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

[davef]

Not a problem, Angela! Apology accepted!

 

[davef]

Vallicanus,
Angela pointed out that North African Jews do not have italian ancestry and I agree with her. The "italian" could be explained by bad methodology on 23andme's part,the possibility that the poster had flat out lied about their ancestry, or that they aren't fully North African jewish, but mixed with italian unlike anyone who is full blooded North African jewish. I had to coin my words precisely so no one misreads them and thinks I was implying that North African jews have italian ancestry which isn't the case.

There isn't ibd sharing between North African jews and italians and I wouldn't gloss too hard over those results.

I apologize for veering off topic.

 

[Angela]

Razib Khan has posted on the paper. See:

See: http://www.unz.com/gnxp/the-origins-of-ashkenazi-jews-near-resolution/#comments

A few tidbits:
"The likely parental populations of Ashkenazi Jews are Roman period peoples of the eastern Mediterranean, particularly the swath of territory from Alexandria up to Anatolia, and, the peoples of the western Mediterranean. That is, Levantines and Iberians & Italians."

Well, maybe it's likely, but we still don't have the proof, in my opinion, whatever the authors may claim.

"Additionally, the more and more we learn about the Middle East, the more likely it seems that Muslim populations, who are often modeled as a parental group, are highly cosmopolitan compared to ancient groups. Recall that Neolithic farmers from the Levant resemble Sardinians more than they do locals, because of later migration from further east in Eurasia, as well as later African gene flow. Using imperfect reference populations will probably skew the results appropriately."

This is probably why we won't solve this mystery until we get ancient dna from the right periods.

"The major change in the past few years is the usage of more genetic information than common genotypes. This paper for example looks at haplotype information. Sequences of variants across the genome. This preserves more recent genetic variation. In other cases you can look at whole genome sequences, and focus on low frequency variants which are extremely informative of recent population differentiation."

I wish he'd addressed the concerns expressed in the paper about how problematic and unhelpful some of these programs are when analyzing southern European genetics.

This person offered one possible explanation:
"Priori assumptions about ME|Euro 50|50 … SW Asian in Ashkenazi is 11-13% … W-Asian in Jews 17-22% and not entirely derivative of the Levant. 252K is fine for Admixture or PCA but rather light for Alder, ChromoPainter and IBD. Ashkenazi ethnogenesis far more nuanced than presented here."

There's also some interesting discussion in the comments about what caused the bottleneck. I agree with the posters who hold it's probably small population size to begin with, combined with the pogroms.

There's also an amusing bit where Khan typically loses his temper again when confronted with the same old argument, made, I would be willing to bet, by the same person who used to repeat this stuff over and over again on this forum, that Italians are the product of at least a partial replacement by Middle Easterners during the Roman Empire, and all based on the fact that there is an increasing number of grave inscriptions with "Greek" names in that period.

I could have written the response myself, although, unlike Khan, I haven't looked at thousands of Italian genomes.

I also don't agree with even "possible" parity in Sicily, given the uniparental story.

Anyway, it's off topic for this thread, so I won't include it here, but the exchange is worth a read.

 

[davef]

Razib Khan has posted on the paper. See:

See: http://www.unz.com/gnxp/the-origins-of-ashkenazi-jews-near-resolution/#comments

A few tidbits:
"The likely parental populations of Ashkenazi Jews are Roman period peoples of the eastern Mediterranean, particularly the swath of territory from Alexandria up to Anatolia, and, the peoples of the western Mediterranean. That is, Levantines and Iberians & Italians."

Well, maybe it's likely, but we still don't have the proof, in my opinion, whatever the authors may claim.

"Additionally, the more and more we learn about the Middle East, the more likely it seems that Muslim populations, who are often modeled as a parental group, are highly cosmopolitan compared to ancient groups. Recall that Neolithic farmers from the Levant resemble Sardinians more than they do locals, because of later migration from further east in Eurasia, as well as later African gene flow. Using imperfect reference populations will probably skew the results appropriately."

This is probably why we won't solve this mystery until we get ancient dna from the right periods.

"The major change in the past few years is the usage of more genetic information than common genotypes. This paper for example looks at haplotype information. Sequences of variants across the genome. This preserves more recent genetic variation. In other cases you can look at whole genome sequences, and focus on low frequency variants which are extremely informative of recent population differentiation."

I wish he'd addressed the concerns expressed in the paper about how problematic and unhelpful some of these programs are when analyzing southern European genetics.

