Time and place of European admixture into the Ashkenazim-Xue et al

[Angela]

Here we go again. :) Can't we just wait for the ancient dna which must surely be on its way?

Oh well, from Xue et al, (including Shai Carmi)

See the following pre-print:
http://biorxiv.org/content/biorxiv/e...63099.full.pdf

"The Ashkenazi Jewish (AJ) population is important in medical genetics due to its high rate of Mendeliandisorders and other unique genetic characteristics. Ashkenazi Jews have appeared in Europe in the 10thcentury, and their ancestry is thought to involve an admixture of European (EU) and Middle-Eastern(ME) groups. However, both the time and place of admixture in Europe are obscure and subject tointense debate. Here, we attempt to characterize the Ashkenazi admixture history using a largeAshkenazi sample and careful application of new and existing methods. Our main approach is based onlocal ancestry inference, assigning each Ashkenazi genomic segment as EU or ME, and comparing allelefrequencies across EU segments to those of different EU populations. The contribution of each EUsource was also evaluated using GLOBETROTTER and analysis of IBD sharing. The time of admixture wasinferred using multiple tools, relying on statistics such as the distributions of EU segment lengths andthe total EU ancestry per chromosome and the correlation of ancestries along the chromosome. Oursimulations demonstrated that distinguishing EU vs ME ancestry is subject to considerable noise at thesingle segment level, but nevertheless, conclusions could be drawn based on chromosome-widestatistics. The predominant source of EU ancestry in AJ was found to be Southern European (≈60-80%),with the rest being likely Eastern European. The inferred admixture time was ≈35 generations ago, butmultiple lines of evidence suggests that it represents an average over two or more admixture events,pre- and post-dating the founder event experienced by AJ in late medieval times, with the prebottleneckadmixture event bounded between 25-55 generations ago."

So, around 800 AD? Is the speculation still that they were in Italy at that time? Does that mean they're not "Middle Eastern" at all? That doesn't seem right based on their yDna.

Of course, if it's closer to 50 generations ago we'd be talking about around 200 AD? Southern European converts to Judaism?

Well, I guess I'll be reading it today while I'm waiting in reception rooms.

[davef]

I swear these studies kill my productivity!

Thanks for posting this but what makes this study special? We already know that doing admixture can make european jewish populations a certain percentage of Eastern European, Southern European and Mideast. That's what fire haired came up with using tools he probably downloaded for free. We don't know which populations are responsible for these scores.
Nothing on you for posting this, I'm just curious as to why they spent so much money to reach conclusions that may as well be reached (in my possibly clueless mind) using a cheap laptop and a few files and apps :).

[davef]

I feel I've been overly harsh (that's how bad work days affect people). This is on fact one of the most interesting articles on this subject I've ever read because it estimates the pre founder admixture event as early as 8 hundred something AD. Assigning a time makes a huge difference and it could've happened while in what italy was back then. Where else could they have gotten the south european score? It certainly couldntve been from Central Europe (painfully obvious). There's the lack of ibd with italians as reported but that's with current day italians, Italy is very very diverse and very large so maby those italians jews do have ibd sharing haven't been found or tested yet (assuming they exist to begin with).

Or maybe they came from an old population that shared a lot with south euros. If the pre founding admixture event didn't happen in italy it would be odd if it happened in Germany assuming jews werent european shifted and purely Levantine.

[Angela]

Some tidbits from the paper:

"Running RFMix on the AJ genomes with our European and Middle-Eastern reference panels and summing up the lengths of all tracts assigned to each ancestry, the genome-wide ancestry was ≈53% EU and ≈47% ME, consistent with an ADMIXTURE analysis (Methods) and our previous estimate based on a smaller sequencing panel."

The problems came when they did the IBD analysis, f4 statistics and even tree mix, and used Alder. Looking at the IBD sharing made it seem as if almost all the admixture was from eastern Europe. Also, there was more IBD sharing with Iberians, Western Europeans, and much more with Eastern Europeans than with Southern Europeans. If you go to the paper and scroll down to the IBD section, they have a graphic of it which is quite startling.

Alder and Tree Mix also produced results consistent with most of the European ancestry stemming from eastern Europeans.

As for f4 statistics, here is what they had to say:

"We measured the f4 statistic for the configuration (X,YRI;AJ,ME), where we used Yoruba (YRI) as an outgroup, and substituted different European regions for X (Figure S4A). The European region that gave the highest value of f4, Eastern Europe (closely matched by Western Europe), is theoretically the one closest to the true source of European gene flow. However, simulations with a dominant (or even exclusive) Southern European source resulted in highest f4 values for Eastern Europe as well. [This discrepancy might be explained, at least partly, by a strong Middle-East to Southern EU migration event [51] (Figure S5)]. Therefore, those results are still consistent with a dominant Southern EU source for AJ."

They then re-analyzed the data with computer simulations, and their conclusion was that "The results of all analyses (once calibrated by simulations) pointed to Southern Europe as the predominant source of European gene flow. At the same time, minor contributions from Western and/or Eastern Europe were also detected, with some analyses (IBD within AJ and between AJ and other sources, and Globetrotter) showing stronger support for an Eastern European source. Based on historical plausibility, these admixture events must have necessarily happened at different times, implying multiple historical events. The inferred admixture time, when modeled by a single event, was between ≈24-40 generations ago by the methods we examined (calibrated mean segment length and ancestry proportions, Alder, and Globetrotter), very close to the time of the AJ bottleneck, previously estimated to ≈25-35 generations ago [9, 16]. Therefore, admixture must have occurred both before and after the bottleneck, with the IBD and Alder analyses suggesting that the Eastern European admixture was more recent."

