
The individual in question is the highest match for four Afghan Pashtuns on my sharing list. Also, HRP0370 is an Afghan Pashtun, and this half Punjabi-half Pashtun individual is her highest match. The two Pakistani Pashtuns I'm sharing with don't have this individual as their highest match. Not sure why she matches so high with so many diverse Afghan Pashtuns, and with myself.

Also, isolation-by-distance is a very robust force in population genetics, Pashtuns who live closer to Punjabis will clearly have greater affinities to Punjabi people than Herati Pashtuns. Nothing that I'd really contest here. Pashtuns aren't exactly endogamous when it comes to inter-ethnic marriage. Pashtuns freely take Punjabi, Tajik, Uzbek, Hazara, Muhajir, Sindhi, and Balochi wives. And the children of such marriages are always regarded as Pashtuns, since Pashtun ethnic identity is patrilineally transmitted. Traditionally, being a Pashtun simply involves having a Pashtun father, speaking the Pashto language, and adhering to Pashtowali, an ancient pre-Islamic tribal code. Most people still take the first two conditions seriously, although few if any urban Pashtuns in Afghanistan or Pakistan still adhere to Pashtowali. Only rural Pashtuns in the tribal belt stretching across the Afghanistan and Pakistan border take Pashtowali very seriously. Pashtowali is a remnant of a much darker and violent time, and absolutely impossible in a contemporary setting. But people in FATA and rural KP-Afghanistan still hold these traditions and rules in great respect, and practice them freely. Pashtowali involves seeking and enacting blood revenge at all costs, answering every personal injury in like terms, actively seeking and maintaining agricultural land or holdings, defending agricultural land or holdings, providing unconditional hospitality to guests, providing sanctuary to refugees of every stripe and any character, and never under any circumstances killing a woman, child, or person of Hindu faith.

Nevertheless, it must be remembered that Pashtun expansion has always involved a push west. Most Afghan Pashtuns claim origins far farther to the east in contemporary Pakistan. In short, "Pashtunization" almost always means demographic expansion into Farsiwan-Tajik areas. This complicates things rather deeply, since it could mean that less South Asian-shifted Afghan Pashtuns are admixed with Farsiwan, rather than more eastern Pashtuns being admixed with people east of the Indus. It's all very interesting, and I hope we can see more genetic research on this region.

parasar
11-26-2013, 06:23 PM
QUOTE=Sein;20877]... Pashtowali is a remnant of a much darker and violent time, and absolutely impossible in a contemporary setting. But people in FATA and rural KP-Afghanistan still hold these traditions and rules in great respect, and practice them freely. Pashtowali involves seeking and enacting blood revenge at all costs, answering every personal injury in like terms, actively seeking and maintaining agricultural land or holdings, defending agricultural land or holdings, providing unconditional hospitality to guests, providing sanctuary to refugees of every stripe and any character, and never under any circumstances killing a woman, child, or person of Hindu faith.

...[/QUOTE]

Very interesting. Is this injunction against killing Hindus applies only to killing Brahmans and Shramans or all Hindus?
I had read something similar about the Kafirs, but did not know of this Pakhtoon tradition.

"When the "chotee" is found on the head of a stranger, we treat him with courtesy, but if he do not wear it we kill him.http://books.google.com/books?id=EeQLAAAAIAAJ&pg=PA7&lpg=PA7&dq=chotee+kafir&source=bl&ots=InQiK_mCY0&sig=wGhVnn0epjzcWRmDms5GEh3SdOU&hl=en&sa=X&ei=gDWVUqb-I9POyAH1kIG4Bg&ved=0CB8Q6AEwAA

soulblighter
11-26-2013, 07:06 PM
The cool thing is that one doesn't even need to look at prehistoric DNA for distinctive and unique results. At K=3-4 in Admixture/Structure, Papuans tend to turn out a mixture of East Eurasians, West Eurasians, with a small touch of Sub-Saharan African. When asking the algorithm to give three or four ancestral populations, it simply makes them out to be a three/four-way mix of the inferred components. People like Australasians don't really fit in a tripartite scheme of Sub-Saharan African-West Eurasian-East Eurasian.

I think you might have something else in mind, the Denisovan genome did not turn out a mixture between Northern Europeans, West Africans, and East Asians. I've never really read or heard of anything along those lines. The Denisovan genome represents an "archaic" Homo sapiens population, the product of admixture between Neanderthals and a completely unknown population. At the end of the day, when you look close enough, everyone is a product of admixture. For example, it's perfectly accurate to describe West Eurasians as intermediate between Sub-Saharan Africans and East Eurasians.

The Malta individual's results are very interesting. If you look at the PCA plot, he gravitates towards the Pakistani samples, rather than Europeans. And his Admixture results do support this. The component modal in low-caste South Indians is the largest fraction of his ancestry. Not sure if this a "real" result, some sort of ancient connection between South Asians, and the population the made a contribution of over 50% autosomal ancestry to Northern Europeans. This is possible. Northern Europeans and South Asians have some sort of deep and vague genetic link, one that appears consistently throughout various analyses. Some people attribute this to "Scythian" admixture, but perhaps this complicates things. Perhaps more interesting, and significantly more ancient, phenomena and events explain the South Asian-Northern European connection? My knowledge of haplogroups is abysmal (perhaps Parasar can chime in), but I was under the impression that R is supposed to have a South Asian origin? If so, that must mean something. If I'm wrong about this, please forgive me, I don't really follow the uniparental genetics literature (I mean, lets be honest, your paternal or maternal haplogroup represents only 0.1% of your ancestors 12 generations ago!).

Then again, maybe this not "real"? Perhaps this reflects huge amounts of genetic drift throughout the centuries. That is also completely possible and reasonable. I hope this aspect is further researched.

There have been multiple migrations out of and into south Asia (and India).
If you saw the latest paper by Greg Magoon et. al, they have found a new haplogroup in a Telugu man, that indicates south India has a deep connection with the y-DNA NO* branch.
This is an amazing find (as big as A00).
I think all of these simplistic Aryan invasion and Out of India theory need to be thrown out the window, and some real work needs to be done.
If R is in itself 240000 years old in Siberia, I wonder how old all the other haplogroups are....
I wish South Asians would pool together and contribute towards Genetic testing of people from the Subcontinent.....but I doubt this will happen anytime soon.
We are missing a huge chunk of useful data from the region. Indians living outside India do not even represent 0.1% of its genetic variation ( as most tend to be upper caste individuals).

newtoboard
11-26-2013, 07:08 PM
There have been multiple migrations out of and into south Asia (and India).
If you saw the latest paper by Greg Magoon et. al, they have found a new haplogroup in a Telugu man, that indicates south India has a deep connection with the y-DNA NO* branch.
This is an amazing find (as big as A00).
I think all of these simplistic Aryan invasion and Out of India theory need to be thrown out the window, and some real work needs to be done.
If R is in itself 240000 years old in Siberia, I wonder how old all the other haplogroups are....
I wish South Asians would pool together and contribute towards Genetic testing of people from the Subcontinent.....but I doubt this will happen anytime soon.
We are missing a huge chunk of useful data from the region. Indians living outside India do not even represent 0.1% of its genetic variation ( as most tend to be upper caste individuals).


There have been multiple migrations out of and into south Asia (and India).
If you saw the latest paper by Greg Magoon et. al, they have found a new haplogroup in a Telugu man, that indicates south India has a deep connection with the y-DNA NO* branch.
This is an amazing find (as big as A00).
I think all of these simplistic Aryan invasion and Out of India theory need to be thrown out the window, and some real work needs to be done.
If R is in itself 240000 years old in Siberia, I wonder how old all the other haplogroups are....
I wish South Asians would pool together and contribute towards Genetic testing of people from the Subcontinent.....but I doubt this will happen anytime soon.
We are missing a huge chunk of data from the region.

Could you post the paper? Did they actually find an NO* lineage?

soulblighter
11-26-2013, 07:12 PM
Could you post the paper? Did they actually find an NO* lineage?

Here it is...I meant to include it in the previous post but forgot! It is a brother to NO* They call it X* for now.

http://biorxiv.org/content/early/2013/11/22/000802.1.full.pdf

parasar
11-26-2013, 07:29 PM
Here it is...I meant to include it in the previous post but forgot! It is a brother to NO* They call it X* for now.

http://biorxiv.org/content/early/2013/11/22/000802.1.full.pdf

Plus, potentially three GHIJK level mutations; and the Papuan HGDP00542 sample which lends further support to a SE Asian origin for P.

Sein
11-27-2013, 12:20 AM
Is this injunction against killing Hindus applies only to killing Brahmans and Shramans or all Hindus?

I'm not quite sure about this, perhaps all Hindus?


There have been multiple migrations out of and into south Asia (and India).
If you saw the latest paper by Greg Magoon et. al, they have found a new haplogroup in a Telugu man, that indicates south India has a deep connection with the y-DNA NO* branch.
This is an amazing find (as big as A00).
I think all of these simplistic Aryan invasion and Out of India theory need to be thrown out the window, and some real work needs to be done.
If R is in itself 240000 years old in Siberia, I wonder how old all the other haplogroups are....
I wish South Asians would pool together and contribute towards Genetic testing of people from the Subcontinent.....but I doubt this will happen anytime soon.
We are missing a huge chunk of useful data from the region. Indians living outside India do not even represent 0.1% of its genetic variation ( as most tend to be upper caste individuals).

I wholeheartedly agree with you, I think the picture is getting richer and more complex by the day. We really need robust representation of South Asia in autosomal data-sets, and I hope that isn't too far on the horizon.

Dr_McNinja
11-27-2013, 03:31 AM
Someone just mentioned something in a comment on the Harappa blog which reminded me, we could really use a lot more data from the Western Punjabi group (Pahari/Potohari/Potwari/Mirpuri/Hindko/etc). Groups including Hindkowans, Kamboj, etc. Kamboj are supposed to be one of the oldest and closest thing to indigenous populations of the Punjab region that we have. I lost touch with a friend I made online way back who was a Kamboj from my village near Faisalabad in Pakistan. They're pretty much indistinguishable from other Punjabis but perhaps there are some surprises in store with their haplogroups or the Pahari populations. The original language of Jatts ("Jatki") was closer to the Western Punjabi tongues than the current popular Eastern/Central dialects. Not to mention the Jati/Jakati gypsies of Russia/Ukraine/Afghanistan. I had run into quite a few Pahari-speaking Pakistani Jatts who had emigrated to Europe online but lost touch with most of them.

What's weird is that the Pahari group has had some tensions both with their Pashtun neighbors to the West and their Central/Eastern-dialect Punjabi brethren to the East. Lots of racism all around. They didn't fare much better in Europe (a lot of Pakistani immigrants to the UK are from them). Many of them have settled in Rawalpindi.

