Ashkenazi Jewish Origins Thread

[billh]

I remember this being a big thread on the old 'genica, so I thought I would try and revive it!

Important studies:

https://www.nature.com/articles/ncomms5835 Shai Carmi et al. (2014)
https://journals.plos.org/plosgenetics/a...en.1006644 (Xue et al. 2017)

Both of these studies found that Ashkenazi Jews are a mix of European and Middle Eastern groups, and that some high or late medieval bottleneck cut the population to mere hundreds.

https://www.biorxiv.org/content/10.1101/...805v1.full (Erfurt Study)
https://www.cell.com/current-biology/pdf...1355-0.pdf (English medieval Jews study)

These studies are the first to include samples of medieval Ashkenazi Jews, finding that they had more variation than modern Ashkenazi Jews. The British Jews were a bit more southerly than modern Ashkenazim, while the Erfurt samples had both more and less southerly samples, split into "Erfurt-EU" and "Erfurt-ME". 

Other studies:

https://www.pnas.org/doi/epdf/10.1073/pnas.2119281119 (I believe they refer to Ashkenazi as "mixed European" here in graphs? Just has some more info on Jewish population bottleneck)
https://www.biorxiv.org/content/10.1101/467761v1.full (Substructure in Ashkenazi Jews -- Western Ashkenazim are less Northern European, have stronger genetic drift)

Haplogroup Studies:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1380291/ (Behar et al.)
-Ashkenazi mtDNA is highly bottlenecked, "four founding lineages" take up 40% of mtDNA pool

However, studies have disagreed on where these haplogroups originate. Some suggest all 4 are from Europe. Others suggest all 4 and a vast majority of Ashkenazi mtDNA is European in origin
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3806353/

Ashkenazi Y-DNA seems to be less controversially Middle Eastern in origin.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3543766/

Although there are oddities in the Ashkenazi haplogroup record. R1a is extremely common among Levites for some reason, although from what I've heard this R1a likely came from the Middle East (perhaps through Indo-Iranic populations). Although, it's odd that it is an Ashkenazi thing in that case. Aaronites 
https://en.wikipedia.org/wiki/Y-chromoso...somal_Levi

There's also evidence of East Asian admixture into Ashkenazi Jews which can't simply be explained by East Asian admixture in Slavs, which allegedly comes from the Silk Road. 
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4323646/

The Khazar Hypothesis has been quite thoroughly debunked, although I vaguely remember hearing about a book dedicated to Ashkenazi mtDNA and Y-DNA which found some Alanic and Khazarian haplogroups in Ashkenazim. If any of you know what came of this, holler at me. 

Anyways, big conversations around Ashkenazi origin include:

-Time and duration of Bottleneck
-Origin of Jewish mtDNA
-Composition of Ashkenazi DNA 
-Route Ashkenazim took through Europe (ex: Rhineland Hypothesis)
-Dating of admixture event 
-Relationship between Ashkenazi Jews and other Jewish groups

Anyways, I guess I'll get this started just as I remember the thread on Anthrogenica regularly bringing up: That a "50/50 model" for Ashkenazim is anachronistic. Most Ashkenazi Jews were not in parts of Italy with a "Northern Italian" profile that people use to generate this model. It's more likely to me that Ashkenazi Jews descend from an Italkim-like population which underwent its mixing during the Roman period and early medieval period, although may have continued this process gradually into the 2nd millennium AD. Of course, the people in urban Italy at this time were more of a central or southern Italian profile. The aforementioned Erfurt study models Ashkenazim as 70% Southern Italian, and 15% Eastern European and West Asian respectively. Of course, ignoring the West Asian component in Southern Italians. This seems to compliment runs I've had on G25 when comparing Southern Italians and Ashkenazim. although G25 seems to be utterly inconsistent with Mediterranean profiles for me. 

Ashkenazim seem to have mixed with Northern Europeans, particularly Eastern Europeans, at a later date. There seems to be some disagreement as to when this happened with the studies. Erfurt suggests it could not have happened after the bottleneck, but earlier studies suggest it must have happened after the bottleneck. I think the most reasonable conclusion is that it happened at the end of the more prolonged bottleneck period suggested in the Erfurt study. This compliments some of the Erfurt samples, who are within this period, having quite substantial Eastern European DNA, while others do not.

But, I would appreciate more input. Especially from people more familiar with Jewish history.

Also, I know that recent world events have made this topic a bit itchy. Please keep all political punditry or motivated arguments over who is indigenous to where out of it.

[Pylsteen]

The book(s) you refer to are from Kevin Brook, I haven't read them, but they seem interesting.

