PDA

View Full Version : Jewish populations in GEDMatch?



pakistani
07-05-2020, 08:11 PM
Hey all, these are my HarappaWorld results:

Population %:

1 S-Indian 35.31
2 Baloch 32.74
3 NE-Euro 8.71
4 Caucasian 6.9
5 Mediterranean 6.11
6 SW-Asian 3.83
7 NE-Asian 2.83
8 Siberian 2.59
9 Beringian 0.74
10 Papuan 0.11
11 San 0.08
12 American 0.04


Single Population:

1 up-muslim (harappa) 5.87
2 nepalese-a (xing) 8.51
3 bengali-brahmin (harappa) 8.61
4 up-brahmin (harappa) 8.89
5 punjabi (harappa) 9.18
6 bihari-muslim (harappa) 9.34
7 gujarati-muslim (harappa) 9.4
8 punjabi-ramgarhia (harappa) 9.65
9 singapore-indian-c (sgvp) 9.73
10 brahmin-uttar-pradesh (metspalu) 9.89
11 kashmiri-pahari (harappa) 9.95
12 punjabi-brahmin (harappa) 10.14
13 punjabi-jatt-muslim (harappa) 10.36
14 kashmiri-pandit (reich) 10.38
15 vaish (reich) 10.6
16 kashmiri (harappa) 10.83
17 up-kshatriya (metspalu) 11.59
18 punjabi-jatt-sikh (harappa) 11.64
19 cochin-jew (behar) 11.8
20 karnataka-brahmin (harappa) 11.82


Mixed Mode Population:

1 88.9% up-brahmin (harappa) + 11.1% morocco-n (henn2012) @ 4.16
2 88.5% up-brahmin (harappa) + 11.5% algeria (henn2012) @ 4.21
3 82.8% up-kshatriya (metspalu) + 17.2% morocco-jew (behar) @ 4.24
4 89.5% up-brahmin (harappa) + 10.5% tunisia (henn2012) @ 4.28
5 89% up-brahmin (harappa) + 11% mozabite (hgdp) @ 4.3
6 88.4% up-brahmin (harappa) + 11.6% libya (henn2012) @ 4.33
7 89.2% up-brahmin (harappa) + 10.8% saharawi (henn2012) @ 4.36
8 88.7% up-brahmin (harappa) + 11.3% moroccan (behar) @ 4.39
9 80.7% gujarati (harappa) + 19.3% ashkenazi (harappa) @ 4.4
10 82.5% up-kshatriya (metspalu) + 17.5% ashkenazy-jew (behar) @ 4.41
11 81.3% gujarati (harappa) + 18.7% tuscan (1000genomes) @ 4.43
12 82.3% gujarati-b (hapmap) + 17.7% morocco-jew (behar) @ 4.44
13 80.6% gujarati (harappa) + 19.4% ashkenazy-jew (behar) @ 4.44
14 81.9% gujarati-b (hapmap) + 18.1% ashkenazy-jew (behar) @ 4.45
15 82.1% gujarati-b (hapmap) + 17.9% ashkenazi (harappa) @ 4.45
16 83% up-kshatriya (metspalu) + 17% sephardic-jew (behar) @ 4.46
17 78.4% kerala-nair (harappa) + 21.6% romanian-a (behar) @ 4.49
18 87.6% bengali-brahmin (harappa) + 12.4% morocco-jew (behar) @ 4.5
19 63.3% dharkar (metspalu) + 36.7% tajik (yunusbayev) @ 4.52
20 84.5% vaish (reich) + 15.5% morocco-jew (behar) @ 4.58

I'm mainly surprised by the amount of secondary Jewish populations in the mixed mode sharing, does anyone know how I should interpret them? I'm a Pakistani for reference and have no known Jewish or North African ancestry (I'm almost certain of it on my mother's side, but much less so on my father's beyond two or three generations)

Kulin
07-05-2020, 08:14 PM
The combination of relatively high Mediterranean and SW Asian is resulting in those populations showing up.