In terms of Globetrotter results, this is what they had to say.

" the overall European ancestry, ≈70% (or ≈67% after calibration by simulations; Supplementary Text S3), is about 15% higher than the LAI-based estimate, as well as our previous results based on whole-genome sequencing [9]. Our detailed calibration simulations (Supplementary Text S3) demonstrate that evidence exists to support either estimate, suggesting that the true fraction of EU ancestry is midway, around ≈60% (see Discussion)."

As to the programs which they attempted to use:

"In summary, we demonstrated that the raw results returned by Alder, the f-statistics, and TreeMix must only be interpreted in light of simulations. Using such simulations, the results were overall consistent with our previous model of an admixture event ≈35 generations ago involving predominantly Southern Europe, with minor contributions of either Western or Eastern Europe.

When simulating Middle-Eastern and European admixture (particularly Southern European), we found many tools to be of limited utility (see, e.g., the section on Alder, f-statistics, and TreeMix, the Methods section on ADMIXTURE, and Supplementary Texts S1 and S2 on LAMP and PCAMask). Further, while we were eventually able to extract useful information off RFMix’s local ancestries, the raw results were not very accurate: the accuracy per SNP was only ≈70- 80%, the mean segment length was more than twice than expected, and the variance of the ancestry proportion per chromosome was underestimated. When jointly analyzing LAI and IBD sharing, the inferred proportion of IBD segments that were either false or had random LAI was as high as ≈35% ((1-λ) in Methods), although fortunately, we were able to account for that in the model."

I've certainly been no fan of Alder (i.e. used in the Hellenthal/Busby papers on recent admixture) but I'll leave it to any "modelers" who read the paper to opine on whether they have an important point here. If they do, then even ancient dna might not give us reliable answers.

To get back to simple basics, if they don't find any appreciable IBD sharing between Ashkenazim and any southern Europeans as defined as Bergamo, Tuscany, Abruzzo, Sicilians or Greeks, I fail to see how they could be responsible for the European admixture into the Ashkenazim. The only southern Europeans who do seem to have some substantial IBD sharing with the Ashkenazim are the Iberians. Is it possible it was Iberian Jews who moved to the Rhineland and formed the nucleus for the Ashkenazi Jews? I don't know why they didn't attempt to model them that way.

Taking the opposite view that the programs they ran without simulations were indeed accurate, and the admixture was mostly eastern European, one wonders what happened to all the eastern European yDna. If it is about 8% one could see that perhaps it just drifted out except for the I1 sub-group. I'm also unclear if it's definite that they picked up their Z93 subclade in the Near East. If it's about 50%, however, then it makes no sense.

Unless, of course, one were to dredge up that old theory that it was Jewish men who went to eastern Europe, and arrived there both without all that many women of their own, and at a time when there wasn't so much hostility and the church didn't yet have all that much control, and so they were able to take a lot of non-Jewish wives.

Well, all I can say is I wouldn't bet any money on any of this. I'll wait for the ancient dna, with the caveat that I want to know if some of these programs are indeed reliable.

Ed. It's late and it's been a long day, so if I made any misrepresentations as to their conclusions, just let me know.

[davef]

Single most important paragraph:
"The results of all analyses (once calibrated by simulations) pointed to Southern Europe as the predominant source of European gene flow. At the same time, minor contributions from Western and/or Eastern Europe were also detected, with some analyses (IBD within AJ and between AJ and other sources, and Globetrotter) showing stronger support for an Eastern European source. Based on historical plausibility, these admixture events must have necessarily happened at different times, implying multiple historical events. The inferred admixture time, when modeled by a single event, was between ≈24-40 generations ago by the methods we examined (calibrated mean segment length and ancestry proportions, Alder, and Globetrotter), very close to the time of the AJ bottleneck, previously estimated to ≈25-35 generations ago [9, 16]. Therefore, admixture must have occurred both before and after the bottleneck, with the IBD and Alder analyses suggesting that the Eastern European admixture was more recent."

Admixture occurred both before and after the bottleneck. We know that after the bottleneck, it was from a Slavic population. But the mystery is whom it was with BEFORE the bottleneck. The southern euro score is the biggest mystery here and needs serious investigation. Its way higher than the Slavic. The first sentence says south Europe is the predominant source but we don't have the ibd data to figure out which Southern European population was the source though as you said, could be spain/portugal.
The Southern European score is the biggest trickster here.

Yeah I need that sleep thing as well.

[davef]

Also, why did they mix in more northerly italians with abruzzo, Sicily and greek? I know abruzzo is more in the center but I question how many people were used in each group because if they used say 90 percent from bergarmo that would produce way different results than if they used 90 percent Sicily.

Sorry, but it never fails to blow my mind that on those gedmatch calculators european jews fit so well as italians (this is every calculator), levantines don't come close (some would be about maybe 5 to 10 percent south italian based on those oracles and often nothing at all) and even the Bronze Age levantines probably don't cut it (I haven't checked their results). I know there's no evidence of mixing with italians based on the weak ibd scores but the south european score couldn't have come from nowhere.