(The Kamboj are a total Central/Eastern Punjabi group speaking Majhi/Doabi/Malwai/etc dialects, I just mention them too because of their different history)

Sapporo
11-27-2013, 04:05 AM
@Dr_McNinja

HRP_0135 is a Pahari speaking Punjabi Rajput that I personally know. He is placed in the Kashmiri Pahari group but he is labeled as a Punjabi Pahari Rajput. I believe he is from around Bhimber in Azad Kashmir. His Y-DNA is R1a1a and his mt-DNA is R0a2.

@Sein

That Western UP/Uttarakhand Tyagi Brahmin I told you about is on HAP now using his Geno 2.0 data. So, Zack accepts Geno 2.0.

There is also a new UP Pathan on HAP using his Ancestry.com data.

https://docs.google.com/spreadsheet/ccc?key=0AuW3R0Ys-P4HdGE4eDh6emt1dUs2U2pXTkVjS0lsV1E&hl=en#gid=0

Sein
11-27-2013, 04:23 AM
@Dr_McNinja - I think you make a great point here, western Punjabi groups should really be sampled. I'm very interested in how Hindokowan from the Peshawar region stack up genetically.

@Sapporo - That's great! I wonder if Zack will make an official announcement regarding accepting Ancestry.com raw-data. Also, this individual is the first person of fully UP Pathan origins whose results we can see. I wonder if he'll be substantially North Indian-like, or similar to Pashtuns. Probably intermediate?

Dr_McNinja
11-27-2013, 04:32 AM
I have a friend of Uttar Pradesh origin who's likely got Pathan heritage. I heard about the 23andMe legal issues, which service should I recommend to him so he can join HAP?

Sapporo
11-27-2013, 04:34 AM
@Sapporp - That's great! I wonder if Zack will make an official announcement regarding accepting Ancestry.com raw-data. Also, this individual is the first person of fully UP Pathan origins whose results we can see. I wonder if he'll be substantially North Indian-like, or similar to Pashtuns. Probably intermediate?

I'm curious as well. As for the individual, I suspect intermediate as well but we'll see in a short while anyways.


I have a friend of Uttar Pradesh origin who's likely got Pathan heritage. I heard about the 23andMe legal issues, which service should I recommend to him so he can join HAP?

I would still recommend 23andMe personally. 23andMe is having issues with the FDA but I'm certain they will sort it out. The issue is due to the marketing of their health testing services which are fairly speculative anyways.

I've heard rumors of a Black Friday sale as well.

Sapporo
11-27-2013, 04:36 AM
Edit: Double post.

Sein
11-27-2013, 05:54 AM
I agree with Sapporo, but if the legal issues do prove serious, FTDNA's Family Finder. Also, I think after 23andMe's "upgrades" to V4, FTDNA raw-data should have more SNPs. "Population Finder" was a terrible test in my case, as I'm not even close to being predominantly Middle Eastern, with some Siberian. But your results made perfect sense, and FTDNA is going to update it by early 2014.

Dr_McNinja
11-27-2013, 08:50 PM
Found a relatively high match for HRP0286 (Afghan Pashtun Mohammadzai/Barakzai). It's an ethnic Georgian @ 75.77%

Sein
11-28-2013, 01:06 AM
On my list, the ethnic Georgian is HRP0286's highest match. Followed by a Pakistani Pashtun and the Half Punjabi-Half Pashtun at 75.71%. Then four individuals are tied at 75.65%, among them Sapporo. I match the ethnic Georgian better than the I match HRP0286.

Dr_McNinja
11-28-2013, 02:24 AM
I've got two @ 75.66% to HRP0286, both Jatts (one Rajasthani Jatt, HRP0359, and another Sikh Gill Jatt, probably Punjabi).

Here's an updated list for HRP0370:


75.80% - 1/2 Pakistani Punjabi, 1/2 Pakistani Pathan/Pashtun (Female)
75.75% - Afghan Pashtun Ghilzai Esmailkhel
75.73% - Haryana Hindu Jatt (HRP0170)
75.70% - 1/2 English, 1/2 Punjabi Jatt (Sandhu)
75.67% - 1/2 Afghan Pashtun, 1/2 Russian (Kyrgyzstan)
75.66% - Iranian (Tehran and Tabriz)
75.66% - Pakistan Punjabi Jatt Muslim (Sandhu)
75.65% - Punjabi Jatt Sikh (Gill)
75.65% - Iranian (Descendant of Kara Khitan Mongols on paternal side)
75.65% - Pakistan Pashtun (HRP0282) (Sein)
75.64% - Punjabi Jatt Muslim (Gill) (HRP0349) (Me)
75.62% - Punjabi Jatt Sikh (Grewal/Mann/Gill) (HRP0341) (Sapporo)
75.62% - Punjabi Khatri
75.61% - Rajasthani Hindu Jatt (HRP0359)
75.59% - 1/2 British, 1/2 Sikh Jatt (Male)
75.59% - 3/4 Pakistani Punjabi, 1/4 Afghan Pashtun
75.58% - Punjabi Sikh Jatt (Cheema)
75.58% - Arora
75.56% - Georgian
75.55% - Punjabi Chaudhry/Mian
75.54% - Haryana Hindu Jatt (HRP0275 or HRP0276)
75.53% - Punjabi Brahmin
75.52% - Punjabi Jatt (Athwal)
75.52% - Khan/Awan (Punjabi Awan?)
75.51% - Rajasthani Hindu Jatt (HRP0360)
75.49% - Afghan Pashtun Achakzai
75.49% - Greek (Black Sea, Pontic Greek, Abkhazia)
75.48% - Georgian
75.48% - Punjabi of mixed ancestry including Tajikstan (husband of 1/2 Pakistani Punjabi, 1/2 Pakistani Pathan Female @ 75.80%)
75.48% - Alevi Kurd
75.48% - Iranian
75.47% - Ukraine, Tatarstan, Chechnya
75.47% - Khetran Baloch (Durrani surname)
75.46% - Half Scandinavian, Half Moroccan+Algerian+South Asian
75.45% - Armenian
75.45% - South Indian of some kind (Rao/Iyer/Ramaswamy/Venkateswaran)
75.44% - Ukraine
75.43% - European
75.43% - Nepalese (Brahmin, I think)
75.43% - Punjabi (Multan)
75.42% - Punjabi Sikh Jatt (Gill)
75.42% - Greek (Pontic/Greek/NE-Turkey/Abkhazia)
75.41% - Greek (Pontic Greek including Southern Russia, Turkey)
75.41% - Afghan Pashtun Mohammadzai (HRP0286)
75.40% - Some type of Indian Punjabi
75.38% - 1/2 British, 1/2 Sikh Jatt (Female)
75.38% - Punjabi Chamar (possibly, definitely Indian)
75.38% - Kashmiri Butt, Muslim
75.38% - Ukraine
75.37% - Ukraine
75.36% - Rajasthani (not Jatt probably)
75.35% - Pakistani Pathan
75.31% - South Indian Karnatka Brahmin
75.30% - Ukraine
75.29% - Punjabi Jatt Muslim (HRP0374) (My mom)
75.29% - Iraq/Bulgaria
75.24% - Pakistani/Indian from Bukhara, Tajikstan in 18th century
75.22% - Bengali
75.21% - South Asian Muslim
75.05% - Mostly Northern European, some Spanish
74.74% - Pakistani Punjabi (Father of female @ 75.80%)
73.77% - Punjabi Brahmin
73.74% - Greek
73.44% - 1/2 Pakistani, 1/8 British, rest South Indian

Dr_McNinja
11-30-2013, 11:53 PM
Was messing around on GEDmatch and found you, Sein, and your Popalzai friend on HRP0370's one-to-many results. Found Paul Gill on mine (and my parents).

Dr_McNinja
11-30-2013, 11:56 PM
Do you know who this is?


S-Indian 25.40%
Baloch 39.25%
Caucasian 17.60%
NE-Euro 10.91%
SE-Asian -
Siberian 0.80%
NE-Asian -
Papuan 0.48%
American 1.44%
Beringian 0.78%
Mediterranean -
SW-Asian 3.34%
San -
E-African -
Pygmy -
W-African - Kit # M053658. Their name is only listed as "M".

Sein
12-01-2013, 12:27 AM
That's my mother's kit.

Dr_McNinja
12-01-2013, 03:28 AM
Do you know who kit M061909 is? Is that HRP0281?

Sein
12-01-2013, 03:52 AM
Not sure, are the results similar to myself and HRP0281?

Sapporo
12-01-2013, 03:56 AM
Not sure, are the results similar to myself and HRP0281?

I don't remember all of your Gedmatch results Sein but the kit's HarappaWorld DIY and Eurogenes is very similar to yours but just very slightly different. That's why I think they are also likely the results of HRP0281.

Sein
12-01-2013, 04:13 AM
Cool! I actually hope this is a different individual, that would be a nice addition in terms of "sample size".

Sein
12-01-2013, 04:34 AM
Just out of curiosity, how'd you find this kit number?

Dr_McNinja
12-01-2013, 04:45 AM
I found a bunch of Pashtun (including yourself and your Popalzai friend) in the GEDmatch one-to-many for HRP0370. Found another two Jatts (both in Harappa somewhere I think) on my one-to-many.

I added them here:

https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdFMtRHVlZDBuQ3lMcjhxMDE4V3JoY lE&usp=sharing

Remarkable similarity between the Kandahar/Popalzai folks, some differences from other Pashtun. Is HRP0286 on GEDmatch?

Sein
12-01-2013, 04:50 AM
"Could you PM me the Pashtun kit numbers?"

Sorry about that, I didn't read closely. HRP0286 is Shenodoah.

Sein
12-01-2013, 05:08 AM
Also, has anyone tried the new PCA functions they've implemented at GEDMatch for every calculator that has an oracle?

Sapporo
12-01-2013, 07:34 AM
Also, has anyone tried the new PCA functions they've implemented at GEDMatch for every calculator that has an oracle?

I've experimented with them but I have no idea how to set them up to mimic Dr. McDonald's world PCA plot he produced for me.

Sein
12-01-2013, 08:18 AM
It would be nice if we could attempt that, but the PCA's in question are based on components, rather than the individual genotype. So they won't be as accurate. For example, here are mine's and M061909's, for HarappaWorld, and the scale is 300% (so that we're visible):

990

991

Both of us have nearly identical positions. But some things are weird. In terms of PC1, both of us are identical to the Baloch and Brahui, and separated from the HAP Pashtuns (of which it's likely we're both a part, and in my case, absolutely certain). The HGDP Pashtuns are also very distant.

