Haplogroups of Moroccan Jews- our project results

[Iẓeḍwan n Nanna-Tuda]

You say G-Z3428 is Iberian in the OP. Yet, it's not even rare amongst Ashkenazim. I also found a Lebanese with it. As for R1b-M269, 7% of Lebanese have it and so do 10% of Iraqis. The following reasoning "E-V13 and some subclades of J2 (like the Illyrian one) are very uncommon in the levant." is what lead to thinking Ashkenazi R1a was European/Khazar. As for studies on Moroccan Jews, they state that the YDNA profile is similar to that of Ashkenazim. I also played with the NMonte calculator on Moroccan Jews and no matter what Italian/Spanish group I chose, I couldn't get the Levantine below 44%. I'll grant you that I2a is probably European.

G-Z3428 : You found 1 (only 1) Lebanese BUT there are hundreds of goyim from NW Africa (10 people as far as I know), Iberia, Italy (and Sardinia), Germany, Austria, Switzerland, England and Netherlands who carry it too. It's much more common in SW Europe and NW Africa than anywhere else. (Cf. some FTDNA's projects such as the G-CTS342's one). Indeed, you could find plenty of Ashkenazim, some Sephardi but 0 Mizrahi ! Curious, isn't it ? 0 Egyptian, 0 Arabian (although they take plenty of tests), 0 Iraki, hardly 1 or 2 people from the most Mediterranean part of the Middle East : the Levant (Crusaders?). You would even have very hard time to find it in SW Europe (Greece, Albania...) !
G-Z3428 has been found in SW Sicily (Bronze Age) ! Its mother clade, G-Z1903, has been found 4 times in SW European's Chalcolithic : in the Iberian (1) and Sicilian Chalcolithic (2) (in Lucena, Southern Spain more precisely and in SW Sicily)... Not to talk about its presence in the Bulgarian and Hungarian Neolithic and Chalcolithic (3)... This clade is European and has to be considered as such. It occurs among European Jews only!
Yes, indeed G-Z3428 is mainly Spanish and Italian. It might reflect Iberian and Italian conversions to Judaism !

G-Z6028 (its great-granddaughter's clade ) has had a strong founder effecter among Ashkenazim (TMRCA : 1200 years!) although it's probably Iberian/West Med, yes, but did you know that some Ashkenazi had Sephardi ancestry ? It would explain :
1/ why is E-L19 (especially E-M81) found among them (4) ?
2/ why is the NA mutation of Parkinson commong among them (5)?
3/ why they carry G-Z3428
4/why they carry R-DF27 (which is 100% Iberian)
5/ partially why more than 81% of Ashkenazim mtDNAs are European (and trace back to Western Europe) (6)

Why didn't you mention that the Lebanese G-Z6028 is closer to the Mende from Sierra Leone than he's to EUROPEAN Jews ? The link between Lebanon and the Sierra Leone might be NW Africa, which means that G-Z6028 is everything but Levantine!

R-M269 :It's indeed present in the Levant (although very rare). Its frequency among Moroccan Jews is substantially higher than in the Levant (18% is everything but low) and they mainly carry SW European such as R-U106 or R-DF27 (which for sure aren't Levantine). Don't forget that the Levant got some European admixture (Romans, crusaders, corsairs)... This European admixture is very likely stronger than it used to be in the Levant prior to the Jewish diaspora

R-DF27 is Iberian. How would you explain that one of its subclade is exclusively found among Sephardim and Ashkenazim (TMRCA : 1150 years!) ? https://www.yfull.com/tree/R-V3476/
Same founder effect as for G-Z6028 !

So, yes, Moroccan Jews are mainly Levantine from their paternal side (something like 55-60%, not more though) but they have a strong West European and NW African contribution!


(1) The genomic history of the Iberian Peninsula reconstructed
(2) The Arrival of Steppe and Iranian Related Ancestry in the Islands of the Western Mediterranean
(3) Eupedia
(4)


(5) https://journals.plos.org/plosone/ar...l.pone.0181335
(6) https://www.nature.com/articles/ncomms3543

[ModernMaskil]

Great thread. Thank you.

[Pine]

G-Z3428 : You found 1 (only 1) Lebanese BUT there are hundreds of goyim from NW Africa (10 people as far as I know), Iberia, Italy (and Sardinia), Germany, Austria, Switzerland, England and Netherlands who carry it too. It's much more common in SW Europe and NW Africa than anywhere else. (Cf. some FTDNA's projects such as the G-CTS342's one). Indeed, you could find plenty of Ashkenazim, some Sephardi but 0 Mizrahi ! Curious, isn't it ? 0 Egyptian, 0 Arabian (although they take plenty of tests), 0 Iraki, hardly 1 or 2 people from the most Mediterranean part of the Middle East : the Levant (Crusaders?). You would even have very hard time to find it in SW Europe (Greece, Albania...) !
G-Z3428 has been found in SW Sicily (Bronze Age) ! Its mother clade, G-Z1903, has been found 4 times in SW European's Chalcolithic : in the Iberian (1) and Sicilian Chalcolithic (2) (in Lucena, Southern Spain more precisely and in SW Sicily)... Not to talk about its presence in the Bulgarian and Hungarian Neolithic and Chalcolithic (3)... This clade is European and has to be considered as such. It occurs among European Jews only!
Yes, indeed G-Z3428 is mainly Spanish and Italian. It might reflect Iberian and Italian conversions to Judaism !