BTW I'm not vouching it came from Italy. And I apologize if I sound like a broken record but my mother has a broken record tendency as well

[Angela]

I get the sense that a lot of effort was expended to get the results from this new method to match the results from their prior paper, and also to match a long held belief by a number of researchers that the Ashkenazi population was formed by the admixture of Jewish men and the Italian women (probably northern Italian women) whom they married and converted.

I'm not sure they're correct that the admixture was between a Middle Eastern population and a southern European population if by southern European they mean Italians and Greeks (Bergamo, Tuscany, Abruzzi, Sicily and Greece are the listed populations that were used). I don't see how it's possible for 50-60 % of the genome of the modern Ashkenazim to be from these people less than 2000 years ago and yet there is almost no IBD sharing with any of them. If the Ashkenazim stem from a southern European population then perhaps it is the Iberians, as there's much more sharing between Ashkenazi Jews and Iberians. (Unless the sharing with the Iberians is mostly gene flow from Jews into Iberian populations during the Moorish occupation?) Strangely, they never modeled the admixture with Iberians.

It's either that or Jewish men arriving in eastern Europe with very few wives and marrying a lot of gentile women before endogamy was enforced.

I also don't understand why the Ashkenazim share more IBD with Eastern Europeans not only in the last 1000 years or so, but also, presumably, 2000 years ago, and why, in the earlier time period they don't share all that much with Levantine populations?

I'd very much like for someone else to redo the IBD analysis on these specific samples, using both their methods and others.

IBD analysis of Ashkenazi Jews-Xue et al.PNG



I have no idea what you're getting at with your first paragraph. Why wouldn't they include northern Italians? They tried to get a combined reference sample that would do a good job of covering all or at least most of Italian genetic variation so that they'd have a better chance of finding some IBD sharing. Even with samples from northern Italy, Tuscany, central Italy and the south, they couldn't find it, but they found it with Iberians even though the total number of samples was less.

As to the admixture utilities posted at gedmatch, and the Oracle results often provided, I'm afraid you seem to be a bit confused about how they should be interpreted. The fact that two groups of people share similar percentages of ancestry from very ancient populations doesn't at all mean that they necessarily share any gene flow in thousands of years. They might, but not necessarily. Nor are possible Oracle "admixtures" meant to be taken literally. This is as mistaken as to assume that because two groups of people cluster near one another on a PCA it means that they share any genes at all in even the last 10,000 years.

You might want to take a look at this article about how not to use Admixture. Even academic authors don't seem to be aware of some of the pitfalls. I posted it here as I think it's very helpful in interpreting or not misinterpreting Admixture results.
http://www.eupedia.com/forum/threads...ted-Cautiously

The truth of the matter, of which many people seem to be totally unaware, is that with any of these analyses, including the ones using dstats like those done by Fire-Haired, if you plug in enough different populations you can get a number of patterns that are a "reasonable" fit, so you can prove just about anything you wish. That's why people have to be wary of the conclusions of even academic papers until they've really thought it through or tried to reproduce the results themselves. I'm waiting for some of the modelers in chief to do that with this paper. When hobbyists are involved, it's buyer beware, in my opinion.

One specific reason that I'm not spending too much time on this is all the calculations are, of course, based on modern genomes. The computations are based, for example, on modern Middle Easterners like Jordanians, Palestinians, Druze etc. However, how similar are these modern groups to the Jews of the first centuries of the Common Era, the period roughly corresponding to the time of the Roman Empire? From the current Lazardis et al paper, the ancient Levantine Bronze Age sample plots among modern Saudis and Egyptians. If the Jews of that era were close to them, then the Ashkenazim are perhaps even more "European" than the authors of this paper propose. What if, however, there has been additional SSA and other genetic intrusion into the Levant since the Jews were scattered? Then, the Ashkenazim may be less "European" admixed.

So, for me, I'm not looking for ultimate answers from this paper. More important to me is whether their criticism of some of these programs for use on southern Europeans, and Jews, is justified, because the tools will be used on ancient dna as well.

[davef]

I get the sense that a lot of effort was expended to get the results from this new method to match the results from their prior paper, and also to match a long held belief by a number of researchers that the Ashkenazi population was formed by the admixture of Jewish men and the Italian women (probably northern Italian women) whom they married and converted.

I'm not sure they're correct that the admixture was between a Middle Eastern population and a southern European population if by southern European they mean Italians and Greeks (Bergamo, Tuscany, Abruzzi, Sicily and Greece are the listed populations that were used). I don't see how it's possible for 50-60 % of the genome of the modern Ashkenazim to be from these people less than 2000 years ago and yet there is almost no IBD sharing with any of them. If the Ashkenazim stem from a southern European population then perhaps it is the Iberians, as there's much more sharing between Ashkenazi Jews and Iberians. (Unless the sharing with the Iberians is mostly gene flow from Jews into Iberian populations during the Moorish occupation?) Strangely, they never modeled the admixture with Iberians.

It's either that or Jewish men arriving in eastern Europe with very few wives and marrying a lot of gentile women before endogamy was enforced.