MfA
01-05-2014, 12:27 AM
Palisto's Globe13 genomemapping tool plotted me on 37.64, 44.35 http://goo.gl/maps/asi5J

http://abload.de/img/desktop_2014_01_05_02mor9z.jpg

ZephyrousMandaru
01-05-2014, 12:57 AM
Palisto's Globe13 genomemapping tool plotted me on 37.64, 44.35 http://goo.gl/maps/asi5J

http://abload.de/img/desktop_2014_01_05_02mor9z.jpg

You're just north of me, I'm in South-Central Iraq 32.77, 45.23.

http://img856.imageshack.us/img856/5113/6f4v.png

ZephyrousMandaru
01-05-2014, 12:43 PM
You're just north of me, I'm in South-Central Iraq 32.77, 45.23.

http://img856.imageshack.us/img856/5113/6f4v.png

Oh shit, never mind lol. There was an error in the Macros' code. This is my actual position.

Zakho, Iraq 37.04, 42.68

http://img809.imageshack.us/img809/9013/pw3b.png

Although that last one is more reflective of my deeper origins.

Dr_McNinja
01-09-2014, 06:07 AM
I was just fooling around and plotted the Eurogenes K36 numbers with a formula on my spreadsheet:

https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.k4m5y2qxNrNY

I used the admixture percentages to weight averages of the various latitudes and longitudes. Not sure what happened with the Tajik, their European admixture pulled them sharply to the West.

Ahaddad
01-09-2014, 07:55 AM
How i put my dad in spatial eurogenes map spa? His results DETAILS
Continent (Subcontinent) Population Percentage Margin of Error
Middle East (Middle Eastern) Druze 47,57% ±2,00%
Middle East (Iranian) Iranian 48,93% ±1,31%
Europe Orcadian, Romanian, Russian, Sardinian, Spanish, Tuscan, Finnish, French, Basque 3,50% ±1,01%


How i can know exactly the point?

Dr_McNinja
01-09-2014, 10:11 PM
I was just fooling around and plotted the Eurogenes K36 numbers with a formula on my spreadsheet:

https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.k4m5y2qxNrNY

I used the admixture percentages to weight averages of the various latitudes and longitudes. Not sure what happened with the Tajik, their European admixture pulled them sharply to the West.I did one with the DIYHarappa results from the spreadsheet: https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.k3WNiCSfJO9s

The spread on the Harappa results seems a little less coherent than the Eurogenes K36 one (the spread within Punjabis and Pashtun is equal to the spread between them which shouldn't be the case). Perhaps that's due to easier identification of coordinates for K36's very specific/localized components. I used the modal populations to find coordinates in Harappa though, so I'm pretty sure that isn't the case.

Sein
01-09-2014, 11:26 PM
Very interesting differences between both of these maps. The Eurogenes K36 map makes more sense, and shows some tight ethnic clustering. I would have expected the HarappaWorld map to be more accurate for all of us, but I guess not.

Dr_McNinja
01-10-2014, 12:21 AM
Very interesting differences between both of these maps. The Eurogenes K36 map makes more sense, and shows some tight ethnic clustering. I would have expected the HarappaWorld map to be more accurate for all of us, but I guess not.Eurogenes K13 isn't bad either: https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.kZfduZmb6wnE The Nepalese/Bihar (and myself, assuming Uttar Pradesh or thereabouts as usual) and Bengali differences show up better. I feel like I could manipulate the K13 results and move/rotate them around to fit well onto the map.

Sein
01-10-2014, 12:34 AM
It would pretty interesting if you could rotate the results to fit the map. Although, I still say that you don't fit in Uttar Pradesh. Your results are slightly different from some Punjabi Jatts, but fairly similar to other Punjabi people of non-Jatt backgrounds. And in the new paper on European genetics, they used 8 Punjabi samples from Lahore. These individuals were rather different from the HGDP samples, and similar to Gujaratis and Uttar Pradeshis. So, with more sampling, you might turn out less South Asian than most Punjabi people. I think your position is more reflective of sampling.

Dr_McNinja
01-10-2014, 03:47 AM
Here's EUTest (Old): https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.kJKzuLSjtWHw This is the most accurate geographically, clustering is pretty good too

Added a couple more individuals to the K36 plot (two Pashtun and an Iranian)

parasar
01-10-2014, 03:47 AM
Eurogenes K13 isn't bad either: https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.kZfduZmb6wnE The Nepalese/Bihar (and myself, assuming Uttar Pradesh or thereabouts as usual) and Bengali differences show up better. I feel like I could manipulate the K13 results and move/rotate them around to fit well onto the map.

Good idea. You could try a coordinate shift and put me near Patna. For the past five hundred years we have pretty strong connection to that neighborhood (both maternal and paternal). Plus as our caste is mentioned in Asokan inscriptions, a speculative connection with Pataliputra may be as old as 2500ybp (eg BAbhan BAbhanesu etc http://books.google.com/books?id=xcMUAAAAYAAJ&pg=PA70).

That could bring most Punjabis in the Punjab (though you would likely land in UP), and the Pakhtoons in Afghanistan/NWFP. The Nepalese sample a little north to me will also fall in place.

Dr_McNinja
01-10-2014, 04:03 AM
Good idea. You could try a coordinate shift and put me near Patna. For the past five hundred years we have pretty strong connection to that neighborhood (both maternal and paternal). Plus as our caste is mentioned in Asokan inscriptions, a speculative connection with Pataliputra may be as old as 2500ybp (eg BAbhan BAbhanesu etc http://books.google.com/books?id=xcMUAAAAYAAJ&pg=PA70).

That could bring most Punjabis in the Punjab (though you would likely land in UP), and the Pakhtoons in Afghanistan/NWFP. The Nepalese sample a little north to me will also fall in place.Unfortunately I have no idea how to do that in Google Maps. The best I can do is just drag all the dots around while selected together but it won't let me rotate. I could try to tinker with the coordinate formula in the spreadsheet. I think using Photoshop would be the easiest if anyone else has it and wants to try. You could just mask around the dots then rotate them around on a fresh map.

Here's the formula I used in the spreadsheet if anyone wants to mess with it:

Latitude: =SUM(VALUE(C3)*53+VALUE(D3)*55.5833+VALUE(E3)*40+V ALUE(F3)*33.8333+VALUE(G3)*22+VALUE(H3)*34+VALUE(I 3)*8.5074+VALUE(J3)*21.0333+VALUE(K3)*60+VALUE(L3) *64-VALUE(M3)*9.5+VALUE(N3)*8.3+VALUE(O3)*7.6833)

Longitude: =SUM(-VALUE(C3)*8+VALUE(D3)*24.1167+VALUE(E3)*9+VALUE(F3 )*35.7667+VALUE(G3)*38+VALUE(H3)*54+VALUE(I3)*76.9 720+VALUE(J3)*105.85+VALUE(K3)*105-VALUE(L3)*150+VALUE(M3)*147.1167+VALUE(N3)*39.1167-VALUE(O3)*0.9833)

Dr_McNinja
01-10-2014, 04:39 AM
Not the right place for this, but for you R1a guys, what SNPs should I be looking into testing at FTDNA? My maternal uncle's Y-DNA12 results just came back and FTDNA predicts R-M198/R1a1a*. I joined these two projects:

http://www.familytreedna.com/public/R1a/default.aspx?vgroup=R1a&section=yresults

http://www.familytreedna.com/public/R1aY-Haplogroup/default.aspx?vgroup=R1aY-Haplogroup&section=yresults

Kit # 311047

Y-DNA37 is still in progress. I'll wait for that before making a stab at which SNPs to test for (getting a Geno 2.0 test to him overseas will be difficult).

DYS393 was 14, interestingly enough. The only Indians I see with that are two Gujaratis. The oldest paternal ancestor on that line was my great-great-grandfather in Jhang/Faisalabad area, who was a Pansota. My family doesn't know anything before that, odds are he emigrated from Hoshiarpur in Indian Punjab if the history I found online is true.

Dr_McNinja
01-10-2014, 06:09 AM
I fixed the K36 plot, I had Siberian/Amerindian plotted differently. There are 3 layers with different ways of plotting it, the last of which is probably the best bet (Siberia and Amerindian as two separate components with Amerindian in the Western hemisphere). This pulls the Pakistani Pashtun further over to the Northwest from the Punjabis. The Tajik seemed to be closest with Siberian/Amerindian plotted together over towards the Northeast. The original had Siberian/Amerindian together over towards the Northwest (Bering Sea in the Western hemisphere).

When I get time tomorrow I'm going to later add one more layer, plotting Amerindian in Brazil. I just kept it at Beringian for this because with South Asians the Amerindian likely represents an Arctic component.

Sein
01-10-2014, 06:20 AM
Is it possible for you to add my phased paternal kit? I'm interested in seeing where my father will cluster (he is a Pakistani Pashtun). Also, M061909 is an Afghan Pashtun, HRP0281. I had the opportunity to talk to them.

Dr_McNinja
01-10-2014, 03:53 PM
Is it possible for you to add my phased paternal kit? I'm interested in seeing where my father will cluster (he is a Pakistani Pashtun). Also, M061909 is an Afghan Pashtun, HRP0281. I had the opportunity to talk to them.What's the GEDmatch #?

Dr_McNinja
01-10-2014, 04:02 PM
I ported over most of the entire Harappa spreadsheet: https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.kU-eYEC4D4yk

I think the lack of a Western European component similar to "Atlantic" is screwing it up a little. Everything is pulled towards the Mediterranean because that picks up some of what would otherwise go to an Atlantic/Western-European component. That's probably pulling the Northern Europeans down and pulling some of the South Asians to the West rather than Northwest.

parasar
01-10-2014, 05:52 PM
Not the right place for this, but for you R1a guys, what SNPs should I be looking into testing at FTDNA? My maternal uncle's Y-DNA12 results just came back and FTDNA predicts R-M198/R1a1a*. I joined these two projects:

http://www.familytreedna.com/public/R1a/default.aspx?vgroup=R1a§ion=yresults

http://www.familytreedna.com/public/R1aY-Haplogroup/default.aspx?vgroup=R1aY-Haplogroup§ion=yresults

Kit # 311047

Y-DNA37 is still in progress. I'll wait for that before making a stab at which SNPs to test for (getting a Geno 2.0 test to him overseas will be difficult).

DYS393 was 14, interestingly enough. The only Indians I see with that are two Gujaratis. The oldest paternal ancestor on that line was my great-great-grandfather in Jhang/Faisalabad area, who was a Pansota. My family doesn't know anything before that, odds are he emigrated from Hoshiarpur in Indian Punjab if the history I found online is true.

311047 Rizwan Pansota (Punjabi Jatt) Pakistan R1a1a 14 25 16 11 11-14 12 12 10 13 11 30
This is not sufficient to figure out the subclade. Once the 37 markers are in, please let me know.