G-Z6028 (its great-granddaughter's clade ) has had a strong founder effecter among Ashkenazim (TMRCA : 1200 years!) although it's probably Iberian/West Med, yes, but did you know that some Ashkenazi had Sephardi ancestry ? It would explain :
1/ why is E-L19 (especially E-M81) found among them (4) ?
2/ why is the NA mutation of Parkinson commong among them (5)?
3/ why they carry G-Z3428
4/why they carry R-DF27 (which is 100% Iberian)
5/ partially why more than 81% of Ashkenazim mtDNAs are European (and trace back to Western Europe) (6)

Why didn't you mention that the Lebanese G-Z6028 is closer to the Mende from Sierra Leone than he's to EUROPEAN Jews ? The link between Lebanon and the Sierra Leone might be NW Africa, which means that G-Z6028 is everything but Levantine!

R-M269 :It's indeed present in the Levant (although very rare). Its frequency among Moroccan Jews is substantially higher than in the Levant (18% is everything but low) and they mainly carry SW European such as R-U106 or R-DF27 (which for sure aren't Levantine). Don't forget that the Levant got some European admixture (Romans, crusaders, corsairs)... This European admixture is very likely stronger than it used to be in the Levant prior to the Jewish diaspora

R-DF27 is Iberian. How would you explain that one of its subclade is exclusively found among Sephardim and Ashkenazim (TMRCA : 1150 years!) ? https://www.yfull.com/tree/R-V3476/
Same founder effect as for G-Z6028 !

So, yes, Moroccan Jews are mainly Levantine from their paternal side (something like 55-60%, not more though) but they have a strong West European and NW African contribution!


(1) The genomic history of the Iberian Peninsula reconstructed
(2) The Arrival of Steppe and Iranian Related Ancestry in the Islands of the Western Mediterranean
(3) Eupedia
(4)


(5) https://journals.plos.org/plosone/ar...l.pone.0181335
(6) https://www.nature.com/articles/ncomms3543

I appreciate your reply. Some things I don't understand about haplos: the haplo stated in your yfull link isn't R-DF27. Is it the same haplo in another format? Also, G-Z3428 exists in the Caucus and one of the new Israelite samples released by Reich is Caucus-admixed. As for the Lebanese sample I found, I'm not sure where it maps and if we're talking about the same one.

As for the stuff I do understand, these are the main issues with your reply:

1. You can't compare frequencies of a diaspora population that has likely shared in the Ashkenazi bottleneck to regular populations. This reasoning is what led everyone to think that Ashkenazi R1a is European.

2. You raise a good question of why these haplos are found in Ashkenazim and Sephardim, but not in Mizrachim. The answer could be attributed to the following combination of factors:
- We simply don't have the full data on Mizrachi YDNA. And I had trouble finding much data on it.
- The former is referred to as the Western(Roman) cluster and the latter to the Eastern (Babylonian) cluster, with the exception of Yemenite Jews, whom belong to their own cluster. Studies indicate that these clusters separated a very long time ago, leaving enough time for these haplos to enter the Western cluster in the Levant. And the selection amongst those exiled to Babylon was likely not uniform, for which the haplos may be a proxy for.

3. I should've mentioned this earlier, but there is a potentially massive issue with your sample selection. Granted, it's unavoidable - you can't see the results of those who don't match SpLoa/his grandmother. The issue is that cousins share ancestry. Because cousins share ancestry, their autosomal composition should on average differ from that of a random sample. Since some autosomal data does come from one's YDNA and MtDNA lines, those should then also differ on average from that of a random sample. I understand that you guys did your best to avoid this problem by only counting one individual per surname, however, if you look at Jewish geneology, this is hardly a sufficient method.

4. You, appropriately, asked me many questions. You asked about the Lebanese sample. You asked about the Roman/Babylonian discrepancy. You then asked about the Sephardi-Ashkenazi YDNA sharing, for which I agree - there is some Sephardi ancestry amongst some or maybe even all Ashkenazim. I can't deny this from seeing my own autosomal results. However, Ashkenazi YDNA has been studied in detail and has been concluded to be almost entirely Middle Eastern. Are the Ashkenazi European-YDNA lines simply over represented amongst Sephardim? Here is a site that attempted a similar method to yours, but didn't go through cousin matches: https://jewishdna.net/FAQ.html. They find that " in the Sephardic group we find 6% from Europe and 6% from Iberia. ".

I now want to ask you a question. Why do your conclusions disagree with the following(some of which may not be indepenent): 1. Studies on Morrocan Jews, some of which directly make statements about their YDNA composition. The autosomal esitmates that derive thereof (usually, majority Levantine ancestry). Hell, not just Sephardic, but "Moroccan Jews" are my first match on some calculators. G25 Nmonte calculations: I couldn't get the Levantine below 44%. I used Lebanese Christians as a proxy for the Levant and Mozabite for the Berber. I varied the Spanish and Italian from north to south, all the way to Sicily.