jkotl0327
07-05-2020, 08:18 PM
Hey all, these are my full HarappaWorld results:

Population %:

1 S-Indian 35.31
2 Baloch 32.74
3 NE-Euro 8.71
4 Caucasian 6.9
5 Mediterranean 6.11
6 SW-Asian 3.83
7 NE-Asian 2.83
8 Siberian 2.59
9 Beringian 0.74
10 Papuan 0.11
11 San 0.08
12 American 0.04


Single Population:

1 up-muslim (harappa) 5.87
2 nepalese-a (xing) 8.51
3 bengali-brahmin (harappa) 8.61
4 up-brahmin (harappa) 8.89
5 punjabi (harappa) 9.18
6 bihari-muslim (harappa) 9.34
7 gujarati-muslim (harappa) 9.4
8 punjabi-ramgarhia (harappa) 9.65
9 singapore-indian-c (sgvp) 9.73
10 brahmin-uttar-pradesh (metspalu) 9.89
11 kashmiri-pahari (harappa) 9.95
12 punjabi-brahmin (harappa) 10.14
13 punjabi-jatt-muslim (harappa) 10.36
14 kashmiri-pandit (reich) 10.38
15 vaish (reich) 10.6
16 kashmiri (harappa) 10.83
17 up-kshatriya (metspalu) 11.59
18 punjabi-jatt-sikh (harappa) 11.64
19 cochin-jew (behar) 11.8
20 karnataka-brahmin (harappa) 11.82


Mixed Mode Population:

1 88.9% up-brahmin (harappa) + 11.1% morocco-n (henn2012) @ 4.16
2 88.5% up-brahmin (harappa) + 11.5% algeria (henn2012) @ 4.21
3 82.8% up-kshatriya (metspalu) + 17.2% morocco-jew (behar) @ 4.24
4 89.5% up-brahmin (harappa) + 10.5% tunisia (henn2012) @ 4.28
5 89% up-brahmin (harappa) + 11% mozabite (hgdp) @ 4.3
6 88.4% up-brahmin (harappa) + 11.6% libya (henn2012) @ 4.33
7 89.2% up-brahmin (harappa) + 10.8% saharawi (henn2012) @ 4.36
8 88.7% up-brahmin (harappa) + 11.3% moroccan (behar) @ 4.39
9 80.7% gujarati (harappa) + 19.3% ashkenazi (harappa) @ 4.4
10 82.5% up-kshatriya (metspalu) + 17.5% ashkenazy-jew (behar) @ 4.41
11 81.3% gujarati (harappa) + 18.7% tuscan (1000genomes) @ 4.43
12 82.3% gujarati-b (hapmap) + 17.7% morocco-jew (behar) @ 4.44
13 80.6% gujarati (harappa) + 19.4% ashkenazy-jew (behar) @ 4.44
14 81.9% gujarati-b (hapmap) + 18.1% ashkenazy-jew (behar) @ 4.45
15 82.1% gujarati-b (hapmap) + 17.9% ashkenazi (harappa) @ 4.45
16 83% up-kshatriya (metspalu) + 17% sephardic-jew (behar) @ 4.46
17 78.4% kerala-nair (harappa) + 21.6% romanian-a (behar) @ 4.49
18 87.6% bengali-brahmin (harappa) + 12.4% morocco-jew (behar) @ 4.5
19 63.3% dharkar (metspalu) + 36.7% tajik (yunusbayev) @ 4.52
20 84.5% vaish (reich) + 15.5% morocco-jew (behar) @ 4.58

I'm mainly surprised by the amount of secondary Jewish populations in the mixed mode sharing, does anyone know how I should interpret them? I'm a Pakistani for reference and have no known Jewish or North African ancestry that I'm aware of (I'm almost certain of it on my mother's side, but much less so on my father's beyond two or three generations).