Some of us at the old forum (I don't know where they all went to, but they are welcome here!!) also tried to research both the autosomal the Y-dna lineages further.

We have of course the very valuable, although in some aspects dated database of Wim Penninx (jewish-dna.net). Last year I used that database in combination with the FamilyTree DNA Discover tree to look at the age and growth rate of Ashkenazi-specific Y-DNA lineages (I remember some of us having done that too, there are about 90 to 100 Ashkenazi-specific lineages, some clearly ancient Jewish, some "Mediterranean", some Western or Central-European, and a few recent Sephardic and Mizrahic). I'm planning to write this all down in a personal paper, although it will likely take some more time.

In short, the thoughts I have myself on Ashkenazi origins and development at the moment based on all the input, may be summarized as follows:

1. Fundamental to Ashkenazim, Sephardim and Italkim, is a core mixture of Middle Eastern and Greek/Southern-Italian-like ancestry. This mixture must be dated to large period (later Iron Age, Hellenistic period, Roman period).
   
2. Additionally, G25 runs seem to show about 5-10% Northern African (Berber-like) admixture, although not confirmed by peer-reviewed studies AFAIK. If legitimate, its source is likely the Hellenistic/Roman jews of places such as Alexandria and Cyrenaica.
   
3. This mixture created the "proto-Western jewish" population, largely living in Roman Italy, ancestral to Italkim, Sephardim and Ashkenazim.
   
4. Studying the TMRCA's of Ashkenazi-specific Y-DNA lineages in the FTDNA discover tool, there is a remarkable peak in TMRCA's between the late 7th century and late 9th century (taking the average estimates). After this period, influx of new lineage remains, but in a lower pace. This initial peak is IMO the start of a specific Ashkenazi population, likely assembled from Italian, Frankish and Visigoth jews. In the old forum, I think it was Riverman and another one who came up with this "Frankish period origin" idea. I also looked at the Ashkenazi-specific lineages that were of older jewish origin (shared with other jewish groups) and for now conclude that they were on average still mostly "common jewish" during the 6th century. So after 600 but before 900 seems a safe bet for the time of origin of the "core" of the Ashkenazi population.
   
5. This founder effect in Y-DNA lineages during the 7th to 9th century probably coincides with the bottleneck in autosomal genes, resulting in a population IMO similar to what was found in the Chapelfield samples (12th century) and Erfurt-ME samples (probably Tzarfat/Rhineland-like). After the bottleneck, a few Western European lineages seem to have entered the Y-DNA gene pool. Immigration from jews from Italy to the Rhineland (such as the Kalonymus family) remained.
   
6. In the same time, a (small) number of these post-bottleneck early Ashkenazim must have migrated to central/eastern Europe, likely taking part in the trade between Germany, Poland, Prague, Hungary, Kievan Rus and the Khazar empire. These Ashkenazim (probably together with other Jewish traders) were IMO likely ancestral to the Kna'anim and the Erfurt-EU cluster. They have picked up a large amount (up to 30%) of Slavic ancestry and some "Asian" through the trade routes along central Asia. The amount of Slavic Y-DNA lineages is however very low. I assume that during a later period Rhineland-like jews started to emigrate to central/eastern Europe as well.
   
7. Besides the Rhineland and Kna'anim groups there was likely more substructure we dont know of yet. In any case, both groups contributed to the modern Ashkenazim in different proportions (a cline is still seen from German jews to Russian jews). Of course there are many complex details due to the many expulsions. Additionally, Sephardim, perhaps Italkim and jews from the Levant/Mesopotamia also came to Central/Eastern Europe.
   
8. Looking at the growth patterns of the Ashkenazi-specific Y-DNA lineages with the FTDNA discover tool, I notice that the internal growth took over new influx from the 9th century onwards, and was especially large between the 1300 and 1600. This probably also reflects the developments in the population growth itself.
   

Well, that went to be a longer post than I thought, I hope I can write down more on the Y-DNA lineages during the next months.

[Riverman]

What I wonder about is how common actually rare yDNA-subclades among Ashkenazi Jews are, which have a deeper Jewish history, to put it that way. Like going back to the Medieval period and earlier.
Because most Ashkenazi lineages are both widespread and well-tested, due to Medieval and later founder events, as well as Ashkenazi Jews being more willing and able to test than many other populations from around the world. But are there actually proven, historically Jewish lineages which are more rare? Its hard to prove they are Jewish without ancient DNA and when there is no more than 1 Jewish tester.
I tried to find such rare lineages with say less than 3 modern testers which are clearly Jewish, but with no luck so far. All proven Jewish subclades have, if consulting FTDNA and YFull, got more testers and experienced a founder event. That's why I'm asking whether there are any known rare ones at all.