Thanks for your reply.

[mihyar]

Do we have a range of Moroccan Jewish GEDmatch results? This thread has me wishing that GEDmatch population averages reported ranges + standard deviations for each component, because the Moroccan Jewish average gives no hint of this diversity.

[Pine]

Do we have a range of Moroccan Jewish GEDmatch results? This thread has me wishing that GEDmatch population averages reported ranges + standard deviations for each component, because the Moroccan Jewish average gives no hint of this diversity.

Someone on Athrogenica might have that.

[Dick]

why no I1

[XYZ.2018]

Any R1a?

[Pine]

G-Z3428 : You found 1 (only 1) Lebanese BUT there are hundreds of goyim from NW Africa (10 people as far as I know), Iberia, Italy (and Sardinia), Germany, Austria, Switzerland, England and Netherlands who carry it too. It's much more common in SW Europe and NW Africa than anywhere else. (Cf. some FTDNA's projects such as the G-CTS342's one). Indeed, you could find plenty of Ashkenazim, some Sephardi but 0 Mizrahi ! Curious, isn't it ? 0 Egyptian, 0 Arabian (although they take plenty of tests), 0 Iraki, hardly 1 or 2 people from the most Mediterranean part of the Middle East : the Levant (Crusaders?). You would even have very hard time to find it in SW Europe (Greece, Albania...) !
G-Z3428 has been found in SW Sicily (Bronze Age) ! Its mother clade, G-Z1903, has been found 4 times in SW European's Chalcolithic : in the Iberian (1) and Sicilian Chalcolithic (2) (in Lucena, Southern Spain more precisely and in SW Sicily)... Not to talk about its presence in the Bulgarian and Hungarian Neolithic and Chalcolithic (3)... This clade is European and has to be considered as such. It occurs among European Jews only!
Yes, indeed G-Z3428 is mainly Spanish and Italian. It might reflect Iberian and Italian conversions to Judaism !

G-Z6028 (its great-granddaughter's clade ) has had a strong founder effecter among Ashkenazim (TMRCA : 1200 years!) although it's probably Iberian/West Med, yes, but did you know that some Ashkenazi had Sephardi ancestry ? It would explain :
1/ why is E-L19 (especially E-M81) found among them (4) ?
2/ why is the NA mutation of Parkinson commong among them (5)?
3/ why they carry G-Z3428
4/why they carry R-DF27 (which is 100% Iberian)
5/ partially why more than 81% of Ashkenazim mtDNAs are European (and trace back to Western Europe) (6)

Why didn't you mention that the Lebanese G-Z6028 is closer to the Mende from Sierra Leone than he's to EUROPEAN Jews ? The link between Lebanon and the Sierra Leone might be NW Africa, which means that G-Z6028 is everything but Levantine!

R-M269 :It's indeed present in the Levant (although very rare). Its frequency among Moroccan Jews is substantially higher than in the Levant (18% is everything but low) and they mainly carry SW European such as R-U106 or R-DF27 (which for sure aren't Levantine). Don't forget that the Levant got some European admixture (Romans, crusaders, corsairs)... This European admixture is very likely stronger than it used to be in the Levant prior to the Jewish diaspora

R-DF27 is Iberian. How would you explain that one of its subclade is exclusively found among Sephardim and Ashkenazim (TMRCA : 1150 years!) ? https://www.yfull.com/tree/R-V3476/
Same founder effect as for G-Z6028 !

So, yes, Moroccan Jews are mainly Levantine from their paternal side (something like 55-60%, not more though) but they have a strong West European and NW African contribution!


(1) The genomic history of the Iberian Peninsula reconstructed
(2) The Arrival of Steppe and Iranian Related Ancestry in the Islands of the Western Mediterranean
(3) Eupedia
(4)


(5) https://journals.plos.org/plosone/ar...l.pone.0181335
(6) https://www.nature.com/articles/ncomms3543

I thought I'd return to this, now that I understand this material better. In all the clades you mention , but R-DF27, the Jews are closest to Levantine members.

R-M269 is a big and old clade, some of the Jewish subclades of it are clearly Judean; others are clearly European; some might be Mesopotamian in origin. Counting all of it as non-Judean is simply dishonest, as it doesn't even approximate the truth. However, here is an example of an Ashkenazi subclade within it and its closest goy is Lebanese:

https://www.yfull.com/tree/R-L943/

The exact same thing is true for G-Z6028:

https://www.yfull.com/tree/G-Z6028/

And for the main Ashkenazi subclade of E-M81:

https://www.yfull.com/tree/E-A5604/

As for not finding Mizrachim in a given clade - this is typical of almost every single Western Jewish clade, no matter how Levantine it is. The reasons for this are: early separation, lack of Mizrachi samples, and the possibility that some clades left/ remained entirely (in Israel).

However, I agree: most of R-DF27 among Jews is European in origin. The numbers in your study don't align with other statistics. Where is all the Middle Eastern R1b?

[mihyar]

Hey—were there any subclades of G besides G-Z3428?