This person offered one possible explanation:
"Priori assumptions about ME|Euro 50|50 … SW Asian in Ashkenazi is 11-13% … W-Asian in Jews 17-22% and not entirely derivative of the Levant. 252K is fine for Admixture or PCA but rather light for Alder, ChromoPainter and IBD. Ashkenazi ethnogenesis far more nuanced than presented here."

WOW!!! I just scoped the date of the article and it came out two days ago!!! Im reading it later, thanks Angela!!!!!

 

[davef]

"Priori assumptions about ME|Euro 50|50 … SW Asian in Ashkenazi is 11-13% … W-Asian in Jews 17-22% and not entirely derivative of the Levant."
Why is he insisting "not entirely derivative of the Levant?" Just curious

Razib also said:
"Ultimately the only reason I’d suggest that this paper is lacking is the imperfection of Middle Eastern source populations. That’s probably increasing the European and decreasing the Middle Eastern fraction somewhat on the margins"
Could you also explain why he said this?

 

[Angela]

"Priori assumptions about ME|Euro 50|50 … SW Asian in Ashkenazi is 11-13% … W-Asian in Jews 17-22% and not entirely derivative of the Levant."
Why is he insisting "not entirely derivative of the Levant?" Just curious

Razib also said:
"Ultimately the only reason I’d suggest that this paper is lacking is the imperfection of Middle Eastern source populations. That’s probably increasing the European and decreasing the Middle Eastern fraction somewhat on the margins"
Could you also explain why he said this?

The Middle Eastern source populations being used are imperfect, as we've already said may times, because they're modern samples, and we don't know how close the Jews of the first millennium BC and the early centuries of the Common Era were to modern Palestinians/Jordanians, or even Druze, although they may be closer, as they have more northern Near East admixture.

If the Jews of the beginning of the Common Era were like modern Palestinians, that's one thing. If they were more like the Druze, for example, being less SSA, more northern Near East, then they probably picked up less admixture in Europe.

The Levant Bronze Age clustered with Saudis. We don't know, however, the genetic signature of any of the inhabitants of the Levant by the late Iron Age. Did they by that time have more Northern Near Eastern affinities, and then they changed again with the Arab invasions and the slave trade?

Even if they were still Levant Bronze Age like in the first millennium BC, had the Jews, in particular, given that a good number of them had become merchants, and now actively proselytized or at least willingly absorbed a lot of non Jewish women into their communities, already changed to a certain degree by the first centuries of the Common Era, particularly in the communities in Anatolia? We just don't know.

In this regard, there were at that time large communities of Jews throughout Anatolia, and a large one even in Cyprus. When the Apostles and other disciples of Jesus traveled the ancient world to preach the Gospel, they worked from the Jewish communities in the beginning. When the Jews of the Levant, after two disastrous wars with Rome, were by the tens of thousands either killed or sold into slavery, the future of the Jewish people may have rested on these already slightly changed diaspora communities.

In that regard I've often wondered why more analysis isn't done with the Italian Jews, who were in Rome from very early times. I suppose by now there's been intermarriage with other European Jews, and so the waters are muddied.

We just don't know. Khan seems to be leaving the possibility open that the Jews had already become a bit more northern and western by the time of the major diaspora movements, although he thinks the differences might be small given that he says it would only change the percentages at the margins.

Of course, none of this answers why we're not seeing any significant IBD sharing from that period with any southern European populations.

 

[davef]

Percentages at the margins? Don't know what that means. Plus, you mentioned that if they were more like "northern near east" then they would've picked up less European admixture...does this mean that the northern near east had more in common with modern europeans?

 

[davef]

If the jews descend from Bronze Age Canaanites, then a little Slavic won't be enough for them to plot in Europe. There had to be some intervention, and its doubtful that it was from " italians". Im not fooled by the " jews are descended from italians/Romans" theory. We do have philistines that were dug up last week, just last week and I would definitely use them if we ever want to figure out the reason why jews plot in Europe. They are an ancient population from Mediterranean Europe who have quite a history with ancient jews.

Neolithic levantines don't plot in Europe, nor do Palestinians, Druze, or other levantines. Far from it.

Ancient genomes will teach us more than anything else. We have them. We must use them. The days of using modern levantines and italians/greeks should come to an end.

 

[Angela]

Percentages at the margins? Don't know what that means. Plus, you mentioned that if they were more like "northern near east" then they would've picked up less European admixture...does this mean that the northern near east had more in common with modern europeans?

