search this blog

Monday, September 26, 2016

Estonian Biocentre Human Genome Diversity Panel (EGDP)


Published along with Pagani et al. 2016, the EGDP dataset is freely available at the Estonian Biocentre website as VCF and PLINK binary files here. It overlaps at ~550K SNPs with Broad MITs/Harvard's Human Origins, and at an impressive ~1.1 million SNPs with the ~1.2 million SNP ancient DNA chip used by the Reich Lab and others.

To see what's what, I ran a Principal Component Analysis (PCA) of all of the samples except the Congo Pygmies. I then removed four Siberians that behaved as if they had very recent European ancestry, and reran the PCA. Below are a few screen grabs from the latter analysis. The datasheet is available here.

I also tested a few of the samples with my Basal-rich K7 model. You can see their results in the spreadsheet here. Look for the individual IDs with the GS prefix.

Using the K7 spreadsheet and nMonte, here's a model for Ashkenazi Jews with some of the new EGDP populations as references, including Avars from the North Caucasus and Arabs from Israel. The Arabs do help to improve the fit, but they're not as important as Samaritans and Tuscans.

Ashkenazi
Anatolia_ChL 7.9
Arab_Israel_1 15.65
Avar 0.6
Bashkir 0.05
Cossack 0
Italian_Tuscan 30.45
Polish 11.75
Samaritan 33.6
Uygur 0

distance%=0.2874 / distance=0.002874

In any case, that looks like a fairly sensible outcome, considering that it only took me a few minutes to put together. I've seen much worse in scientific literature.

173 comments:

Nirjhar007 said...

Very good. BTW how does Ashkenazi do with CHG? . Do they show more than other Jewish groups?.

Tesmos said...

How is your new calculator going? any ETA?

Richard Rocca said...

@David, your comment about the importance of Tuscans... could some Western German (Palatinate) ancestry be getting lumped into Tuscans as a best fit? Either way, this correlates with the following pretty well...

http://biorxiv.org/content/early/2016/07/10/063099

Our model of the AJ admixture history is presented in Figure 7. Under our model, admixture in Europe first happened in Southern Europe, and was followed by a founder event and a minor admixture event (likely) in Eastern Europe. Admixture in Southern Europe possibly occurred in Italy, given the continued presence of Jews there and the proposed Italian source of the early Rhineland Ashkenazi communities.

Gioiello said...

@ Davidski

Davidski, as you can see, some people share the same thinking:
" Admixture in Southern Europe possibly occurred in Italy, given the continued presence of Jews there and the proposed Italian source of the early Rhineland Ashkenazi communities" (Rocca).
"Today, 02:42 PM Agamemnon replied to a thread L817 in the Arabian Peninsula in J1-M267
That's to be expected, L816 looks uniquely Jewish and its TMRCA fits within the time frame of the Ashkenazi bottleneck".

These two people think that Europeans derive from Jews (and not the other way around) and that if Jews have a MRCA about 1 thousand years ago, i.e. 1 thousand years after the diaspora, it is due to a bottleneck and not to introgression from European or other people, also some Khazars or Eastern Europeans: it is well known, at least to whom undertands a little of genetics, that the high percentage of CCR5delta32 in Ashenazic people came from Eastern Europe, being that mutation very low in what was the Roman Empire and practically to 0,00% in Middle East.
I think having demonstrated where the single haplotypes both for the Y and the mt derive. My 10000 letters are at disposal of everyone.
It is hilarious to read something in the Anthrogenica in French about me (yes. Gioiello Tognoni, with name and surname, and Agamemnon's answer).

Gioiello said...

@ Davidski

Your work has been fast and interesting, but, as Italy had two refugia (one in North from where Tuscans derive, coming Etruscans from the Alpine zone where we found Rhaetians) and one in the South (that of mt HV and much other) and of course Poles has intakes from Baltic refugium and perhaps some other from Samara and all the R1a present in Ashklenazic people), I think that if you run Southern Italy and Mediterranean Isles, you'll find much less Arab and Samaritan and the percentage of a recent paper (full of Jewish scholars), i.e. 70% Europeans (45% Italians) couldn't be far from the truth.

ben-canaan said...

That is pretty damn impressive, Davidski. I second Richard's request, though — how about a W/SW German (/E French) reference, too? There's plenty of "Atlantic" ancestry in Ashkenazim, and I'd really be surprised if it turned out to be all Italian and Polish.

andrew said...

What is the source of the Samaritan sample?

I would assume that it is a sample of modern Levantine people belonging to a particular ethnic group, but this isn't obvious.

Lukasz M said...

Poles in those data are Estonian Poles?