True, DYS393=14 is not that common overall. It is probably old.
It is common in the UP Chamar who form the bulk of agricultural caste of UP. Southern India and the southern Indus region show that repeat too.
Eg. cha-06 cha India North South Asia Caste-low Indo-European M017 R1a1 16 12 17 13 24 10 11 14 10 9 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1380230/table/TB3/

Dr_McNinja
01-10-2014, 07:43 PM
311047 Rizwan Pansota (Punjabi Jatt) Pakistan R1a1a 14 25 16 11 11-14 12 12 10 13 11 30
This is not sufficient to figure out the subclade. Once the 37 markers are in, please let me know.

True, DYS393=14 is not that common overall. It is probably old.
It is common in the UP Chamar who form the bulk of agricultural caste of UP. Southern India and the southern Indus region show that repeat too.
Eg. cha-06 cha India North South Asia Caste-low Indo-European M017 R1a1 16 12 17 13 24 10 11 14 10 9 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1380230/table/TB3/
The only one that matches it exactly there is #410:

410 iyr-48 iyr India South South Asia Caste-high Dravidian M017 R1a1 16 12 17 13 25 11 11 14 10 9

My uncle's values:

DYS19 = 16
DYS388 = 12
DYS389 = 13 ... 30 (?? Don't know how FTDNA's values correlate to that study)
DYS390 = 25
DYS391 = 11
DYS392 = 11
DYS393 = 14
DYS439 = 10

parasar
01-10-2014, 08:31 PM
The only one that matches it exactly there is #410:

410 iyr-48 iyr India South South Asia Caste-high Dravidian M017 R1a1 16 12 17 13 25 11 11 14 10 9

My uncle's values:

DYS19 = 16
DYS388 = 12
DYS389 = 13 ... 30 (?? Don't know how FTDNA's values correlate to that study)
DYS390 = 25
DYS391 = 11
DYS392 = 11
DYS393 = 14
DYS439 = 10

That is 13, 17, same as iyr-48

Sein
01-10-2014, 09:12 PM
What's the GEDmatch #?

PM760347P1

I also have kit numbers for an Afghan Tajik, and an individual who is 75% Afghan Pashtun+25% Afghan Tajik. I'll pm you the kit numbers (it's best that these two Afghan individual's kit numbers should remain private).

Dr_McNinja
01-14-2014, 05:15 AM
I adjusted the map a little, I fixed a mistake with the Caucasian component and moved "Amerindian" to the home of the Ket and Selkup people (all the way north inbetween Mongolia and Kazakhstan). They average 3-4% Amerindian on HAP, right behind the Greenlanders (who are right behind the Native Americans) and just ahead of Altaians, Tajiks, and then a few South Asian populations (Haryana and Rajasthani Jatts average nearly 2% Amerindian which is significant I think... it's why the Jatts were all curving to the Southwest on the previous version). I also moved Mediterranean a little north towards Western France to make it like an "Atlantic-Mediterranean" (average of Sardinian, Basque, and North Sea/North Atlantic).

Both maps are available as layers on the same map, the one labeled "ADJUSTED" incorporates these changes.

https://mapsengine.google.com/map/u/0/edit?mid=z1DL_IP3p3uA.kU-eYEC4D4yk

I added two Afghan Tajiks as well, courtesy of Sein.

Two remaining issues are that the individuals plotted to the west of Balochistan will now pull back towards the East because of higher Baloch admixture. Not sure what to do about Papuan and SW-Asian since I doubt these represent actual Papuan or Arab geographic ancestry signals and are likely just noise (at least the Arctic components can point towards, well, the Arctic).

Dr_McNinja
01-14-2014, 05:46 AM
One noticeable pattern is that one section of Jatts, Brahmins, Punjabis, and Uttar Pradesh overlap with lots of mixed individuals and other similar groups (Kashmiri, Khatri, Baloch, Sindhi, etc). The Muslim Punjabi Jatts specifically overlap with the Brahmins and Punjabis indicating either mixing with these groups (simple theory) or similar long-term North Indian ancestry. The latter explanation is possible if you consider that many Sikh Punjabi Jatts hail from western areas now in Pakistan and many Muslim Punjabi Jatts hail from eastern areas in India. It's probably a bit of both.

I had read that North India (Uttar Pradesh specifically) had a cosmopolitan upper-caste Muslim nature to it under the Mughals and even others. There seems to be lots of mixed Pathan/Pashtun ancestry in the area as well. The highest likelihood of Jatts converting to Islam (before the political events of the 20th century) would be from these areas I would wager, especially during the times when the powerful Indian states heavily recruited for their armies from these races. So North India might have had all these factors going at the same time (political, economic, and ideological opportunity).

The other thing is that the ethnic groups (at least in the Harappa calculator) form a gradient. There are few Pashtun individuals but I imagine if we had like a dozen more they'd fill up the spot between Iranians and Jatts, with the Pakistani ones overlapping with the Jatts. If the Pakistani/Muslim Jatts represent more North Indian origins, then the gradient makes even more sense.

Sein
01-14-2014, 06:43 AM
Great work with the maps! They look really nice, and the distribution is fascinating.

You make a great point concerning the Indian Sikh Jatts. My mother is a Pakistani Pashtun, and she does overlap with the Indian Sikh Jatts. If it's of any relevance, some Dodecad Oracles construe me as 55% Indian Sikh Jatt, and 45% Tajikistani Tajik. I get the Tajiks due to my Siberian admixture. Excluding that, I would probably be construed as 55% Indian Sikh Jatt+45% Iranian. Basically, a hybrid of both populations. I think your idea about Sikh Jatts is very plausible. It would explain why Sikh Jatts seem to be less South Asian than Muslim Jatts from Pakistan. That, or greater endogamy among Sikhs?

Now that you've plotted individuals using components, perhaps you could try a HarappaWorld map based on reference population coordinates? For example, we know that the HGDP Pashtuns are from the Kurram Agency. Most researchers think the samples come from Parachinar, and one of the samples is strong evidence for this (HGDP00220, Parachinar formerly attracted many Hazara migrants from Afghanistan). Parachinar is basically in Afghanistan. Well, very close to the Afghan border. The Brahui samples seem to be from Kalat in Balochistan. The Sindhi samples are presumably from Mirpur Khas. The public Tajik samples are from different parts of Tajikistan, so one could use Dushanbe. So would this be possible, plotting the sampling location coordinates, using the average results for those populations, and then plotting all of us in relation to the actual sampling locations?

MfA
01-14-2014, 12:26 PM
https://mapsengine.google.com/map/u/0/edit?mid=z1DL_IP3p3uA.kU-eYEC4D4yk


Interesting thanks.. Here's me, It seems that I'm the most western of Kurds here and also the closest one to Ezidi Kurd between the Assyrians and Armenians.

http://abload.de/img/desktop_2014_01_14_14w2ji3.png

Here's my admixture results.


3.84% S Indian
23.68% Baloch
40.98% Caucasian
3.62% NE Euro
0.00% SE Asian
1.19% Siberian
0.61% NE Asian
1.02% Papuan
1.43% American
1.73% Beringian
8.48% Mediterranean
12.05% SW Asian
0.60% San
0.60% E African
0.18% Pygmy
0.00% W African

BTW there are 2 different set of calculators for Harappa and one is before the batch number of 200s and later. I'm from the latter iteration and there is a subtle difference between two calcs. Although Zack don't think there is a meaningful difference, I don't agree with him. Most of the Kurds here is from first batch unlike me. If you think you can update it later, here's my first batch results.



2.40% S-Indian
24.40% Baloch
43.43% Caucasian
4.65% NE-Euro
0.00% SE-Asian
0.54% Siberian
0.51% NE-Asian
0.07% Papuan
0.75% American
1.20% Beringian
9.32% Mediterranean
12.57% SW-Asian
0.00% San
0.00% E-African
0.13% Pygmy
0.00% W-African

Dr_McNinja
01-14-2014, 02:53 PM
Great work with the maps! They look really nice, and the distribution is fascinating.

You make a great point concerning the Indian Sikh Jatts. My mother is a Pakistani Pashtun, and she does overlap with the Indian Sikh Jatts. If it's of any relevance, some Dodecad Oracles construe me as 55% Indian Sikh Jatt, and 45% Tajikistani Tajik. I get the Tajiks due to my Siberian admixture. Excluding that, I would probably be construed as 55% Indian Sikh Jatt+45% Iranian. Basically, a hybrid of both populations. I think your idea about Sikh Jatts is very plausible. It would explain why Sikh Jatts seem to be less South Asian than Muslim Jatts from Pakistan. That, or greater endogamy among Sikhs?Either are possibilities. In my family's case, they actually converted from Sikhism in the 19th century, and I can trace that back to my great-great-great grandparents on my maternal side at the latest (around the same or later on my paternal side). Of course they did mix with locals wherever they were (there are a lot of grandparents when you go back that far) so I can't track down everyone (especially the women), just the major patrilineal ancestors.


Now that you've plotted individuals using components, perhaps you could try a HarappaWorld map based on reference population coordinates? For example, we know that the HGDP Pashtuns are from the Kurram Agency. Most researchers think the samples come from Parachinar, and one of the samples is strong evidence for this (HGDP00220, Parachinar formerly attracted many Hazara migrants from Afghanistan). Parachinar is basically in Afghanistan. Well, very close to the Afghan border. The Brahui samples seem to be from Kalat in Balochistan. The Sindhi samples are presumably from Mirpur Khas. The public Tajik samples are from different parts of Tajikistan, so one could use Dushanbe. So would this be possible, plotting the sampling location coordinates, using the average results for those populations, and then plotting all of us in relation to the actual sampling locations?The only thing I can think of is to use the reference population coordinates as part of the weighted average. The averages of their current plotting locations are more or less useless since it's so far off geographically. I don't think Harappa can be plotted coherently on a geographical map, it's more like the PCA plots. I might tinker with some of the Eurogenes by moving the location of South Asian and Balochistan components to see if I can place them right.

Ahaddad
01-14-2014, 05:01 PM
Where fit in Spatial Map the Uzbekistan Jews???

Dr_McNinja
01-14-2014, 06:59 PM
I was messing around and got this from the EUTest results:

https://mapsengine.google.com/map/u/0/edit?mid=z1DL_IP3p3uA.kJKzuLSjtWHw (make sure it's on the 'UPDATED' layer)

I don't understand exactly what I just did.

