How much Middle Eastern ancestry do Ashkenazim have?

[heksindhi]

Nice model but it's more complicated than that. North Africa, mesopotamia, and East Asian samples should be included as well as a slavic sample.

Also, the Italian samples are really mixed. Maybe try an Italic IA sample and a mycenean sample

You're right of course. I wasn't attempting an elemental model. My goal was to determine the amount of additional admixture the Ashkenazim picked up as they migrated into Europe or
Levant > Roman Empire > Francia > Slavic Europe


I think the proportions are accurate in that sense, at least for the German Ashkenazi average. As they move further east, and pick up additional European/Slavic, the Mediterranean drops which shows up in the model as an artificial increase in the Levantine. Using the Italic IA sample would fix that.

In any case, as other have already noted, 35%-40% seems about right.

[The Angel's Advocate]

I'd say it ranges from 35% to 45-50%.
Some Gedrosia Near East Neolithic K13 results of Ashkenazis:

#1
Admix Results (sorted):

# Population Percent
1 CHG_EEF 28.87
2 ANATOLIA_NEOLITHIC 20.34
3 NATUFIAN 16.93
4 IRAN_NEOLITHIC 15
5 EHG 7.18
6 SHG_WHG 5.11
7 SIBERIAN 3.65
8 ANCESTRAL_INDIAN 1.9
9 SUB_SAHARAN 0.62
10 PAPUAN 0.39

#2
Admix Results (sorted):

# Population Percent
1 CHG_EEF 32.6
2 ANATOLIA_NEOLITHIC 18.21
3 NATUFIAN 17.18
4 IRAN_NEOLITHIC 15.98
5 SHG_WHG 5.69
6 EHG 5.52
7 SUB_SAHARAN 2.01
8 SIBERIAN 1.84
9 PAPUAN 0.67
10 KARITIANA 0.31

#3
Admix Results (sorted):

# Population Percent
1 CHG_EEF 29.45
2 ANATOLIA_NEOLITHIC 19.97
3 NATUFIAN 19.83
4 IRAN_NEOLITHIC 15.33
5 SHG_WHG 5.81
6 EHG 5.11
7 SUB_SAHARAN 1.97
8 SIBERIAN 1.06
9 ANCESTRAL_INDIAN 0.97
10 KARITIANA 0.32
11 PAPUAN 0.19

Also Shai Carmi says: "Our most recent findings suggest that the Ashkenazi Jewish ancestry is roughly ≈40% Middle Eastern, ≈40% Southern European (or a closely related ancestry), and ≈20% Eastern European. Mixing with Southern Europeans occurred around ≈30-50 generations ago, and was followed by a dramatic decrease in population size around ≈25-30 generations ago, probably representing the early Jewish settlement in Poland. This founder event was followed by an extremely rapid expansion and mixing with (likely) Eastern Europeans."

[alexfritz]

Nice model but it's more complicated than that. North Africa, mesopotamia, and East Asian samples should be included as well as a slavic sample.

Also, the Italian samples are really mixed. Maybe try an Italic IA sample and a mycenean sample

south italians (or aegeans) are actually a good ref. as thats exactly how the romans looked like, italy (cf viminatium / marathon) at the time of the jewish diaspora leaving judaea (?70-136AD)
italic ia was a thing of the past by then

[Levantine_Viking]

south italians (or aegeans) are actually a good ref. as thats exactly how the romans looked like, italy (cf viminatium / marathon) at the time of the jewish diaspora leaving judaea (?70-136AD)
italic ia was a thing of the past by then

When modeling the "mena" admixture of ashkenazim its best to leave out modern Italians.

[Levantine_Viking]

south italians (or aegeans) are actually a good ref. as thats exactly how the romans looked like, italy (cf viminatium / marathon) at the time of the jewish diaspora leaving judaea (?70-136AD)
italic ia was a thing of the past by then

Imperial Rome is way too mixed to use when you're trying to separate the components especially with g25.

[Levantine_Viking]

You're right of course. I wasn't attempting an elemental model. My goal was to determine the amount of additional admixture the Ashkenazim picked up as they migrated into Europe or
Levant > Roman Empire > Francia > Slavic Europe


I think the proportions are accurate in that sense, at least for the German Ashkenazi average. As they move further east, and pick up additional European/Slavic, the Mediterranean drops which shows up in the model as an artificial increase in the Levantine. Using the Italic IA sample would fix that.

In any case, as other have already noted, 35%-40% seems about right.

Campania though isn't a good sample to use because it's modern south Italy and that in of itself is similar to ashkenazim. I think a Greek proxy might be better.

[alexfritz]

When modeling the "mena" admixture of ashkenazim its best to leave out modern Italians.

judging by the model you can use both a mena and italian source
and yes you prob could use other sources for ashkenazim but if there was a roman input after the jewish diaspora ventured out of judaea into the other provinces modern south italians (or aegeans) will fully cover that bill

Imperial Rome is way too mixed to use when you're trying to separate the components especially with g25.

~50% Italic_IA ~50% ancient Near East (alalakh/israel) all on one cline
though the more near east pulled cluster will prob have a higher level of near east but both clusters are closely related nonetheless (based on the papers)

[Aben Aboo]

deleted

[Seabass]

Also Shai Carmi says: "Our most recent findings suggest that the Ashkenazi Jewish ancestry is roughly ≈40% Middle Eastern, ≈40% Southern European (or a closely related ancestry), and ≈20% Eastern European. Mixing with Southern Europeans occurred around ≈30-50 generations ago, and was followed by a dramatic decrease in population size around ≈25-30 generations ago, probably representing the early Jewish settlement in Poland. This founder event was followed by an extremely rapid expansion and mixing with (likely) Eastern Europeans."

Unless he means 20% Non-Southern European (which I think is what he probably should mean...) maybe. But 20% Eastern European I can already tell is unlikely. Eastern Ashkenazi Jews are not at all 20% shifted over to Slavs from Western Ashkenazi Jews. I know this because they are hardly even 15% shifted to non-Southern Europeans from Italian Jews.

[Riverman]

Unless he means 20% Non-Southern European (which I think is what he probably should mean...) maybe. But 20% Eastern European I can already tell is unlikely. Eastern Ashkenazi Jews are not at all 20% shifted over to Slavs from Western Ashkenazi Jews. I know this because they are hardly even 15% shifted to non-Southern Europeans from Italian Jews.

I think 5-10 percent Western-Central European and 10-15 Central-Eastern is the range. But the total admixture from North of the Alps wont exceed 20 %. Its rather about the ratio of Western vs Eastern which is more doubtful.