It seems to be plotting you as mostly (80-85% S. Asian) with the rest from a MENA source. Pakistanis have some MENA admixture, albeit not from Jews or N. Africans but probably from Afghanis and E. Iranic peoples. It's just that rather than placing you as 60% S. Asian and 40% Afghan, or whatever the proportions of a Pakistani usually are, it seems to think that giving you more S. Asian and less MENA, just with a more Western-shifted MENA group is a better fit for you. You don't have to interpret this literally. In fact, the 19th line that gives you 63% S. Asian and 37% Central Asian is probably the most accurate here in my opinion. You should get an opinion from someone like Tipirneni who is more informed about this though.

pakistani
07-05-2020, 08:27 PM
It seems to be plotting you as mostly (80-85% S. Asian) with the rest from a MENA source. Pakistanis have some MENA admixture, albeit not from Jews or N. Africans but probably from Afghanis and E. Iranic peoples. It's just that rather than placing you as 60% S. Asian and 40% Afghan, or whatever the proportions of a Pakistani usually are, it seems to think that giving you more S. Asian and less MENA, just with a more Western-shifted MENA group is a better fit for you. You don't have to interpret this literally. In fact, the 19th line that gives you 63% S. Asian and 37% Central Asian is probably the most accurate here in my opinion. You should get an opinion from someone like Tipirneni who is more informed about this though.

The MENA ancestors I know of came from Hadhramaut in the 1500s and Baghdad in the 1700s. I have ancestors who came from Bukhara between the 13th and 15th centuries. I potentially have recent Pashtun ancestry so that could also explain it. I don't know how useful it is but I ran EthioHelix K10 + Palestinian and got these results:

1 libya 5.62
2 egyptans 5.91
3 egypt 7.4
4 algeria 13.8
5 sahara-occ 15.02
6 moroccans 15.82
7 morocco-n 15.84
8 mozabite 18.27
9 palestinian 24.6
10 morocco-s 30.52

jkotl0327
07-05-2020, 08:35 PM
The MENA ancestors I know of came from Hadhramaut in the 1500s and Baghdad in the 1700s. I have ancestors who came from Bukhara between the 13th and 15th centuries. I potentially have recent Pashtun ancestry so that could also explain it. I don't know how useful it is but I ran EthioHelix K10 + Palestinian and got these results:

1 libya 5.62
2 egyptans 5.91
3 egypt 7.4
4 algeria 13.8
5 sahara-occ 15.02
6 moroccans 15.82
7 morocco-n 15.84
8 mozabite 18.27
9 palestinian 24.6
10 morocco-s 30.52

That is pretty ancient so theoretically it should be barely noticeable. Ethiohelix + Pal is for mixed MENA and African people so probably not great for you. If you want to find out if you are more MENA shifted than average, I would use Vahaduo GEDMATCH calculators and try HarappaWorld updated. For sources use either Gujjar Pakistan or Baloch Pakistan (I don't know which one you are), Yemenese, and Iraqi Arab. Also, to account for the possibility that you are actually more S. Asian than the average Pakistani, I would use an Indian component (maybe Uttar Pradesh). See what happens.

jkotl0327
07-05-2020, 08:53 PM
The MENA ancestors I know of came from Hadhramaut in the 1500s and Baghdad in the 1700s. I have ancestors who came from Bukhara between the 13th and 15th centuries. I potentially have recent Pashtun ancestry so that could also explain it. I don't know how useful it is but I ran EthioHelix K10 + Palestinian and got these results:

1 libya 5.62
2 egyptans 5.91
3 egypt 7.4
4 algeria 13.8
5 sahara-occ 15.02
6 moroccans 15.82
7 morocco-n 15.84
8 mozabite 18.27
9 palestinian 24.6
10 morocco-s 30.52

I was a bit curious so I tried it myself. Using Gujjar Pakistan, you came out more India-shifted than average, and it seems that your best model is 99.6% Uttar Pradesh Muslim and 0.4% Yemenese. Even though those are Muslims, they are Indian Muslims so I would say that you are more India-shifted than the average Pakistani (maybe E. Pakistani or recent Indian ancestors?) with the possibility of some distant Yemen or Arab ancestry. Uttar Pradesh Muslims by the way come out as about two-thirds Pakistani and one-third Hindu Uttar Pradesh.