I used this formula for latitude:

( (WestAsian*GeorgiaLatitude) + (GeorgiaLatitude+(WestAsian*GeorgiaLatitude)) ) / 2

And for longitude:

( (SouthAsian*SouthIndiaLongitude) + (SouthIndiaLongitude+(SouthAsian*SouthIndiaLongitu de)) ) / 2

Coordinates: Georgia: 42.0167, 43.7333 and South India: 8.5074, 76.972

That came out really well, but I don't know what I did or how to include the other components (which would probably push the Nepali back into place I figure, their East Asian isn't being counted). I don't know why using the latitude of the West Asian component (Georgia) and longitude of the South Asian component (South India in this case) comes really close to the actual geography. This doesn't work when using Balochistan for a West Asian component.

Sein
01-14-2014, 09:12 PM
I was messing around and got this from the EUTest results:

https://mapsengine.google.com/map/u/0/edit?mid=z1DL_IP3p3uA.kJKzuLSjtWHw (make sure it's on the 'UPDATED' layer)

I don't understand exactly what I just did.

I used this formula for latitude:

( (WestAsian*GeorgiaLatitude) + (GeorgiaLatitude+(WestAsian*GeorgiaLatitude)) ) / 2

And for longitude:

( (SouthAsian*SouthIndiaLongitude) + (SouthIndiaLongitude+(SouthAsian*SouthIndiaLongitu de)) ) / 2

Coordinates: Georgia: 42.0167, 43.7333 and South India: 8.5074, 76.972

That came out really well, but I don't know what I did or how to include the other components (which would probably push the Nepali back into place I figure, their East Asian isn't being counted). I don't know why using the latitude of the West Asian component (Georgia) and longitude of the South Asian component (South India in this case) comes really close to the actual geography. This doesn't work when using Balochistan for a West Asian component.

This is probably the most accurate map yet! Rather striking, the Bengali individual is in Bangladesh, and my mother is slightly south of the Peshawar-Kohat region, almost on the Indus (which is absolutely accurate in her case). In fact, I think my mother's spot is in Mianwali District, Punjab. The Iranian is also in a perfect spot. Also, pretty cool that it puts all Pashtuns west of the Indus, and all Punjabis east of it. And the two Kandahari Durrani Pashtun results are very suggestive, in light of the long sojourn of the Durrani in Herat, and eastern Khorasan. That's where they lived after the Ghilzai defeated and expelled them from southeastern Afghanistan, and they returned after Ahmad Shah Durrani consolidated power. Perhaps they mixed with local Farsiwan in the Herat region?

I'm on the eastern edge of Ghazni, near Paktika. No clue if I should take that seriously, the only way to verify this, in my case, would be finding out HRP0281/M061909's provincial origins in Afghanistan, as well as their tribal affiliation. HRP0281/M061909 is basically identical to myself, we only differ by 1% for various components, so if they originate from Ghazni-Logar-Paktika-Zabul, this map is spot on. Unfortunately, HRP0281 wasn't able to shed any light on this topic.

It seems that a Georgian-South Indian axis works very well, for all of us.

Dr_McNinja
01-15-2014, 02:23 AM
And the two Kandahari Durrani Pashtun results are very suggestive, in light of the long sojourn of the Durrani in Herat, and eastern Khorasan. That's where they lived after the Ghilzai defeated and expelled them from southeastern Afghanistan, and they returned after Ahmad Shah Durrani consolidated power. Perhaps they mixed with local Farsiwan in the Herat region?I'm pretty sure that's what happened because Dr. McDonald placed HRP0370 in that same spot in Iran over the border from Herat. Either that or being that far out helped them remain endogamous because Iranians don't have European admixture like these Pashtun do, it's more similar to Jatts and Brahmins. There's no way for ancient European admixture to get into the region any more recently than that, I think it had to have been present a long time.


It seems that a Georgian-South Indian axis works very well, for all of us.Yeah, except for those with significant East Asian, African, Arab, or other ancestry. The European seems to not really matter, it might be because the European signals are so old and widespread they don't have any geographic relevance in the region, only ethnic (kind of like Baloch does in Western Europe). The Haryana/Rajasthani Jatts would probably be placed quite close to their home (considering some of them had low Caucasian numbers). The only place where it might matter would be in cases like mine, including the European admixture would pull me into Punjab/Uttar Pradesh/Brahmin territory (as I am on the Harappa map), but going just by South Asian-West Asian, it puts me near Nepal. It's actually right near Lucknow and I used Lucknow as the center of "North India" on the FTDNA population finder results who averaged me out to within a few miles of my village near Faisalabad. So I think you might have been right along, my admixture is Punjabi, but a weird kind (McDonald used that West Euro component to average out my Indian component to a place in Pakistan... I wonder if other Punjabis are done this way or whether they get West Asian/India mixes instead). Who knows if I have roots that far north in India, perhaps my Y-DNA research might reveal something. If Geno 2.0 confirms I'm J2-M241, my DYS617 value is unheard of (every other J2-M241 I've seen is 12 or 13, I'm 10). Nepal is ground zero for the Y-STR variance of J2b2.

On the other hand, I'm now a little suspicious that FTDNA, DNA-Tribes, etc might have been purposely trying to average out my admixture in this way. It's a little too coincidental or perhaps they're the experts and were doing something right all along. What would make my family different is that our European admixture is geographically relevant without which we wind up over in Nepal, and in other Punjabis it would move them out of Punjab and west of the Indus. (EDIT: My dad's phased kit is between Haryana and Uttar Pradesh too)

I also still think the relatively high Amerindian values in Jatts is relevant and could possibly be a clue to a time of origin (maybe relative to Siberian and Beringian).

Dr_McNinja
01-15-2014, 02:44 AM
I updated it and added the 3/4 Pashtun, 1/4 Tajik person. They're right on top of HRP0281. The pure Afghan Tajik won't work because of his huge Siberian admixture.

EDIT: Just added your phased paternal kit too.

Sein
01-15-2014, 08:49 AM
The only place where it might matter would be in cases like mine, including the European admixture would pull me into Punjab/Uttar Pradesh/Brahmin territory (as I am on the Harappa map), but going just by South Asian-West Asian, it puts me near Nepal. It's actually right near Lucknow and I used Lucknow as the center of "North India" on the FTDNA population finder results who averaged me out to within a few miles of my village near Faisalabad. So I think you might have been right along, my admixture is Punjabi, but a weird kind (McDonald used that West Euro component to average out my Indian component to a place in Pakistan... I wonder if other Punjabis are done this way or whether they get West Asian/India mixes instead). Who knows if I have roots that far north in India, perhaps my Y-DNA research might reveal something. If Geno 2.0 confirms I'm J2-M241, my DYS617 value is unheard of (every other J2-M241 I've seen is 12 or 13, I'm 10). Nepal is ground zero for the Y-STR variance of J2b2...What would make my family different is that our European admixture is geographically relevant without which we wind up over in Nepal, and in other Punjabis it would move them out of Punjab and west of the Indus. (EDIT: My dad's phased kit is between Haryana and Uttar Pradesh too)

I think you've pretty much hit the nail right on the head.


...because Iranians don't have European admixture like these Pashtun do, it's more similar to Jatts and Brahmins.

This is actually quite interesting, and it is the one thing that I find fascinating about Pashtun genetics. For example, David at Eurogenes experimented with his data-set using MCLUST, and this is what he found for people from our neck of the woods:
https://docs.google.com/spreadsheet/ccc?key=0Ato3EYTdM8lQdE8xQ2N2VDBFUUQzS2RmRkhBVmNuZ Wc&usp=sharing

The great thing is that he also used the new Afghan Pashtun samples from Kunduz-Takhar-Baghlan. These Pashtuns have always interested me, since they live very far away from the traditional Pashtun region of southeastern Afghanistan/northwestern Pakistan. They also seem to intermarry with Tajiks and Uzbeks quite frequently, so they should be an interesting group. Hopefully, Zack will post their results for HarappaWorld in due time. That being said, what strikes me about their results is the Pashtun-Sindhi-Punjabi Jatt affinity. All of these Afghan samples get classified in the "Pashtun-Sindhi" cluster, and are solid members of this cluster, with exception to one individual who might be a mislabeled sample, and another individual who is recently admixed. An Afghan Tajik is also a solid member of the "Pashtun-Sindhi" cluster, although they might actually be a mislabeled Pashtun. Since these individuals fall into the "Pashtun-Sindhi" cluster, it is a sure bet that HRP0370, and my Kandahari friend, are also going to be members of the "Pashtun-Sindhi" cluster. So, even Pashtuns who are significantly more West Asian than Sindhis and Punjabi Jatts, still get classified with Sindhis and Punjabi Jatts, and don't show a strong affinity to Iranians. Why? We see this in Y-DNA as well, Pashtuns cluster strongly with northwestern South Asians, and are rather different from Iranians. That much was demonstrated in the recent Afghan genetics paper, which sampled quite a few Afghan Pashtuns for Y-DNA analysis.

Just a random side note, but it seems that Afghan and Pakistani Pashtuns are identical in terms of Y-DNA. No real differences whatsoever. But the mtDNA does differ. For Pakistani Pashtuns, 55.6% of their mtDNA is West Eurasian, 39.1% is South Asian, and 5.2% is East Asian. Macro-haplogroup R accounts for around 61% of Pakistani Pashtun mtDNA (I guess some R-derived linage are considered South Asian? My knowledge of haplogroups is very weak), macro-haplogroup M accounts for around 31% of Pakistani Pashtun mtDNA, and macro-haplogroup N accounts for around 8% of Pakistani Pashtun mtDNA. The most common haplogroups are HV and U7. These Pakistani Pashtuns are mostly from KP/NWFP. Afghan Pashtuns have somewhat different stats, at least the ones sampled from the far north. These percentages are based on the same Afghan Pashtuns from Kunduz-Takhar-Baghlan. For these Afghan Pashtuns, 51.1% of their mtDNA is West Eurasian, 27.8% is South Asian, and 10% is East Asian. Macro-haplogroup R accounts for around 70% of these Afghan Pashtuns' mtDNA, macro-haplogroup M accounts for around 23% of these Afghan Pashtuns' mtDNA, and macro-haplogroup N accounts for around 7% of these Afghan Pashtuns' mtDNA. The most common haplogroups are U2 and R0. I think this means all Pashtuns share an identical paternal heritage, but may have different maternal ancestries. The Afghan Pashtuns they sampled marry Tajik and Uzbek women quite frequently, so this might explain why their mtDNA is less West Eurasian, less South Asian, and more East Asian. But as expected, no real differences were detected for Y-DNA in comparison to Pakistani Pashtuns, who were sampled from KP/NWFP.

parasar
01-15-2014, 03:57 PM
....
Just a random side note, but it seems that Afghan and Pakistani Pashtuns are identical in terms of Y-DNA. No real differences whatsoever. But the mtDNA does differ. For Pakistani Pashtuns, 55.6% of their mtDNA is West Eurasian, 39.1% is South Asian, and 5.2% is East Asian. Macro-haplogroup R accounts for around 61% of Pakistani Pashtun mtDNA (I guess some R-derived linage are considered South Asian? My knowledge of haplogroups is very weak), macro-haplogroup M accounts for around 31% of Pakistani Pashtun mtDNA, and macro-haplogroup N accounts for around 8% of Pakistani Pashtun mtDNA. The most common haplogroups are HV and U7. These Pakistani Pashtuns are mostly from KP/NWFP. Afghan Pashtuns have somewhat different stats, at least the ones sampled from the far north. These percentages are based on the same Afghan Pashtuns from Kunduz-Takhar-Baghlan. For these Afghan Pashtuns, 51.1% of their mtDNA is West Eurasian, 27.8% is South Asian, and 10% is East Asian. Macro-haplogroup R accounts for around 70% of these Afghan Pashtuns' mtDNA, macro-haplogroup M accounts for around 23% of these Afghan Pashtuns' mtDNA, and macro-haplogroup N accounts for around 7% of these Afghan Pashtuns' mtDNA. The most common haplogroups are U2 and R0. I think this means all Pashtuns share an identical paternal heritage, but may have different maternal ancestries. The Afghan Pashtuns they sampled marry Tajik and Uzbek women quite frequently, so this might explain why their mtDNA is less West Eurasian, less South Asian, and more East Asian. But as expected, no real differences were detected for Y-DNA in comparison to Pakistani Pashtuns, who were sampled from KP/NWFP.

mtDNA appears to have spread in layers too, much as Y-DNA.
The first mtDNA layer I think is N, and is pan-Eurasian. M spread later from E/SE Asia.