The criteria would be:
- Less than 4, but more than 1 tester with FTDNA and YFull being combined
- A TMRCA pointing to an old branching event from within the Jewish population, not a recent (after 1700) proselyte, NPE or other such scenarios.

[Chad]

Does anyone have the genotype data for the English Jews?

[RBHeadge]

What I wonder about is how common actually rare yDNA-subclades among Ashkenazi Jews are, which have a deeper Jewish history, to put it that way. Like going back to the Medieval period and earlier.
Because most Ashkenazi lineages are both widespread and well-tested, due to Medieval and later founder events, as well as Ashkenazi Jews being more willing and able to test than many other populations from around the world. But are there actually proven, historically Jewish lineages which are more rare? Its hard to prove they are Jewish without ancient DNA and when there is no more than 1 Jewish tester.

jewishdna.net still periodically adds new branches to their list. Similarly the numbers of kits in pre-existing branches also increase. They get their data view the Avotaynu Project and other academic endeavors. I can't easily find to the page that explains how they collate and verify the veracity of the submitted kits, but IIRC it also involved atDNA and was pretty strict. Since they keep adding YDNA branches, its stands to reason that they are still finding lesser populated survivor branches.

[Riverman]

What I wonder about is how common actually rare yDNA-subclades among Ashkenazi Jews are, which have a deeper Jewish history, to put it that way. Like going back to the Medieval period and earlier.
Because most Ashkenazi lineages are both widespread and well-tested, due to Medieval and later founder events, as well as Ashkenazi Jews being more willing and able to test than many other populations from around the world. But are there actually proven, historically Jewish lineages which are more rare? Its hard to prove they are Jewish without ancient DNA and when there is no more than 1 Jewish tester.

jewishdna.net still periodically adds new branches to their list. Similarly the numbers of kits in pre-existing branches also increase. They get their data view the Avotaynu Project and other academic endeavors. I can't easily find to the page that explains how they collate and verify the veracity of the submitted kits, but IIRC it also involved atDNA and was pretty strict. Since they keep adding YDNA branches, its stands to reason that they are still finding lesser populated survivor branches.

Thank you. But if it's an extremely small number or even more important, has a very recent TMRCA, it likely is a later convert, NPE etc., after the formation of the Ashkenazi population. They include such lineages too.
Without enough samples its sometimes difficult to tell which is which, since for a perfect dating it needs both Ashkenazi Jewish and non-Ashkenazi Jewish testers to estimate when it entered the Ashkenazi gene pool.

[Riverman]

I just looked up the Rothschild branch according to FTDNA and for that branch emerges an interesting pattern:

https://discover.familytreedna.com/y-dna/J-Y23457/tree

A first founder event about 1200 AD, then a split into a German, central and Polish branch. The central group looks like by far the biggest, demographically most successful, with people from around Poland-Russia moving back to Western Europe secondarily. The upstream founder of J-Y15234 lived in the 9th century, therefore again in the early Frankish era, like practically all major Ashkenazi founder lineages.

I think that's a pattern which repeats itself, with the original foundation between 600-900 and then splits after 1000-1300 AD. Many of the Jewish regional branches with a clearly Ashkenazi profile had very little contacts after 1300 AD, which means the relationships between these communities need to date before 1200-1400 in my opinion.

[Pylsteen]

One of the "Oppenheimer" lineages claiming descent from the the Worms/Frankfurt Oppenheimers (15th century) belongs to Q-BZ37, part of the large Q-Y2232 lineage, which is one of the oldest and largest Ashkenazi lineages (except for downstream Sephardic Q-BZ72). In fact, it is so old (3rd century AD) that it likely predates the Ashkenazi founder effect. 

Back to the Oppenheimers: the discovery tree doesn't show many downstream German lineages. But, if we take another clear Frankfurt-Jewish lineage, R-BY13143 (the Epstein branch), the number of Central/East European downstream lineages also exceeds the number of German ones. Can we say that the branches that remained in Germany are undertested?

[Riverman]

One of the "Oppenheimer" lineages claiming descent from the the Worms/Frankfurt Oppenheimers (15th century) belongs to Q-BZ37, part of the large Q-Y2232 lineage, which is one of the oldest and largest Ashkenazi lineages (except for downstream Sephardic Q-BZ72). In fact, it is so old (3rd century AD) that it likely predates the Ashkenazi founder effect. 