I think it's a pretty well understood term; it means "a bit more". It comes from the financial sector. Obviously, he can't quantify it without ancient genomes.

https://answers.yahoo.com/question/index?qid=20110928124639AA2tFEH

As for your second question, let's take the example of just the SSA contribution to modern Levantine genomes. By some analyses, the Palestinians are 7-8% SSA, Jordanians about 7%, Lebanese 3 1/2%, but the Samaritans are 2%, the Druse 2%. If the Jews of the first centuries of the Common Era were more like the Druze than the Palestinians, then they would need less Southern European admixture to pull them to their present position.
https://docs.google.com/spreadsheet...1sy2-kSWVhNJvanIsWtJU/edit?pref=2&pli=1#gid=0

So, as I suggested above, perhaps there was a change from the Levant Bronze Age to the Iron Age and then a change back to a more Bronze Age profile with the Muslim invasions and the Arab slave trade. Or maybe not. We just don't know yet.

We could try to look at what we know about the ethnogenesis of the Samaritans and the Druze, but we don't know a whole lot. The Samaritans claim descent from the tribes of Ephraim and Manasseh, but the "Jews" of the period said they were half Jews descended from the Jews who remained in Palestine when most Jews were taken into captivity in Babylonia, and the "foreigners" settled on the land by the Assyrians.
https://en.wikipedia.org/wiki/Samaritans

You might find this blogpost by Razib Khan interesting.
http://blogs.discovermagazine.com/g...ws-3000-years-of-common-history/#.V5IX_rgrLNM

As to the Druze, I don't think anyone really understands their ethnogensis. They speak Arabic, their religion grew out of Ismailism, but a lot of Muslims don't consider them Muslim. Was there input into their genome from Iran? We don't know. They incorporated a lot of Plato's ideas, and they could be considered Gnostics on one level, but does that mean they absorbed "Greek" ancestry? I doubt it, but who knows. Or, were they perhaps just relatively isolated highlanders whose adoption of this religion and subsequent endogamy meant that they were not as affected by yet another flow north from Arabia? Again, I don't know, and I don't think anyone else does either.

I'm content to wait for some more clarity from ancient DNA, but then I don't have any "theory" or "agenda" to push.

 

[davef]

Thanks angela! That makes perfect sense, and that's pretty much in line with what popped in my mind when razib said that if they were more like Druze who are closer to eiropeans than most other levantines, then less european would've gotten them where they are. That's easy.

And regarding another comment, someone mentioned that 297k isn't enough for ibd, but I don't quite know why he's complaining about the ibd results. They seem pretty consistent with other papers and make historical sense.


I'm not agenda driven either. What I like to do is put ideas on the table and ask educated people of their opinions about those ideas. I wouldn't come to this forum if I was agenda driven because agendas are frowned upon, and I would get nowhere because (thankfully) someone like you would intervene and point out the holes.

 

[Angela]

The IBD results don't make sense because there is not enough IBD sharing with any Europeans to pull the Ashkenazim away from even the Druze, much less Palestinians.

It's not precisely a function of the Druze being more "European"; it's more a question of the Druze being more "West Eurasian" than the modern Palestinians. The Druze have very little "European", per se.

I don't know if it was you, but I think someone also said that the Early Farmers plot in the Middle East. Actually, they don't at all. You can see it with this plot from Lazaridis, where the EN, who only picked up about 7% new gene flow after leaving the Near East, are not near the modern Near Easterners at all.

Also, in terms of the Druze, they don't plot anywhere near those farmers either. The Druze are not a remnant population of Anatolian farmers. The closest we have to that is the Sardinians.
http://1.bp.blogspot.com/-QJs5tV0H3t4/VIwY7BlNxyI/AAAAAAAAByQ/LakjgsBYRNc/s1600/2dkweaw.jpg

The Druze have also been pulled toward Africans as well as towards ANE.

 

[davef]

Wait, do you even need ibd to pull away? I thought that ashkenazi pull away in admixture tests which don't account for ibd and I understand those tests can't be taken literally and don't say enough.

I guess what you meant by that statement is that in order to plot closer to europeans, one must have genetic input from an actual european population which would result in ibd sharing with that population.... . Don't worry, that's not news to me !!

I bet ibd with certain ancient europeans and levantines (especially levantines) would be much higher. Willing to bet money on that. I'm not hand picking modern ones. Why can't it be? It would be crazy to score that much south european (35 percent) and middle eastern (about half) and not ibd with any european or middle eastern group that existed back then.
Why would it not be crazy? That question was not rhetorical.

Ok yeah general west Eurasian that can be found from British to Saudis replacing ssa would definitely move someone more towards certain purer Eurasians.