Looking at N, some its oldest branches are in Australia, Papua, and East Asia. R is an N derivative but that too is present in Australia and SE Asia. Plus we have the 40000 year old Tianyuan cave mtDNA line that is an N derivative B.

http://www.pnas.org/content/110/6/2223/F1.large.jpg

Comparing pairwise nucleotide differences for chromosome 21 sequences, we see that pan-Eurasian affiliation with modern non-African populations that Tianyuan has indicating that Tianyuan was very close to all humans that spread throughout Eurasia ~45000ybp.
Tianyuan 35,938 33,390 31,059 30,333 30,297 23,168 22,906 (Sardinian) 23,756 21,944 (Karitiana) 22,802 23,339
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3568306/table/t01/

So it would be difficult overall to call N or R, east or west Eurasian, as they are pan Eurasian. R0 and U, I used to think were South and West Eurasian, but with paleolithic U spread as far as Baikal and Spain, perhaps U was pan-north Eurasian.

soulblighter
01-15-2014, 04:14 PM
I believed mtDNA R has highest diversity in south Asia? So I don't really consider it a West asian mtDNA haplotype.
Regarding Iranian vs Pashtun vs Indian, I do think there is a connection between Iranian and South Indian Brahmin groups. I don't know if this would extend to R-Z2123 or not...
Looks at the G2a types *(from FTDNA and the Afghan paper) compared in this post here http://www.anthrogenica.com/showthread.php?1102-List-of-public-G2a-P303-samples-negative-for-subgroups&p=17711&viewfull=1#post17711

parasar
01-15-2014, 05:10 PM
I believed mtDNA R has highest diversity in south Asia? So I don't really consider it a West asian mtDNA haplotype.


I doubt its origin in South Asia though. Its age is too old to be of west or south asian origin. B was certainly present in eastern Asia 40000ybp, so I think a E/SE Asian origin for R is more likely.



Regarding Iranian vs Pashtun vs Indian, I do think there is a connection between Iranian and South Indian Brahmin groups. I don't know if this would extend to R-Z2123 or not...
Looks at the G2a types *(from FTDNA and the Afghan paper) compared in this post here http://www.anthrogenica.com/showthread.php?1102-List-of-public-G2a-P303-samples-negative-for-subgroups&p=17711&viewfull=1#post17711

Have you compared their DYS461?

G2a looks to be like J2a, first an early separation of South Asian and South European branches, and then a later influx from Persia.

soulblighter
01-15-2014, 05:46 PM
Have you compared their DYS461?

G2a looks to be like J2a, first an early separation of South Asian and South European branches, and then a later influx from Persia.

Here are the DYS461 values

Kalash=8
Brahui=10
Iranians= mostly 10, but one 12
Turkmen/Uzbek/Kyrgyz = mostly 11, but one 10
Hazara=10
Tajik= one 9 and one 10
South Indians=??? sadly neither of us went to 111 markers.

parasar
01-15-2014, 07:09 PM
Here are the DYS461 values

Kalash=8
Brahui=10
Iranians= mostly 10, but one 12
Turkmen/Uzbek/Kyrgyz = mostly 11, but one 10
Hazara=10
Tajik= one 9 and one 10
South Indians=??? sadly neither of us went to 111 markers.

Thanks.

To the Kalash I would add 2 - which means that they are like Iranians.
To the Brahui (if they are from the same testing as the Kalash), I would also add 2 which means they are separate.
South Indians = Turkmen/Uzbek/Kyrgyz!

South Indians
Caste-middle Dravidian P15 G2 17 12 18 12 22 10 11 14 11 9
Caste-high Dravidian P15 G2 16 12 16 12 22 10 11 15 12 9
Caste-high Dravidian P15 G2 16 12 16 12 22 10 11 15 13 9
Caste-high Dravidian P15 G2 15 12 16 12 22 10 11 16 13 9
Caste-high Dravidian P15 G2 15 12 16 12 22 10 11 15 13 9
Caste-high Dravidian P15 G2 15 12 16 12 22 10 11 15 11 9
Caste-high Dravidian P15 G2 15 12 16 12 22 10 11 15 12 9
Caste-high Dravidian P15 G2 15 12 16 12 22 10 11 16 13 9

Indo-European
Caste-low Indo-European P15 G2 14 12 16 12 22 11 11 13 12 8
South Asia Indo-European P15 G2 17 12 16 12 23 10 11 14 11 8
South Asia Indo-European P15 G2 16 12 16 12 23 10 11 14 11 8
South Asia Indo-European P15 G2 16 12 16 12 23 10 11 14 11 8
South Asia Indo-European P15 G2 16 12 16 12 23 10 11 14 11 8
South Asia Indo-European P15 G2 15 12 16 12 21 11 11 14 11 9

Brahui - akin to Dravidian
South Asia akin to Dravidian P15 G2 15 12 18 12 22 10 11 13 11 10
South Asia akin to Dravidian P15 G2 15 12 17 12 23 10 11 13 11 10
South Asia akin to Dravidian P15 G2 14 12 19 12 22 10 11 13 11 10
Ref. Sengupta data



within haplogroup G, the Punjabi G-L166 sample, HG02681, here shares four L166-specific SNPs with the 5,300-year-old G-L166 Iceman mummy Ötzi, found in the Italian part of the Ötztal Alps [31] (data from ERP001144), and the same number with a subset of G-L91 samples from Corsica, Sardinia or Tuscany, inferred to be also G-L166 [2]. These SNPs (PF3178, Z6134, Z6213 and Z6287) are ancestral in the Puerto Rican L91+, L166- sample, HG01311.
http://biorxiv.org/content/biorxiv/early/2013/11/22/000802.1.full.pdf

soulblighter
01-15-2014, 07:27 PM
Thanks.

To the Kalash I would add 2 - which means that they are like Iranians.
To the Brahui (if they are from the same testing as the Kalash), I would also add 2 which means they are separate.
South Indians = Turkmen/Uzbek/Kyrgyz!


I don't follow. Why are we adding 2 to the values? All the data I posted are from the same Afghan paper. So the Iranian samples would also need to have 2 added if that be the case!

parasar
01-15-2014, 07:52 PM
I don't follow. Why are we adding 2 to the values? All the data I posted are from the same Afghan paper. So the Iranian samples would also need to have 2 added if that be the case!

Yes, repeat of 14 for Iranians then is possible. I will take a look at the Afghan paper.

Edit:
Took a look. There is something odd going on.
Cristofaro et al say: "Additionally, a total of 39 Y-STRS (DYS385a-b, DYS388, DYS389a, DYS389B, DYS390, DYS391, DYS392, DYS393, DYS394/19, DYS426, DYS437, DYS438, DYS439, DYS441, DYS442, DYS444, DYS445, DYS446, DYS447, DYS448, DYS449, DYS452, DYS454, DYS455, DYS456, DYS458, DYS459a-b, DYS460, DYS461n(TAGA)n, DYS462, DYS463, GGAAT1B07, YCAIIa-b, YGATAA10, YGATAC4/Y_DYS635, YGATAH4) were genotyped using two multiplex reactions." http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0076748

But that was clearly not done on HGDP-CEPH samples as so many "nd" show up. http://www.plosone.org/article/fetchSingleRepresentation.action?uri=info:doi/10.1371/journal.pone.0076748.s015
I suspect they did not STR genotype them at all, but imported the data (with the equivalency DYS461n(TAGA)n=DYSA7.2 without accounting for the difference of 2)
See eg. "DYS461=11 (equivalent to DYSA7.2=9)" http://archiver.rootsweb.ancestry.com/th/read/GENEALOGY-DNA/2007-05/1178970992

Therefore the discrepancy I mentioned in post 219.

Dr_McNinja
01-15-2014, 08:19 PM
My Geno 2.0 results came out today ( http://www.anthrogenica.com/showthread.php?1269-Somewhat-unusual-results-for-my-group&p=27190&viewfull=1#post27190 )

Just wondering, what's the SNP difference in the autosomal 23andMe data versus Geno 2? How much fewer does Geno 2 have? Is it that the conversion (DIYDodecad patch) uses fewer or that they test fewer?

toast
01-15-2014, 09:22 PM
I believed mtDNA R has highest diversity in south Asia? So I don't really consider it a West asian mtDNA haplotype.
Regarding Iranian vs Pashtun vs Indian, I do think there is a connection between Iranian and South Indian Brahmin groups. I don't know if this would extend to R-Z2123 or not...
Looks at the G2a types *(from FTDNA and the Afghan paper) compared in this post here http://www.anthrogenica.com/showthread.php?1102-List-of-public-G2a-P303-samples-negative-for-subgroups&p=17711&viewfull=1#post17711

The connection is because South indian brahmins trace their male lineages back to north and west India if I'm not wrong...which are basically the entry points from west asia into south asia. South Indian upper-castes all trace their ancestry back to the north, so these iranians came in the same way that the parsi zoroastrians came to india with the knowledge of the same routes, from there they migrated deeper into India, unless you want to tell me there was a direct route between persia to ancient south India which was more popular.

besides this must have happened a long time back.. long before Iranians accepted Islam, so the connection is so old it's useless in terms of autosomal analysis it could be interpreted as some sort of ancient neolithic signal. I'd like to know why iranians would become absorbed into south indian dravidian brahmins (out of all the brahmin groups in India) in the first place? why not kashmiris who are closer? what are the stories relating iran to south indian brahmins? what are the names of these brahmin groups? do these groups still resemble iranians today?