Back to the Oppenheimers: the discovery tree doesn't show many downstream German lineages. But, if we take another clear Frankfurt-Jewish lineage, R-BY13143 (the Epstein branch), the number of Central/East European downstream lineages also exceeds the number of German ones. Can we say that the branches that remained in Germany are undertested?

Testing in Germany is generally worse than in other regions, but on the other hand, most emigrated and moved to countries where testing is more common, plus they belong on average to the socially better of people even relative to the Jewish average.
So my best guess is that they simply didn't reproduce as successfully, on average, as those which remained in a more traditional context in the East.

Its basically a similar pattern, even if not as extreme, as if you compare German lineages which moved out with Mennonites and Amish vs. regular bourgeois Germans. Its pretty obvious which side of the family with have been biologically more successful and more widespread in most if not all instances.

My hunch is that the same applies to some German vs. Eastern European Jewish lineages. The latter had way more offspring on average.

[Piquerobi]

An interesting Family Tree DNA group with yDNA results of Jewish families from Frankfurt. Frankfurt is located in the area where the Ashkenazi group is supposed to have sprung.

https://www.familytreedna.com/groups/jew...furt/about

[RBHeadge]

I just looked up the Rothschild branch according to FTDNA and for that branch emerges an interesting pattern:

https://discover.familytreedna.com/y-dna/J-Y23457/tree

A first founder event about 1200 AD, then a split into a German, central and Polish branch. The central group looks like by far the biggest, demographically most successful, with people from around Poland-Russia moving back to Western Europe secondarily.

I manage a STR kit under Y23457>Z36098 and I've spent a bunch of time researching it. The Y23457 and Y23457>Z36098 branches differs from other AJ branches in that it mostly stayed along the Rhine in Germany (or returned so quickly that it didn't matter for our purposes). There were probably two separate migrations east to Poland-Lithuania. One in the 1400's and another in the 1600-1700s during all of the various wars (30 Years, Spanish Succession, etc.)  and the subsequent Jewish expulsions of the period. I doubt there was much, if any, back-migrated from Poland-Lithuania. I don't consider migration up/down the Rhine to matter in this context.

The upstream founder of J-Y15234 lived in the 9th century, therefore again in the early Frankish era, like practically all major Ashkenazi founder lineages.

I think that's a pattern which repeats itself, with the original foundation between 600-900 and then splits after 1000-1300 AD. Many of the Jewish regional branches with a clearly Ashkenazi profile had very little contacts after 1300 AD, which means the relationships between these communities need to date before 1200-1400 in my opinion.

Yes, I agree that it has all the typical AJ branching events from c600 CE onward.

[RBHeadge]

Can we say that the branches that remained in Germany are undertested?

Testing in Germany is generally worse than in other regions

Almost none of those families are still living in Germany. The kits we see now are from those who immigrated prior to 1940 and their descendants.

So my best guess is that they simply didn't reproduce as successfully, on average, as those which remained in a more traditional context in the East.

Its basically a similar pattern, even if not as extreme, as if you compare German lineages which moved out with Mennonites and Amish vs. regular bourgeois Germans. Its pretty obvious which side of the family with have been biologically more successful and more widespread in most if not all instances.

My hunch is that the same applies to some German vs. Eastern European Jewish lineages. The latter had way more offspring on average.

I manage a lot of YDNA and atDNA kits and they're either Pennsylvania Dutch kits or AJ and their family trees. Both groups had more offspring and surviving offspring than those that remained in Germany.

[Tomenable]

Has anyone saved this paper about Jews in the HRE and in the PLC (Poland) before it was deleted?:

https://faculty.runi.ac.il/faculty/eckst...pt2018.pdf

"Server Error
404 - File or directory not found.
The resource you are looking for might have been removed, had its name changed, or is temporarily unavailable."

I can't find it anywhere (so probably I didn't download it) and it wasn't saved by Web Archive.

[Freb]

I say we’ve gotten real lucky with the Erfurt and Norwich results I don’t think we will ever get something like this in a very long time.. or perhaps never.

[szin]

Kevin Brook demonstrates that SOME Ashkenazim score small amounts of East Asian on DNA tests, POSSIBLY indicating roots in the Khazarian Empire but not necessarily. However, this amount of admix is minute, suggesting that the Khazarian impact on the Ashkenazi genepool is minimal- IF that's even the source.

Interestingly, Brook also demonstrates that some Ashkenazim score small amounts of SSA, and that some may have SSA-derived mtDNA haplogroups. SSA admix has long been conjectured based on historical and Biblical evidence (you may remember that Moses married a "Kushite"), but only now are we getting evidence that some Ashkenazim have it.