I also don't understand why the Ashkenazim share more IBD with Eastern Europeans not only in the last 1000 years or so, but also, presumably, 2000 years ago, and why, in the earlier time period they don't share all that much with Levantine populations?

I'd very much like for someone else to redo the IBD analysis on these specific samples, using both their methods and others.

IBD analysis of Ashkenazi Jews-Xue et al.PNG



I have no idea what you're getting at with your first paragraph. Why wouldn't they include northern Italians? They tried to get a combined reference sample that would do a good job of covering all or at least most of Italian genetic variation so that they'd have a better chance of finding some IBD sharing. Even with samples from northern Italy, Tuscany, central Italy and the south, they couldn't find it, but they found it with Iberians even though the total number of samples was less.

As to the admixture utilities posted at gedmatch, and the Oracle results often provided, I'm afraid you seem to be a bit confused about how they should be interpreted. The fact that two groups of people share similar percentages of ancestry from very ancient populations doesn't at all mean that they necessarily share any gene flow in thousands of years. They might, but not necessarily. Nor are possible Oracle "admixtures" meant to be taken literally. This is as mistaken as to assume that because two groups of people cluster near one another on a PCA it means that they share any genes at all in even the last 10,000 years.

You might want to take a look at this article about how not to use Admixture. Even academic authors don't seem to be aware of some of the pitfalls. I posted it here as I think it's very helpful in interpreting or not misinterpreting Admixture results.
http://www.eupedia.com/forum/threads...ted-Cautiously

The truth of the matter, of which many people seem to be totally unaware, is that with any of these analyses, including the ones using dstats like those done by Fire-Haired, if you plug in enough different populations you can get a number of patterns that are a "reasonable" fit, so you can prove just about anything you wish. That's why people have to be wary of the conclusions of even academic papers until they've really thought it through or tried to reproduce the results themselves. I'm waiting for some of the modelers in chief to do that with this paper. When hobbyists are involved, it's buyer beware, in my opinion.

One specific reason that I'm not spending too much time on this is all the calculations are, of course, based on modern genomes. The computations are based, for example, on modern Middle Easterners like Jordanians, Palestinians, Druze etc. However, how similar are these modern groups to the Jews of the first centuries of the Common Era, the period roughly corresponding to the time of the Roman Empire? From the current Lazardis et al paper, the ancient Levantine Bronze Age sample plots among modern Saudis and Egyptians. If the Jews of that era were close to them, then the Ashkenazim are perhaps even more "European" than the authors of this paper propose. What if, however, there has been additional SSA and other genetic intrusion into the Levant since the Jews were scattered? Then, the Ashkenazim may be less "European" admixed.

So, for me, I'm not looking for ultimate answers from this paper. More important to me is whether their criticism of some of these programs for use on southern Europeans, and Jews, is justified, because the tools will be used on ancient dna as well.

You said if you plug in enough different populations you can get enough different patterns that are a "reasonable" fit...not quite understanding what you mean.

And I don't quite understand the link you posted. It's not getting through to my skull. I'm good at admitting I'm too "dumb" or inexperienced to understand anything written with regards to this subject. It's gobbledygook to me lol.

The while ibd with Eastern Europeans within the last 2000 years doesn't seem right at all. Weren't jews in Rome at that time? That's ridiculous. Now the Southern European source is Iberian...it keeps changing every year. Sorry to sound this way but with the sub par week I've been cursed with and this, I'm feeling frustrated. Nothing on you, but wait till next year and they'll conclude that Ashkenazim are from a sacred order of lizard people.

Pulling hair out in three...two...one.,,

[davef]

These posts are merely frustrations of someone who descends from endogamous groups and thus I don't know how I came to be throughout these thousands of years. All of my ancestors are from "backwoods" villages lol. By backwoods, think the equivalent of Appalachia in the US. Endogamous to the extreme.
Dont know the better word for backwoods in case someone doesn't approve of that word. I apologize.

[davef]

But for one thing I do know better than to take oracle results literally, so if someone from france posts an oracle result of 50% English 50% Spanish I'm not going to assume that this french individual has one spanish parent and one english parent and neither should he/she. I think once Alan explained why a neolithic Levantine individual scored 30 percent sardinian I think i started getting what these things are telling us, that is 30 percent of a Sardinians genome is from these neolithic Levantine individuals. Obviously it's not the case that the neolithic Levantine had one parent who was 100 percent Yemenite jewish and one who was 40 percent Yemenite jewish 60 percent sardinian. Common sense really, were there Yemenite Jews (or Jews at all) and Sardinians in the Stone Age? Of course not.
EDIT: I might add that I'm absolutely disgusted with people who manipulate the data in order to fulfill their prophecies because if Joe Bob uses a calculator that is way off center against his 23 and me results and winds up getting a bit of a "surprise" (such as getting 3 percent Native American when his ancestors were colonials from England and Germany for centuries and millennia) he's going to be convinced that he has Native ancestry when it could just be a preshistoric connection. He'll be digging thru papers for years trying to find that Indian in the tree only to realize he's been wasting time. Note there are countless colonial Americans who wish to join a tribe.
In my view, if you convince someone that he isn't of the ethnic background (s) he always believed he was a part of, that could change how he sees himself. Do you get my drift?