Sein
01-16-2014, 01:28 PM
Just an additional note concerning Pashtun mtDNA. I found the percentages for the HGDP Pashtuns, but only 19 of them. For the HGDP Pashtuns, 74% of their mtDNA is West Eurasian, 21% of their mtDNA is South Asian, and 5% of their mtDNA is East Asian. Macro-haplogroup R accounts for 63% of HGDP Pashtun mtDNA, macro-haplogroup M accounts for 21% of HGDP Pashtun mtDNA, and macro-haplogroup N accounts for 16% of HGDP Pashtun mtDNA. The most common haplogroups are W and U.

The HGDP results completely caught me by surprise! They are much more West Eurasian than Pan-Pakistani Pashtun samples, and the northern Pashtun samples from Afghanistan. And they are less South Asian than both groups. Of course, a big caveat should be noted. My previous estimates were based on 90 Afghan Pashtuns from Kunduz-Baghlan, and 230 Pakistani Pashtuns from across KP/NWFP+FATA. But my HGDP estimates are based on only 19 individuals. Still, rather surprised at the differences.

Interestingly, based on 21 HGDP Sindhi samples, 45% belong to haplogroup M. 16% of the HGDP Pashtuns are M. That's a huge gap at the Indus. Does anybody know how much mtDNA M Punjabis have?

newtoboard
01-16-2014, 01:33 PM
About 50% mtDNA M for Punjabis I believe.

newtoboard
01-16-2014, 01:37 PM
I believed mtDNA R has highest diversity in south Asia? So I don't really consider it a West asian mtDNA haplotype.
Regarding Iranian vs Pashtun vs Indian, I do think there is a connection between Iranian and South Indian Brahmin groups. I don't know if this would extend to R-Z2123 or not...
Looks at the G2a types *(from FTDNA and the Afghan paper) compared in this post here http://www.anthrogenica.com/showthread.php?1102-List-of-public-G2a-P303-samples-negative-for-subgroups&p=17711&viewfull=1#post17711

I think some people consider it West Asian because of the recent paper on Iranian mtDNA concluded (or did people conclude based on that) that mtDNA R2 (which is the ancestor of JT) originated in Iran. I would wait until more such studies are conducted on populations from the Caucasus, South Asia, Arabian peninsula, Central Asia and the Levant before drawing that conclusion as well as wait for some ancient mtDNA.

newtoboard
01-16-2014, 01:42 PM
Also most R subclades are found in South/Southeast Asia so like I said I would be very hesitant to draw the conclusion that mtdna R2 is West Asian. The older paper on mtDNA in the Asian crossroads was a strong supporter of mtdna R(including R2 I believer) being very old in South Asia.There is some mtDNA M west of the Indus as well as some U2a-U2c which could have been fellow travelers with mtDNA R2.

On a related note it is king of cool to know that there are probably some South Asians out there who have yDNA R1a and mtDNA R1a as well as some with yDNA R2 and mtDNA R2.

soulblighter
01-16-2014, 01:49 PM
The connection is because South indian brahmins trace their male lineages back to north and west India if I'm not wrong...which are basically the entry points from west asia into south asia. South Indian upper-castes all trace their ancestry back to the north, so these iranians came in the same way that the parsi zoroastrians came to india with the knowledge of the same routes, from there they migrated deeper into India, unless you want to tell me there was a direct route between persia to ancient south India which was more popular.

besides this must have happened a long time back.. long before Iranians accepted Islam, so the connection is so old it's useless in terms of autosomal analysis it could be interpreted as some sort of ancient neolithic signal. I'd like to know why iranians would become absorbed into south indian dravidian brahmins (out of all the brahmin groups in India) in the first place? why not kashmiris who are closer? what are the stories relating iran to south indian brahmins? what are the names of these brahmin groups? do these groups still resemble iranians today?

The post had nothing to do with Autosomal DNA, and obviously there are no stories or connections between Iranians and south Indians. Obviously any connection between these sample is extremely old (probably even before Zoroastrianism appeared in present day Iran, and Vedism in India).
Also, it is not clear the origin of that y-DNA G2a motif is necessarily from present day Iran. It could as well be that they appeared in present day Iran and south India from elsewhere (Armenia?, Central Asia?).


Also most R subclades are found in South/Southeast Asia so like I said I would be very hesitant to draw the conclusion that mtdna R2 is West Asian. The older paper on mtDNA in the Asian crossroads was a strong supporter of mtdna R(including R2 I believer) being very old in South Asia.There is some mtDNA M west of the Indus as well as some U2a-U2c which could have been fellow travelers with mtDNA R2.

On a related note it is king of cool to know that there are probably some South Asians out there who have yDNA R1a and mtDNA R1a as well as some with yDNA R2 and mtDNA R2.

I guess we will hold our horses and wait. I don't know what the Japanense are up to with the Farmana DNA, but hopefully a lot of data will come out this year!

newtoboard
01-16-2014, 01:54 PM
G2a is most likely pre Iranian in South India. There were no such things as Iranians when G2a entered South Asia. And if some G2a came via the BMAC it never spoke an Iranian language either. There was likely very little difference between Neolithic Iranian, Anatolian, Mesopotamian and Caucasian groups anyways.

Dr_McNinja
01-16-2014, 01:55 PM
Parasar, my uncle's 37 marker results are in:

http://www.familytreedna.com/public/R1aY-Haplogroup/default.aspx?vgroup=R1aY-Haplogroup&section=yresults

http://www.familytreedna.com/public/R1a/default.aspx?section=yresults

Kit # 311047

What do you think?

newtoboard
01-16-2014, 01:58 PM
The post had nothing to do with Autosomal DNA, and obviously there are no stories or connections between Iranians and south Indians. Obviously any connection between these sample is extremely old (probably even before Zoroastrianism appeared in present day Iran, and Vedism in India).
Also, it is not clear the origin of that y-DNA G2a motif is necessarily from present day Iran. It could as well be that they appeared in present day Iran and south India from elsewhere (Armenia?, Central Asia?).



I guess we will hold our horses and wait. I don't know what the Japanense are up to with the Farmana DNA, but hopefully a lot of data will come out this year!

Do you know which institution was doing the study? Maybe someone can contact them to see what is going on.

parasar
01-16-2014, 05:14 PM
Parasar, my uncle's 37 marker results are in:

http://www.familytreedna.com/public/R1aY-Haplogroup/default.aspx?vgroup=R1aY-Haplogroup§ion=yresults

http://www.familytreedna.com/public/R1a/default.aspx?section=yresults

Kit # 311047

What do you think?

R1a1a 14 25 16 11 11-14 12 12 10 13 11 30 15 9-10 11 11 24 14 20 33 12-14-16-16 11 11 19-23 15 17 17 20 34-39 13 11

Most likely not L657. Z2123 is possible. My advice would be to first test for Z94.

Dr_McNinja
01-16-2014, 05:33 PM
R1a1a 14 25 16 11 11-14 12 12 10 13 11 30 15 9-10 11 11 24 14 20 33 12-14-16-16 11 11 19-23 15 17 17 20 34-39 13 11

Most likely not L657. Z2123 is possible. My advice would be to first test for Z94.Thanks, I'll do that first.

I found this post of yours interesting: http://www.anthrogenica.com/showthread.php?1507-Some-provisional-calculations-for-haplogroup-R1a-based-on-the-first-FGC-result&p=26441&viewfull=1#post26441


What we find for the most part from FTDNA R1a1 and Subclades project dataset:

For L657
DYS456=15
DYS458=16

But for Z93+, L657-
DYS456=16
DYS458=15Are there any other such patterns with the other SNPs? Like Z94+ or Z2124+?

parasar
01-16-2014, 06:24 PM
Thanks, I'll do that first.

I found this post of yours interesting: http://www.anthrogenica.com/showthread.php?1507-Some-provisional-calculations-for-haplogroup-R1a-based-on-the-first-FGC-result&p=26441&viewfull=1#post26441

Are there any other such patterns with the other SNPs? Like Z94+ or Z2124+?

First I always look at the geographical location. If South Asian, I assume Z94+. For Z94+Z2124+ I look for CDY 75 or less. In your uncle's case it is 34+39=73.
The Pakhtoon Z12124 are YF395S1=16,17 (rather than 17,17).

But overall it is difficult to predict as the the R1a1 family/clan was initially very compact, with very few mutations separating the branches we are seeing, so on the STR side there is a lot of overlap.

newtoboard
01-16-2014, 06:27 PM
R1a1a 14 25 16 11 11-14 12 12 10 13 11 30 15 9-10 11 11 24 14 20 33 12-14-16-16 11 11 19-23 15 17 17 20 34-39 13 11

Most likely not L657. Z2123 is possible. My advice would be to first test for Z94.

Why should he test Z94 first? Like you said South Asians are almost always Z94+. There is just one Z93+, Z94- South Asian (a Punjabi from Lahore).

soulblighter
01-16-2014, 06:46 PM
Do you know which institution was doing the study? Maybe someone can contact them to see what is going on.

Toshiki Osada is leading it.
http://archives.chikyu.ac.jp/archives/AnnualReport/Viewer.do?prkbn=P&id=15&jekbn=E
http://www.chikyu.ac.jp/indus/Indus_project/index.html

parasar
01-16-2014, 07:02 PM
Why should he test Z94 first? Like you said South Asians are almost always Z94+. There is just one Z93+, Z94- South Asian (a Punjabi from Lahore).

Just to nail it down. I ran the matches, there are none, and by GD most Z94+ were down the list. The closer ones are Z283, but are likely coincidental.
Plus there is always that Z93+, Z94- possibility, as that just one PJL you mention constitutes ~1.61% of the M17 South Asian samples in 1000Genomes.

Dr_McNinja
01-17-2014, 11:13 PM
Getting back on topic for a second, I added the new samples from the Afghan study to the Harappa map:

https://mapsengine.google.com/map/edit?mid=z1DL_IP3p3uA.kU-eYEC4D4yk

One of them is like directly overlayed with HRP0370 so it's hard to tell the dots apart without zooming in.

Dr_McNinja
01-22-2014, 04:34 PM
I was compiling this for a little while: https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdG8yX1hkLXNVbFVsR1JuOURBdnpad EE&usp=sharing

What do you guys think? Do you have some hg data you can contribute?

parasar
01-22-2014, 06:19 PM
I was compiling this for a little while: https://docs.google.com/spreadsheet/ccc?key=0AuXBmvmgdkfVdG8yX1hkLXNVbFVsR1JuOURBdnpad EE&usp=sharing

What do you guys think? Do you have some hg data you can contribute?