pakistani
07-05-2020, 08:56 PM
I was a bit curious so I tried it myself. Using Gujjar Pakistan, you came out more India-shifted than average, and it seems that your best model is 99.6% Uttar Pradesh Muslim and 0.4% Yemenese. Even though those are Muslims, they are Indian Muslims so I would say that you are more India-shifted than the average Pakistani (maybe E. Pakistani or recent Indian ancestors?) with the possibility of some distant Yemen or Arab ancestry. Uttar Pradesh Muslims by the way come out as about two-thirds Pakistani and one-third Hindu Uttar Pradesh.


Correct, both sides of my family migrated from Delhi and UP after 1947. My 23andMe results for reference:

38278

MonkeyDLuffy
07-05-2020, 09:12 PM
I was a bit curious so I tried it myself. Using Gujjar Pakistan, you came out more India-shifted than average, and it seems that your best model is 99.6% Uttar Pradesh Muslim and 0.4% Yemenese. Even though those are Muslims, they are Indian Muslims so I would say that you are more India-shifted than the average Pakistani (maybe E. Pakistani or recent Indian ancestors?) with the possibility of some distant Yemen or Arab ancestry. Uttar Pradesh Muslims by the way come out as about two-thirds Pakistani and one-third Hindu Uttar Pradesh.

What's the average Pakistani?

jkotl0327
07-05-2020, 09:17 PM
Correct, both sides of my family migrated from Delhi and UP after 1947. My 23andMe results for reference:

38278

It's interesting that your top two regions are more western than your actual origin, maybe because you are Muslim? The NW Asian might be evidence of the Iraqi and the broadly Arab of Yemen ancestry but it also might be too distant to detect. Also, running you as UP Muslim and Egypt actually gives a smaller distance, and you get 98.6% UP Muslim with 1.4% Egypt, which is exactly the total of your Arab/Egyptian/Levantine on 23andMe. You've never heard anything about Egyptian ancestry?

Magnetic
07-05-2020, 09:30 PM
I have jewish shift too lol

85.3% Kurdish + 14.7% Sephardic_Jewish @ 2.02
85.6% Kurdish + 14.4% Tunisian_Jewish @ 2.09
86.5% Kurdish + 13.5% Italian_Jewish @ 2.13
86.7% Kurdish + 13.3% Ashkenazi @ 2.46

83.1% Kurd (Dodecad) + 16.9% Ashkenazi (Dodecad) @ 1.34
83.4% Kurd (Dodecad) + 16.6% Ashkenazy_Jews @ 1.44
81.7% Kurd (Dodecad) + 18.3% Sephardic_Jews (Behar) @ 1.52

pakistani
07-05-2020, 09:46 PM
It's interesting that your top two regions are more western than your actual origin, maybe because you are Muslim? The NW Asian might be evidence of the Iraqi and the broadly Arab of Yemen ancestry but it also might be too distant to detect. Also, running you as UP Muslim and Egypt actually gives a smaller distance, and you get 98.6% UP Muslim with 1.4% Egypt, which is exactly the total of your Arab/Egyptian/Levantine on 23andMe. You've never heard anything about Egyptian ancestry?

I can almost guarantee I have no Egyptian ancestry on my mother's side, or at least any recent enough to show up. I'm thinking it might be on my father's side, specifically on a maternal branch, since I don't know much there beyond a couple generations. Two of my great-great grandfathers on that side claimed to have Pashtun ancestry, one had the surname Syed (all three were from UP), and the last one along with my great-great grandmothers I know nothing about.

jkotl0327
07-05-2020, 10:16 PM
What's the average Pakistani?