[LeBrok]

These posts are merely frustrations of someone who descends from endogamous groups and thus I don't know how I came to be throughout these thousands of years. All of my ancestors are from "backwoods" villages lol. By backwoods, think the equivalent of Appalachia in the US. Endogamous to the extreme.
Dont know the better word for backwoods in case someone doesn't approve of that word. I apologize.

We all came from hunter gatherers groups and later farmers, mostly secluded in their small villages. All pretty much endogamous for long periods, except short times of mixing with new blood.
Watch any Amazon tribes video on Youtube, the ones still separated from civilization in strict reserves. They've been endogamous for thousands of years. All people in the tribe looks like bunch of brother and sisters, almost twins. And yet they existed like this for thousands of years. Being endogamous can't be that bad as people say?

[davef]

Oh no, I'm not ashamed of being from endogamous groups or saying its a bad thing...it's just hard to figure out the origins of endogamous groups.

[Tomenable]

one wonders what happened to all the eastern European yDna.

We do have few Y-DNA samples from the Khazar Khaganate (the Saltovo-Mayaki archaeological culture).

Saltovo-Mayaki archaeological culture was located in South-Eastern Europe (Pontic-Caspian region):

https://www.academia.edu/15713987/Аф...15._С.146-153

Individual - haplogroups (all samples dated to years ca. 800 - 900 AD):

A80301 - R1a1a1b2a (Y-DNA), I4a (mtDNA)
A80302 - D4m2 (mtDNA)
A80410 - G (Y-DNA)
A80411 - J2a (Y-DNA)

The population of the Khazar Khaganate / of the Saltovo-Mayaki culture, was ethnically heterogeneous:

https://en.wikipedia.org/wiki/Saltovo-Mayaki

"Saltovo-Mayaki is the name given by archaeologists to the early medieval culture of the Pontic steppe region roughly between the Don and the Dnieper Rivers. Their culture was a melting pot of Onogur, Khazar, Pecheneg, Magyar, Alan, and Slavic influences. During the ninth century the Saltovo-Mayaki culture was closely associated with the Khazar Khaganate, and archaeological sites from this period are one way that historians track the geographic scope of Khazar influence."

====================

The most common Jewish subclade of R1a, is this: R1a1a1b2a2b1a1 M582 (aka CTS6).

It descends from R1a1a1b2a Z94 - one sample of which was found in Saltovo-Mayaki.

[Angela]

We do have few Y-DNA samples from the Khazar Khaganate (the Saltovo-Mayaki archaeological culture).

Saltovo-Mayaki archaeological culture was located in South-Eastern Europe (Pontic-Caspian region):

https://www.academia.edu/15713987/Афанасьев_Г.Е._Вень_Ш._Тун_С._ Ван_Л._Вэй_Л._Добровольская_� �.В._Коробов_Д.С._Решетова_И.К ._Ли_Х.._Хазарские_конфедера� �ы_в_бассейне_Дона_Естестве� �нонаучные_методы_исследов� �ния_и_парадигма_современно й_археологии._М._2015._С.146-153

Individual - haplogroups (all samples dated to years ca. 800 - 900 AD):

A80301 - R1a1a1b2a (Y-DNA), I4a (mtDNA)
A80302 - D4m2 (mtDNA)
A80410 - G (Y-DNA)
A80411 - J2a (Y-DNA)

The population of the Khazar Khaganate / of the Saltovo-Mayaki culture, was ethnically heterogeneous:

https://en.wikipedia.org/wiki/Saltovo-Mayaki

"Saltovo-Mayaki is the name given by archaeologists to the early medieval culture of the Pontic steppe region roughly between the Don and the Dnieper Rivers. Their culture was a melting pot of Onogur, Khazar, Pecheneg, Magyar, Alan, and Slavic influences. During the ninth century the Saltovo-Mayaki culture was closely associated with the Khazar Khaganate, and archaeological sites from this period are one way that historians track the geographic scope of Khazar influence."

====================

The most common Jewish subclade of R1a, is this: R1a1a1b2a2b1a1 M582 (aka CTS6).

It descends from R1a1a1b2a Z94 - one sample of which was found in Saltovo-Mayaki.

It's also found in the Near East, isn't it? If so, it could have entered the Jewish genome with "Iranian like" migrations into the Near East. So, again, inconclusive.

Also, it doesn't form a large enough percentage of the y lineages of Ashkenazim to account for 50-60% of their autosomal inheritance, although if it entered their population in eastern Europe it might explain the 15% or so that is proposed with some of these Xue models.

I suppose I should say that I do think it is possible that the ancient Hebrews had rather more eastern Anatolian inheritance than the "Canaanites", since the mythology of their origins as expressed in the Old Testament would allow for that, with the connections to the upper Euphrates and Hurrian culture.* However, I do think there was some ancient interaction with Egyptians, so you would think it would have been somewhere in between these people and the Bronze Age Levantine sample, but again, who knows.

*http://www.jewishvirtuallibrary.org/...y/hebrews.html

[Tomenable]

^ I think testing some ancient DNA samples from actual Jewish burials could help.

There are old Jewish graves in Roman catacombs, and there are also these graves:

http://www.eupedia.com/forum/threads...l=1#post475572

It's also found in the Near East, isn't it?

Yes. R1a-Z94 has a very widespread distribution.