BTW I made come dendrogram from your datasheet. Looks good.
Only Albos position is curious:)

http://antropologia-fizyczna.pl/images/arkusze/tree.png

Richard Rocca said...

So when the authors masked out the Middle Eastern ancestry, they were left with the following:

"Our results suggest a model of at least two events of European admixture. One event slightly
pre-dated a late medieval founder event and was likely from a Southern European source. Another event post-dated the founder event and was likely in Eastern Europe."

This would fit pretty well with the Italian_Tuscan and Polish that David found.

ben-canaan said...

3 of the reference populations for W. European are all British, and 2 of the reference populations for Middle Eastern are Egyptians and Bedouin. So I have my concerns with their estimates.

Nathan Paul said...

Samaritan 33.6

Samaritans are present day Palestinians.
So Palestinians are the original Israelites.

Now start.

ben-canaan said...

@Nathan,

Samaritans and Palestinians are not one in the same. They're obviously related, but there are substantial differences. Samaritans seem to be closest related to Lebanese Christians.

MomOfZoha said...

@ben-canaan:
"Samaritans and Palestinians are not one in the same. They're obviously related, but there are substantial differences. Samaritans seem to be closest related to Lebanese Christians."

I don't think so. Here are some Dodecad admixture results for same person (who happens to be Anatolian Turkish):
Kurds + Nogais + Palestinian + Tuscan 2.756143
Iranians + Nogais + SAMARITANS + Tuscan 2.826605
Kurds + Nogais + Syrians + Tuscan 2.832603
Nogais + Syrians + TSI30 + Uzbekistan_Jews 2.843426
Iranian_Jews + Iraq_Jews + Nogais + North_Italian 2.851023
O_Italian + SAMARITANS + Turkmens + Turks 2.861119

Closest 3-way admixture yields smaller distance (though I know all distances are large):
50% Iraq_Jews +25% N_Italian +25% Nogais 2.616107

And smallest distance over all projects is achieved by Harappa:
kurd + morocco-jew + turk-aydin + turk-aydin 2.189714
kurd + sephardic-jew + turk-aydin + turk-aydin 2.440306
kurd + morocco-jew + turk-aydin + turk-aydin 2.533944

Historically, Samaritans are said to come from Iraq -- nothing to do with most Lebanese Christians.

@Nathan Paul:
No matter how one dissects Ashkenazi genetics, it can not in good faith be used to justify modern day human rights violations. I would apply the same logic to Anatolian Turks as well, by the way, my own apparent "genetics" on both accounts notwithstanding.

It would be terrifying, otherwise, if nations begin to use "genetic litmus tests" of citizenship, don't you think?

Davidski said...

Please don't post Dodecad-based models in debates here. Dodecad tests suffer from the calculator effect and generally cause a lot of confusion as a result.

Davidski said...

@Nirjhar

Sephardi Jews show more affinity to CHG...

Israel_Natufian CHG Moroccan_Jew Ashkenazi_Jew 0.0079 2.84
Israel_Natufian CHG Sephardi_Jew Ashkenazi_Jew -0.007 -2.837

Levant_Neolithic CHG Moroccan_Jew Ashkenazi_Jew 0.0056 2.57
Levant_Neolithic CHG Sephardi_Jew Ashkenazi_Jew -0.0075 -3.737

Barcin_Neolithic CHG Moroccan_Jew Ashkenazi_Jew 0.0019 1.022
Barcin_Neolithic CHG Sephardi_Jew Ashkenazi_Jew -0.0105 -6.269

Jordan_EBA CHG Moroccan_Jew Ashkenazi_Jew 0.0032 1.527
Jordan_EBA CHG Sephardi_Jew Ashkenazi_Jew -0.006 -3.189

@Tesmos

I'm aiming for the middle of next month.

@Richard & ben-canaan

Adding eastern French improves the fit, but makes the Polish admixture go up. Germans don't affect the model, but I don't have any verified Germans from around the Rhine, and I don't want to guess what they're like. Ideally, we need some ancient samples from Iron Age and Roman Levant, Roman Italy and Greece, and Medieval western Germany to settle this once and for all.

Ashkenazi
Anatolia_ChL 0.7
Arab_Israel_1 18.65
Avar 0.2
Bashkir 0
Cossack 0
French_East 1.8
German 0
Italian_Tuscan 24.85
Polish 15.7
Samaritan 38.1
Uygur 0

distance%=0.2817 / distance=0.002817

@andrew

The Samaritans I have are from Israel, including a couple of Cohens, who appear less admixed than the others. That's all I know.

@Lukasz

Yeah, they're Estonian Poles, but apparently they were chosen because they're representative of a wider Polish population sample.