 

[davef]

I see you changed your post. I don't think I said that, but I did once suggest that parts of the Middle East used to be more european or Anatolian farmer related even after most of the Anatolian farmers left, but with Bronze Age levantines we know how that's wrong,
Ashkenazim sat pretty close to the Anatolians. They are about on par with Sicilians when it comes to Anatolian-ness. Don't take this as suggesting they have sicilian ancestry. Yeah, Sardinians are definitely the closest.

 

[oreo_cookie]

Ashkenazim sat pretty close to the Anatolians. They are about on par with Sicilians when it comes to Anatolian-ness. Don't take this as suggesting they have sicilian ancestry. Yeah, Sardinians are definitely the closest.

This right here is why people assume that Ashkenazim have Italian DNA. Because Sicilians and Ashkenazim on PCA plots, are both roughly halfway between North Italians and modern day Levantines, people say "well, Ashkenazim are Levantine on the male side, North Italian on the female side, and thus plot like Sicilians when you mix it all up." But Sicilians have very low IBD sharing with the Levant, which implies that West Asian input into Sicily occurred in the Bronze Age rather than from Phoenicians or anything more recent, and that the Sikels, Sicanians, and Elymians at the time of contact with Phoenicians had already been genetically shifted toward West Asia.

At this point, Ashkenazim would not have existed... they'd still be Israelites. My guess is that ancient Israelites were probably closer to Druze than to Lebanese Arabs or Palestinians, and thus we should look at Druze or Cypriots as a good proxy, which means that you'd need closer to 1/3 South European admixture, not 1/2, to explain European Jews' plotting today. They only land in Sicily because Sicilians have enough West Asian affinity that they drift away from the core European cluster, it's not due to any directly shared ancestry. The same would be true of Cretans, who also are very close to Ashkenazim autosomally.

IBD sharing, above all, reflects how recent the exchange took place. Sicilians DO have high IBD sharing with Maghrebi people, implying that the North African input into the population is more recent. I am unclear if Ashkenazim have the same.

Not that this is scientific, but I have seen a few Druze results on GEDmatch and compared to Lebanese Arabs, they shift toward Iberia, Sardinia, etc. and their overall closest population is Cyprus.

 

[davef]

They could've used cypriots and crete for their south european sample pool. It would've been the more common sense approach. It helps that those two islands are way closer to the levant than italy.

Crete similar autosomally? Do you have data and papers to back this?

Yes I don't subscribe to italian anvestry in jews either.

 

[oreo_cookie]

They could've used cypriots and crete for their south european sample pool. It would've been the more common sense approach. It helps that those two islands are way closer to the levant than italy.

Crete similar autosomally? Do you have data and papers to back this?

Yes I don't subscribe to italian anvestry in jews either.

In terms of autosomal DNA Cretans are close to Sicilians (but again, their y-dna pool shows Slavic influence via R1a and I2), Cypriots are not even in the European cluster on any PCA plot, they are basically Levantines themselves.

 

[Angela]

In terms of autosomal DNA Cretans are close to Sicilians (but again, their y-dna pool shows Slavic influence via R1a and I2), Cypriots are not even in the European cluster on any PCA plot, they are basically Levantines themselves.

Point number 1: This is not a thread on Italian DNA, although I'm sure you would prefer that it were...

Point number 2: Even if it were, here on this Board we don't give much credence to pronouncements for which proof has not been offered.

Point number 3: Proof consists of academic papers, not "personal" admixture calculator results of unknown people, whose ancestry and results are impossible to verify, and which even if verifiable do not come from a random statistical sample.

Now, this thead is about Xue et al. Feel free to post about it if you have something new to add. We really don't need to hear the same point made over and over again, especially on a summer Saturday some of us might actually want to go out and enjoy.

 

[oreo_cookie]

davef brought up Sicily, not me, and he should be informed that Cypriots are not genetically typical Southern European, they are significantly Levantine themselves so it would never make sense to use them as a reference sample for "European" DNA in Jews. Anyway, I do not have anything to add in here other than what I said, so I won't post in this thread again.

 

[davef]

Oh actually you caught my eye when you said Cretans are autosomally similar to Ashkenazim, so I asked if you have a paper to confirm. It would be interesting if ibd sharing is higher there in comparison to the low ibd sharing reported in the paper with south europeans.

 

[oreo_cookie]

Oh actually you caught my eye when you said Cretans are autosomally similar to Ashkenazim, so I asked if you have a paper to confirm. It would be interesting if ibd sharing is higher there in comparison to the low ibd sharing reported in the paper with south europeans.

If I find it I'll send it -- if you have other questions message me. I don't want to derail the thread any more than I have, my apologies to other users here.