We need higher resolution of SNPs to determine where did the Jewish branch come from.

[davef]

It's also found in the Near East, isn't it? If so, it could have entered the Jewish genome with "Iranian like" migrations into the Near East. So, again, inconclusive.

Also, it doesn't form a large enough percentage of the y lineages of Ashkenazim to account for 50-60% of their autosomal inheritance, although if it entered their population in eastern Europe it might explain the 15% or so that is proposed with some of these Xue models.

I suppose I should say that I do think it is possible that the ancient Hebrews had rather more eastern Anatolian inheritance than the "Canaanites", since the mythology of their origins as expressed in the Old Testament would allow for that, with the connections to the upper Euphrates and Hurrian culture.* However, I do think there was some ancient interaction with Egyptians, so you would think it would have been somewhere in between these people and the Bronze Age Levantine sample, but again, who knows.

*http://www.jewishvirtuallibrary.org/...y/hebrews.html

I don't think we can prove that those Bronze Age levantines represented the majority of Canaanites well. And the low ibd sharing with Italy and ironically the Middle East (the Middle East ibd being even lower) in and with the Ashkenazim being more Southern European and middle eastern than Northern European (some of the northerly stuff could've came later) according to the admixture test might just lead to the possibility that some of the Middle East was more european like than it is today. I've seen gedmatch numbers from today's levantines and they don't get much of any EEF in some calculators and if they get any european percentages it would be something like 5 percent some southern euro group.
My biggest bet favors that the Southern European score did not come from Southern Europeans, that includes greeks, italians, sicily, etc though the Iberian ibd could indicate some spanish but in this study Southern European means italian or greek.
I'm not insisting on the idea in the previous paragraph above, it's just something i kinda have more faith in. I can't insist on anything without more actual genomes, but I won't bet as much on the "South European" coming from Europe.

Btw I was only suggesting that perhaps parts of the Middle East was more euro like. I'm not like those boneheads on other forums who try to claim that all the great civilizations of the Middle East were once like eiropeans until they started mixing with slaves or some other quacky idea. If I wanted to convince people that events like these happened, I wouldn't post them here since Angela knows all the tricks in the book.

[Angela]

I don't think we can prove that those Bronze Age levantines represented the majority of Canaanites well. And the low ibd sharing with Italy and ironically the Middle East (the Middle East ibd being even lower) in and with the Ashkenazim being more Southern European and middle eastern than Northern European (some of the northerly stuff could've came later) according to the admixture test might just lead to the possibility that some of the Middle East was more european like than it is today. I've seen gedmatch numbers from today's levantines and they don't get much of any EEF in some calculators and if they get any european percentages it would be something like 5 percent some southern euro group.
My biggest bet favors that the Southern European score did not come from Southern Europeans, that includes greeks, italians, sicily, etc though the Iberian ibd could indicate some spanish but in this study Southern European means italian or greek.
I'm not insisting on the idea in the previous paragraph above, it's just something i kinda have more faith in. I can't insist on anything without more actual genomes, but I won't bet as much on the "South European" coming from Europe.

Btw I was only suggesting that perhaps parts of the Middle East was more euro like. I'm not like those boneheads on other forums who try to claim that all the great civilizations of the Middle East were once like eiropeans until they started mixing with slaves or some other quacky idea. If I wanted to convince people that events like these happened, I wouldn't post them here since Angela knows all the tricks in the book.

Interesting choice of words: are a knowledge of archaeology, history, and genetics, and the use of reason and logic "tricks"???

As to the Canaanites, what better evidence could there be about the autosomal signature of the Canaanites than an ancient sample from the same time and place as the Canaanites? I don't think it gets any better than that. That indicates the Canaanites and other peoples of the Levant at that time were probably very Saudi like and Egyptian like, and lo and behold, the Palestinians and Jordanians in the analysis at 23andme cluster not with Anatolians or Syrians or Iranians or people from the Caucasus, but with Saudis and Egyptians, although they plot a bit "north" of them (not that much, however). I used to think that was because of a lot of influx from Arabia into the Levant with the Muslim migrations, but maybe not, or at least not to any great extent. The Lebanese are probably a decent example of a "Canaanite" like population. They plot a bit north of the Jordanians, etc., but they might have a bit of "Hellenic" admixture and/or a bit of Crusader.

If Iranians are still mostly Iranian Chalcolithic, and some of the Bedouin, if you subtract out the recent SSA are pretty darn Natufian like, then I think it's possible that Canaanites (Lebanese) and Palestinians aren't that far removed from the Levantine Bronze.

The Jews might have been different. Did you read the article from the Jewish Virtual Library? High levels of admixture from the more northern Near East are indeed possible for them, and given their extreme endogamy might have become fixed. It may also be that they absorbed some ancestry from the Sea Peoples as they were very close to the coast at the time. I think we may be able to get a hint about their admixture at that time by looking at the Samaritans and the Druse, who are quite different from the Palestinians and the Jordanians.

All of this is going to have to wait for ancient dna.

Btw, I have no idea how you could think the Jews of that time might have been "European" like, a statement you've made before. There is to my knowledge no documented "European" migration to the Levant large or small in the relevant time frame. Until these "Sea Peoples" dna is analyzed we won't know if they came from Crete or Anatolia, and we certainly won't know how much ancestry they could have contributed to the Jews and whether that would have been enough to turn them into a "European" population. We'll have to wait and see. Even most of the "Hellenic" people were just Hellenized Near Easterners. Did these "European like" Jews drop out of the sky from a space ship? Even wild speculations have to be based on some data.