That Kara Khitai ancestry is interesting, and looks supported by the Y type.

Sapporo
01-23-2014, 02:40 AM
@Dr_McNinja

I have another Punjabi Jatt (Sikh) for you to add to that list. His 23andMe name is uday Shergill and his Y-DNA is L3* *L1c-M357 and mt-DNA is M4a. He just sent Zack his raw data the other day. You might want to add Kashmiri Pandits and Punjabi Brahmins to the list as well. I have 2 Punjabi Brahmins and 4-5 Kashmiri Pandits. Also, G Gill's mt-DNA is U7.

Also, have a Punjabi Jatt (N Bains) who has Y-DNA R2 and mt-DNA U2b2 and another Punjabi Jatt (J Lali) who has Y-DNA R1a1a and mt-DNA U5b2a2. I have a bunch of other Punjabi Jatts who you don't have on the list but I'll pm you them later or take screenshots of my 23andMe sharing lists for Y-DNA and mt-DNA and email them to you.

Dr_McNinja
01-23-2014, 03:53 AM
@Dr_McNinja

I have another Punjabi Jatt (Sikh) for you to add to that list. His 23andMe name is uday Shergill and his Y-DNA is L3* *L1c-M357 and mt-DNA is M4a. He just sent Zack his raw data the other day. You might want to add Kashmiri Pandits and Punjabi Brahmins to the list as well. I have 2 Punjabi Brahmins and 4-5 Kashmiri Pandits. Also, G Gill's mt-DNA is U7.

Also, have a Punjabi Jatt (N Bains) who has Y-DNA R2 and mt-DNA U2b2 and another Punjabi Jatt (J Lali) who has Y-DNA R1a1a and mt-DNA U5b2a2. I have a bunch of other Punjabi Jatts who you don't have on the list but I'll pm you them later or take screenshots of my 23andMe sharing lists for Y-DNA and mt-DNA and email them to you.Thanks! I figured you'd have a bunch :P PM me the other Jatts, Pandits, Brahmins, anyone with an established surname/clan affiliation from South Asia and I'll add them.

Dr_McNinja
02-05-2014, 11:58 AM
I averaged together rough coordinates (very rough, I used Lahore for all Punjabis) for the top 60 individuals in my One-To-Many Gene Comparison list on 23andMe and got these coordinates:

29.9836666667, 70.0466666667

Did the same for HRP0370 (with 58 individuals) and got this:

34.1756896552, 63.5103448276

That turned out pretty well. I could've been much more thorough since many of the 23andMe users actually give exact ancestral locations.

I'm sharing with something like 120 individuals, including many from the Caucasus region, Iran, and a few Europeans (and a bunch of Bengalis, South Indians, Nepal, etc).

Here were the locations:

Me:
Closest (76.00-75.95) - (Uttar Pradesh/Punjab, Jatt, Bengali, Jatt)
Next closest - (75.95-75.90) - (Pahari Jatt, Jatt, Jatt)
Third (75.90-75.85) - (Punjabi Pathan, Gujarat, Kashmir)
Fourth (75.85-75.80) - (Rajasthan Brahmin, Karnatka Brahmin, Punjabi Khatri, Rajasthan Jatt, Punjab Jatt, Khan, Jatt, Punjab Brahmin, Jatt)
Fifth (75.80-75.75) - (Pakistan, Jatt, Jatt, Rajasthan Jatt, Pakistan, Afghanistan, South India, Rajasthan, Andhra Pradesh, Haryana Jatt, Punjabi Khatri)
Sixth (75.75-75.70) - (Jatt, Jatt, Punjab, Punjab Chamar, Pakhtun, Pakhtun, Jatt, Haryana Jatt, Afghan/Kyrgyz, Nepal)
Seventh (75.70-75.65) - (Jatt/English, Nepal, Nepal, Pakistan/Punjab, Punjabi Arora, Pathan, Punjab Kamboj)
Eighth (75.65-75.60) - (Bengali, Punjab, Pakistan, Afghanistan, Iran, Jatt/English)
Ninth (75.60-75.55) - (Iran, Afghanistan, South India, Punjab, Balochistan, Georgia, Europe)^ I started off planning to weight these then just averaged them all together anyway

HRP0370:
Pakhtun + Pathan + Afghanistan + Punjab + Punjab + English/Jatt + Afghan/Kyrgyz + Iran + Punjab + Punjab + Iran + Pakhtun + Punjab + Punjab + Pathan + Punjab + Punjab + Punjab + Iran + Pathan + Haryana + Pathan + English/Jatt + Punjab + Punjab + Europe + Kurd + Pakistan + Punjab + Pahari + Haryana + Punjab + Punjab + Caucasus + Pathan + Punjab + Uttar Pradesh + Rajasthan + Iraq + Caucasus + Pakistan + Afganistan + Caucasus + Pakistan + Kurd + Pakistan + Caucasus + Punjab + Caucasus + Balochistan + Caucasus + Nepal + South India + Europe + Rajasthan + Caucasus + Gujarat + Afghanistan + Europe

MitchellSince1893
03-10-2016, 06:11 PM
FYI, it appears SPA has been updated as my parents and I get totally different results compared to 2013.

firemonkey
03-18-2016, 09:23 PM
Finally got it to work thanks to instructions from petalpusher at Apricity. Seems accurate.

8217

Aldric
03-01-2018, 11:16 PM
Not bad...

https://s13.postimg.org/r2o5o7kav/Untitled.jpg

Chromosome breakdown:

genotype<-read.table("adna.txt",as.is=T)
index<-which(genotype[,2]==1);gen1<-genotype[index,];write.table(gen1,"adna _1.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==2);gen2<-genotype[index,];write.table(gen2,"adna _2.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==3);gen3<-genotype[index,];write.table(gen3,"adna _3.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==4);gen4<-genotype[index,];write.table(gen4,"adna _4.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==5);gen5<-genotype[index,];write.table(gen5,"adna _5.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==6);gen6<-genotype[index,];write.table(gen6,"adna _6.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==7);gen7<-genotype[index,];write.table(gen7,"adna _7.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==8);gen8<-genotype[index,];write.table(gen8,"adna _8.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==9);gen9<-genotype[index,];write.table(gen9,"adna _9.txt",sep="\t",quote =F,row.names=F,col.names=F)
index<-which(genotype[,2]==10);gen10<-genotype[index,];write.table(gen10,"adna _10.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==11);gen11<-genotype[index,];write.table(gen11,"adna _11.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==12);gen12<-genotype[index,];write.table(gen12,"adna_12.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==13);gen13<-genotype[index,];write.table(gen13,"adna _13.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==14);gen14<-genotype[index,];write.table(gen14,"adna _14.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==14);gen14<-genotype[index,];write.table(gen14,"adna _14.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==15);gen15<-genotype[index,];write.table(gen15,"adna _15.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==16);gen16<-genotype[index,];write.table(gen16,"adna _16.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==17);gen17<-genotype[index,];write.table(gen17,"adna _17.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==18);gen18<-genotype[index,];write.table(gen18,"adna _18.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==19);gen19<-genotype[index,];write.table(gen19,"adna _19.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==20);gen20<-genotype[index,];write.table(gen20,"adna _20.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==21);gen21<-genotype[index,];write.table(gen21,"adna _21.txt",sep="\t",quote=F,row.names=F,col.names=F)
index<-which(genotype[,2]==22);gen22<-genotype[index,];write.table(gen22,"adna_22.txt",sep="\t",quote=F,row.names=F,col.names=F)


*Change .txt file name*

Chromosome breakdown:

https://s13.postimg.org/q8bb0esbb/Untitled.jpg

FIREYWOTAN
07-06-2018, 09:00 AM
Thank you for opening up a field of all possibilities. Recent discoveries and there announcements sounded like it might fit.
Abstract
Homo naledi is a previously-unknown species of extinct hominin discovered within the Dinaledi Chamber of the Rising Star cave system, Cradle of Humankind, South Africa. This species is characterized by body mass and stature similar to small-bodied human populations but a small endocranial volume similar to australopiths. Cranial morphology of H. naledi is unique, but most similar to early Homo species including Homo erectus, Homo habilis or Homo rudolfensis. While primitive, the dentition is generally small and simple in occlusal morphology. H. naledi has humanlike manipulatory adaptations of the hand and wrist. It also exhibits a humanlike foot and lower limb. These humanlike aspects are contrasted in the postcrania with a more primitive or australopith-like trunk, shoulder, pelvis and proximal femur. Representing at least 15 individuals with most skeletal elements repeated multiple times, this is the largest assemblage of a single species of hominins yet discovered in Africa.

https://doi.org/10.7554/eLife.09560.001
40 Open annotations. The current annotation count on this page is 40.
eLife digest
Modern humans, or Homo sapiens, are now the only living species in their genus. But as recently as 100,000 years ago, there were several other species that belonged to the genus Homo. Together with modern humans, these extinct human species, our immediate ancestors and their close relatives, are collectively referred to as ‘hominins’.

Now Berger et al. report the recent discovery of an extinct species from the genus Homo that was unearthed from deep underground in what has been named the Dinaledi Chamber, in the Rising Star cave system in South Africa. The species was named Homo naledi; ‘naledi’ means ‘star’ in Sotho (also called Sesotho), which is one of the languages spoken in South Africa.

The unearthed fossils were from at least 15 individuals and include multiple examples of most of the bones in the skeleton. Based on this wide range of specimens from a single site, Berger et al. describe Homo naledi as being similar in size and weight to a small modern human, with human-like hands and feet. Furthermore, while the skull had several unique features, it had a small braincase that was most similar in size to other early hominin species that lived between four million and two million years ago. Homo naledi's ribcage, shoulders and pelvis also more closely resembled those of earlier hominin species than those of modern humans.

The Homo naledi fossils are the largest collection of a single species of hominin that has been discovered in Africa so far and, in a related study, Dirks et al. describe the setting and context for these fossils. However, since the age of the fossils remains unclear, one of the next challenges will be to date the remains to provide more information about the early evolution of humans and their close relatives.

https://doi.org/10.7554/eLife.09560.002
Introduction
Fossil hominins were first recognized in the Dinaledi Chamber in the Rising Star cave system in October 2013. During a relatively short excavation, our team recovered an extensive collection of 1550 hominin specimens, representing nearly every element of the skeleton multiple times (Figure 1), including many complete elements and morphologically informative fragments, some in articulation, as well as smaller fragments many of which could be refit into more complete elements. The collection is a morphologically homogeneous sample that can be attributed to no previously-known hominin species. Here we describe this new species, Homo naledi. We have not defined H. naledi narrowly based on a single jaw or skull because the entire body of material has informed our understanding of its biology.