I used Gujjar Pakistan, as I'm not an expert on Pakistan, that may not have been the ideal component to use. The only other one there was Baloch Pakistan which seemed too western-shifted to be accurate for "average Pakistani."

jkotl0327
07-05-2020, 10:17 PM
I have jewish shift too lol

85.3% Kurdish + 14.7% Sephardic_Jewish @ 2.02
85.6% Kurdish + 14.4% Tunisian_Jewish @ 2.09
86.5% Kurdish + 13.5% Italian_Jewish @ 2.13
86.7% Kurdish + 13.3% Ashkenazi @ 2.46

83.1% Kurd (Dodecad) + 16.9% Ashkenazi (Dodecad) @ 1.34
83.4% Kurd (Dodecad) + 16.6% Ashkenazy_Jews @ 1.44
81.7% Kurd (Dodecad) + 18.3% Sephardic_Jews (Behar) @ 1.52

No known Jewish ancestry? Or are Kurds and Jews close enough that this might not really mean much?

MonkeyDLuffy
07-05-2020, 10:51 PM
I used Gujjar Pakistan, as I'm not an expert on Pakistan, that may not have been the ideal component to use. The only other one there was Baloch Pakistan which seemed too western-shifted to be accurate for "average Pakistani."

He certainly has more than 1 source of West Asian ancestry. Indian muslims, especially the ones with legit ancestry from west asia are hard to model with accurate populations if there are no proper documentation of their family line. The jewish populations showing up are result of High SW asian he carries.

In terms of the Kurdish user result, I'm not sure what those kurd samples are based on, probably Iraqi kurd. So I'd expect a Turkish Kurd to have extra Baltic/Southern European like mix, which would show up as diiferent European Jews in oracles.

StillWater
07-05-2020, 11:32 PM
In terms of the Kurdish user result, I'm not sure what those kurd samples are based on, probably Iraqi kurd. So I'd expect a Turkish Kurd to have extra Baltic/Southern European like mix, which would show up as diiferent European Jews in oracles.

He'd then probably score Balkan instead. He is scoring Western Jewish populations because of an East Med shift.

Magnetic
07-05-2020, 11:46 PM
No known Jewish ancestry? Or are Kurds and Jews close enough that this might not really mean much?

no I dont have any known jewish ancestry . my haplo is also common among jews probably . E-L29 . but not sure


He'd then probably score Balkan instead. He is scoring Western Jewish populations because of an East Med shift.

well we kurds are a people with high east med :) you probably mean extra shift towards the east med . probably . I am geographically from a northwest province maybe that is part of the reason dunno löl

Magnetic
07-05-2020, 11:57 PM
btw.

mdlp world 22

# Primary Population (source) Secondary Population (source) Distance

1 79.3% Kurd (derived) + 20.7% Ashkenazim (derived) @ 3.07
2 79.6% Kurd (derived) + 20.4% Jew_Francestrale (derived) @ 3.37
3 78.9% Kurd (derived) + 21.1% Jew_Italia (derived) @ 3.42

jkotl0327
07-05-2020, 11:58 PM
He certainly has more than 1 source of West Asian ancestry. Indian muslims, especially the ones with legit ancestry from west asia are hard to model with accurate populations if there are no proper documentation of their family line. The jewish populations showing up are result of High SW asian he carries.

In terms of the Kurdish user result, I'm not sure what those kurd samples are based on, probably Iraqi kurd. So I'd expect a Turkish Kurd to have extra Baltic/Southern European like mix, which would show up as diiferent European Jews in oracles.

I see what you are saying, he was very close to Uttar Pradesh Muslim, but is the MENA ancestry of Pakistanis/Indian Muslims more SW Asian or E. Iranic? When I think of SW Asia I think of the Levant/Mesopotamia area, but I always thought that most of the MENA would have been from Afghanistan/Iran. I have to agree with Stillwater though, the Jewish from Magnetic's results seems more likely to reflect East Med than Balkan ancestry. Also, Jews don't have a lot of Baltic, especially Sephardi Jews, which he also got.