No offense meant, but you place way too much emphasis on admixture runs from gedmatch, in my opinion. They're one, and not the most reliable, of the genetic tools available. Also, the fact that today's Near Easterners get so little "EEF", or maybe you mean "Anatolian Neolithic", or Early European farmer" just shows they aren't "European". EEF is preeminently a European marker today, although the majority of the genes in that cluster entered Europe with farmers from Anatolia. After the Anatolian farmers left the Near East ten thousand or so years ago there were massive demographic changes there, although mainly through the agency of another "Basal Eurasian" rich farmer group, and then more minor amounts of ancestry from EHG through the Indo-Europeans. You can also add in some additional ASI and SSA. Now, during the latter part of the first millennium, and then into the Roman Era, they increasingly became traders, and there was at that time no prescription on marriage with "foreign" women, so the genome of diaspora Jews could have been different from that of the Palestine Jews. That's why we need their genomes as well, although again, if the admixture was with Italians or Greeks why is there no real IBD sharing? Italians definitely seem out, which will disappoint all my Ashkenazi friends, but perhaps the "Greek" population used for the comparison did not include island Greeks, Cretans, or Pontic Greeks, and that would make a difference.

[davef]

Thanks!
Yeah, I meant that you can't be fooled easily when I made the tricks in the book statement.

Ok so around the time when the Bronze Age levantines were around, nobody in the Mideast during that time had much (if at all) to share with europeans. The anatolian farmers were gone from there thousands of years prior.

Now that you brought that up, the Southern European score could've in fact been from mixing with an actual, unknown european population, since jews seem to have much more in common with europeans than modern levantines. It's just a matter of getting ancient genomes at this stage.

I know by now that gedmatch or admixture isn't the ultimate predictor, but it does point out that jews have more in common with Southern Europe than levantines do, at least to me it does. That's as far as it goes. It fails to point out the source. Ancient genomes and ibd would be pretty much the only way short of a time machine we can figure out the population that contributed to the "south european score."

[srdceleva]

one majory counfusion in this thread for this whole discussion is the IBD sharing between Aj and the other ME and EU populations. When i was reading the study, when they were discussing IBD sharing between the populations it didnt seem to be an issue at all for the authors that AJ shared the largest IBD segments with Eastern europeans, then next with Iberians, then next with southern europeans, and then next with middle easterners. This makes total sense according to the evidence suggested in the study, that AJ had more recent ancestry from Eastern europeans then slightly older or pre bottleneck ancestry southern europeans, and then ancestry from the original levantine Jews. Obviously you will share the longest strands of your dna with recent people who admixed with you, so this makes total sense that if Eastern eurpeans mixed most recently with AJ that they will share the largest identical strands of dna with them. If admixture was from along time ago such as the ME admixture in AJ then you would expect for modern ME and AJ to share very little IBD segments as the length of Identically shared ancestry would be broken up thousands of times and would not remain identical anymore. Higher shared IBD with Iberia is easily explained away by the recent gene flow in Iberia from Sephardic Jews. The sephardic Jewish ancestry in modern Iberians is quite significant. Maybe I dont understand how it all works quite well enough but thats what I understood from it. I still do think its best to wait for ancient dna evidence.

[davef]

Srdceleva, that was absolutely brilliant. If you relate the order of the ibd lengths to the order of the proposed admixture events, you'll see how they match. The shortest is with levantines, but jews are from the Middle East, and derive from Bronze Age levantines. When they left, their Levantine cousins became Lebanese, etc and formed different sequences and eventually would drift from the jews overtime. Southern European (one that is about as southerly as south italians, not saying its from them) came next, and over time the south euro contributors drifted out genetically from European Jews so ibd got worse overtime. Pattern applies with the other contributors.
Not saying this happened for sure, we lack hard evidence but your idea is very interesting and makes a lot of sense.

[Angela]

one majory counfusion in this thread for this whole discussion is the IBD sharing between Aj and the other ME and EU populations. When i was reading the study, when they were discussing IBD sharing between the populations it didnt seem to be an issue at all for the authors that AJ shared the largest IBD segments with Eastern europeans, then next with Iberians, then next with southern europeans, and then next with middle easterners. This makes total sense according to the evidence suggested in the study, that AJ had more recent ancestry from Eastern europeans then slightly older or pre bottleneck ancestry southern europeans, and then ancestry from the original levantine Jews. Obviously you will share the longest strands of your dna with recent people who admixed with you, so this makes total sense that if Eastern eurpeans mixed most recently with AJ that they will share the largest identical strands of dna with them. If admixture was from along time ago such as the ME admixture in AJ then you would expect for modern ME and AJ to share very little IBD segments as the length of Identically shared ancestry would be broken up thousands of times and would not remain identical anymore. Higher shared IBD with Iberia is easily explained away by the recent gene flow in Iberia from Sephardic Jews. The sephardic Jewish ancestry in modern Iberians is quite significant. Maybe I dont understand how it all works quite well enough but thats what I understood from it. I still do think its best to wait for ancient dna evidence.

Take a look at the graphic of the IBD sharing, and also read the caption:

IBD analysis of Ashkenazi Jews-Xue et al.PNG
Both the less than 7cm and more than 7cm sharing is highest with Eastern Europeans.

Also, the length of the segments is not the only important factor; one also has to consider the mean number of segments.

[srdceleva]

Srdceleva, that was absolutely brilliant. If you relate the order of the ibd lengths to the order of the proposed admixture events, you'll see how they match. The shortest is with levantines, but jews are from the Middle East, and derive from Bronze Age levantines. When they left, their Levantine cousins became Lebanese, etc and formed different sequences and eventually would drift from the jews overtime. Southern European (one that is about as southerly as south italians, not saying its from them) came next, and over time the south euro contributors drifted out genetically from European Jews so ibd got worse overtime. Pattern applies with the other contributors.
Not saying this happened for sure, we lack hard evidence but your idea is very interesting and makes a lot of sense.

thanks for the nice comments

Take a look at the graphic of the IBD sharing, and also read the caption:

IBD analysis of Ashkenazi Jews-Xue et al.PNG
Both the less than 7cm and more than 7cm sharing is highest with Eastern Europeans.

Also, the length of the segments is not the only important factor; one also has to consider the mean number of segments.

i still think it falls in line with what I said, and in the description above the graph it says also the same conclusion, that this shows that most of the recent genetic influence in AJ is from Eastern Europe.

[davef]

thanks for the nice comments



i still think it falls in line with what I said, and in the description above the graph it says also the same conclusion, that this shows that most of the recent genetic influence in AJ is from Eastern Europe.

Yeah, they both show Eastern Europe>(Iberia or Western Europe-they seem to have different ranks in both charts, but they are somewhat close)>Southern Europe>MidEast
which makes sense given what we believe happened during the European journey.

And I don't know if we should say low ibd leads to not having any gene flow from a certain place. To say they haven't gotten any from Southern Europe is like saying they haven't gotten any from the MidEast, which is odd since that's where they are from...but we have the challenge of using modern populations for comparison, since Jews have been endogamous for centuries and other populations, even the ones where the Jews MAY have gotten what they got went their own separate ways genetically and drifted out. If they would be so kind as to replace the modern Levantines with Bronze Age Levantines, and the modern Southern Euros with ancient ones (Angela brought up the island types) the IBD should be higher between either of these populations or both. Jews descend from non-Jews. Period. It's not like they're a separate hominid species, or crash landed on Planet Earth from the outer reaches of the galaxy.

"We...bring...you...peace" (finger glows)

Let me know any issues you have with what I have said

[srdceleva]

Yeah, they both show Eastern Europe>(Iberia or Western Europe-they seem to have different ranks in both charts, but they are somewhat close)>Southern Europe>MidEast
which makes sense given what we believe happened during the European journey.

And I don't know if we should say low ibd leads to not having any gene flow from a certain place. To say they haven't gotten any from Southern Europe is like saying they haven't gotten any from the MidEast, which is odd since that's where they are from...but we have the challenge of using modern populations for comparison, since Jews have been endogamous for centuries and other populations, even the ones where the Jews MAY have gotten what they got went their own separate ways genetically and drifted out. If they would be so kind as to replace the modern Levantines with Bronze Age Levantines, and the modern Southern Euros with ancient ones (Angela brought up the island types) the IBD should be higher between either of these populations or both. Jews descend from non-Jews. Period. It's not like they're a separate hominid species, or crash landed on Planet Earth from the outer reaches of the galaxy.

"We...bring...you...peace" (finger glows)

Let me know any issues you have with what I have said

yes i think were on the right track. From my understanding IBD and proving recent genetic relatedness in general, for example between cousins or close family, is based on the fact that you share large portions of identical dna with that person, and as angela pointed out, the frequency of these segments. This proves that you have a recent genetic relationship. You may be almost one hundred percent of a different ethnicity as someone but if you have a recent common ancestor from not too long ago you will share higher rates of IBD with that person than with someone of your own ethnicity who isnt as close of a cousin because youve been seperated from them for a long time. The Jews have literally been seperated form levantine peoples for over 2000 years. As time goes on that dna gets broken up into so many parts that the length of identical dna becomes very small, though you are of the same ethnic group. This makes total sense. My overall genetic make up may resemble a guy from eastern europe, but because im a quarter american there maybe be many americans who i share more recent ancestry with than many eastern europeans.

[davef]

I know, it makes total sense. They are descended from Bronze Age levantines and I bet if the calculator used in this student were used on the Bronze Age, they might not get any european. Maybe a small percentage of south european in the single digits. Ibd sharing with modern south Europe is small but it's certainly higher than with the modern Middle East.

Yeah ibd is "enough" sharing in a row as opposed to scattered short segments here and there, which admixture calculators go by (by state). The 47 percent Mideast is IBS sharing with the levant so 47 percent of the total genome is admired with levant, and one can be tricked into thinking that a gang of people who were 3/4 south italian 1/4 north euro mingled with a group of people who were of full Lebanese descent and together they formed Ashkenazim.

Also you said you are 1/4 American...you mean Native American ,colonial (English/german or some other northern descent), or having a grandparent born in the US of another background besides the colonial/native?