Magnetic
07-06-2020, 12:04 AM
the Jewish from Magnetic's results seems more likely to reflect East Med than Balkan ancestry. Also, Jews don't have a lot of Baltic, especially Sephardi Jews, which he also got.

I dont understand what you mean . we kurds have east med ancestry by default . or do you mean that I have EXTRA ancestry from the med coast ?

as you see I score a lot of anatolia and levant

compare with the lebanese . I have even a little more Anatolian than them

me :

https://up.picr.de/38914124oi.jpg

Lebanese :

https://up.picr.de/38917810tg.jpg


https://up.picr.de/38917811eh.jpg

Buxoro
07-06-2020, 12:08 AM
The MENA ancestors I know of came from Hadhramaut in the 1500s and Baghdad in the 1700s. I have ancestors who came from Bukhara between the 13th and 15th centuries. I potentially have recent Pashtun ancestry so that could also explain it. I don't know how useful it is but I ran EthioHelix K10 + Palestinian and got these results:

1 libya 5.62
2 egyptans 5.91
3 egypt 7.4
4 algeria 13.8
5 sahara-occ 15.02
6 moroccans 15.82
7 morocco-n 15.84
8 mozabite 18.27
9 palestinian 24.6
10 morocco-s 30.52

Bukharan people are not Mena, back then and now
EthioHelix is unreliable for you and for most people too

Harappa world tends to give a lot of people with some med/sw asian jewish admixture results, your case is not out of ordinary

Kulin
07-06-2020, 12:10 AM
I see what you are saying, he was very close to Uttar Pradesh Muslim, but is the MENA ancestry of Pakistanis/Indian Muslims more SW Asian or E. Iranic? When I think of SW Asia I think of the Levant/Mesopotamia area, but I always thought that most of the MENA would have been from Afghanistan/Iran. I have to agree with Stillwater though, the Jewish from Magnetic's results seems more likely to reflect East Med than Balkan ancestry. Also, Jews don't have a lot of Baltic, especially Sephardi Jews, which he also got.

User 'Pakistani' claims Syed ancestry (which is how the "jewish-like" signal is showing up in his results). He's allegedly of Bukhari and Hadhramuti/yemeni Syed lineage. Syed = descendants of Muhammad. Most South Asian Syeds have ancestry from places in Central Asia/Iran (e.g. Bukhara/Samarqand/Nishapur/Gilan etc) or from Hadhramut in Yemen.

jkotl0327
07-06-2020, 12:41 AM
I dont understand what you mean . we kurds have east med ancestry by default . or do you mean that I have EXTRA ancestry from the med coast ?

as you see I score a lot of anatolia and levant

compare with the lebanese . I have even a little more Anatolian than them

me :

https://up.picr.de/38914124oi.jpg

Lebanese :

https://up.picr.de/38917810tg.jpg


https://up.picr.de/38917811eh.jpg

Yes I meant that maybe you are more E. Med shifted than the average Kurd.

jkotl0327
07-06-2020, 12:44 AM
User 'Pakistani' claims Syed ancestry (which is how the "jewish-like" signal is showing up in his results). He's allegedly of Bukhari and Hadhramuti/yemeni Syed lineage. Syed = descendants of Muhammad. Most South Asian Syeds have ancestry from places in Central Asia/Iran (e.g. Bukhara/Samarqand/Nishapur/Gilan etc) or from Hadhramut in Yemen.

Yeah I didn't really know what syed was, but he could be modeled as 99.6% Uttar Pradesh Muslim and 0.4% Yemen. That might just be a slight Arabian-shift, not necessarily specific Yemeni ancestry of course. He did say that his ancestry is really UP/Delhi Muslims.

MonkeyDLuffy
07-06-2020, 01:25 AM
Yes I meant that maybe you are more E. Med shifted than the average Kurd.

Kurds are diverse bunch, I'd expect Turkish kurds to be a bit more leaning towards Europe than Iraqi kurds.