@Nathan Paul

Samaritans are not Palestinians. They're clearly different from my Palestinian population sample, with less African and CHG/Iran_Neolithic ancestry.

You have an awful posting history in the comments here, with lots of posts that look like straight up trolling attempts. You need to think more carefully before posting here again.

ben-canaan said...

Surprising, but persuasive. Easy to imagine a scenario in which by the time most of the ancestors of Ashkenazim settle in the Rhineland/Bavaria, the Germans are thoroughly Christian and ethnic separation is a given — whereas in the Slavic and Baltic lands, things remain permeable for much longer.

Curious to see a Greek/Turkish Jewish or Italian Jewish run, too. I think you and I have disagreed about the Iberianness of Sephardim — what happens when you include both Spaniards and Tuscans as reference pops?

Richard Rocca said...

@David...

Adding eastern French improves the fit, but makes the Polish admixture go up. Germans don't affect the model, but I don't have any verified Germans from around the Rhine, and I don't want to guess what they're like. Ideally, we need some ancient samples from Iron Age and Roman Levant, Roman Italy and Greece, and Medieval western Germany to settle this once and for all.

If your French_East samples are from Strasbourg, these would be German speaking French and considered part of the Rhineland population.

MomOfZoha said...

@Davidski:
Apologies. Your Eurogenes K13 supports what I stated even more clearly given the pattern of "Samaritan occurrences", again on the very same Anatolian Turkish person (you've seen this before):
1 Georgian + Nogay + SAMARITAN + South_Italian @ 3.267398
2 Algerian_Jewish + Georgian + Nogay + Syrian @ 3.401815
3 Georgian + Nogay + SAMARITAN + Sephardic_Jewish @ 3.491206
4 Georgian + Nogay + Sephardic_Jewish + Syrian @ 3.542657
5 East_Sicilian + Georgian + Nogay + SAMARITAN @ 3.544914

I posted Dodecad simply because that got lower distances. In any case, looking for "Lebanese Christian" as any component across all projects for that person fails, despite similar distances being obtained when "Samaritan" (which cropped up with some surprising frequency) was replaced (by the calculators themselves) with "Palestinian", "Syrian", "Iraqi Jewish", and "Iraqi Mandean". Even with large total admixture distances, the tiny differences between admixture vectors do mean something.

My point was just that I don't see why a non-Iraqi origin (e.g. Lebanese Christian, which are of all different kinds, by the way) is likelier than their Iraqi origin (at least in part).

And, by the way, all "Jewish" components of the (Muslim-born, and Muslim across the known generations) Central Anatolian Turk that I mention are totally and completely surprising, just as I was blindsided by the same pattern in the other "Muslim Turks" I tested -- not Ashkenazi but Sephardic/Iraqi/Iranian Jewish components, 25-50%, with the occasional additional Iraqi Mandean and Samaritan interspersed... FYI

Davidski said...

@MomOfZoha

Calcs that suffer from the calculator effect often produce models with lower distances. That's because the ancestry proportions that they report for the test samples are less exaggerated compared to those of the reference samples, which makes it easier for the oracle algorithms to find better statistical fits.

But it's a pointless exercise because the models are inherently flawed.

@Richard

I don't know where this French_East reference population was sampled. It's from the new Human Origins dataset though.

@ben-canaan

Last models. Don't have Greek or Italian Jews in this dataset. Just some Sephardi Jews. It would probably be possible to find even better statistical models, but you'd have to experiment a lot.

Sephardic_Jew
Anatolia_ChL 11.55
Arab_Israel_1 20.85
Avar 0
Bashkir 0
Cossack 0
French_East 2.25
German 0
Italian_Tuscan 1.4
Polish 0
Samaritan 36
Spanish 27.7
Uygur 0.25

distance%=0.1775 / distance=0.001775

Ashkenazi_Jew
Anatolia_ChL 0
Arab_Israel_1 14.8
Avar 1.5
Bashkir 0
Cossack 0
French_East 0
German 0
Italian_Tuscan 8.35
Polish 12.85
Samaritan 27.85
Sephardic_Jew 26.45
Spanish 8.2
Uygur 0

distance%=0.2718 / distance=0.002718

Gioiello said...

@ Davidski

It seems to me clear that in all your plots, even in the high variability, there is an invariance, i.e. the percentage of Samaritans in European Jews (Ashkenazim and Sephardim):
Samaritan 33.6
Samaritan 38.1
Samaritan 36
Samaritan 27.85
that it is about 30% and coincides with the first calculation of Wu et al. (70% European whose 45% Italian), thus 30% from pre-diaspora Jews.
Of course all the other percentages cannot be taken for old ones, neither the Arab percentage, which demonstrates also that was an introgression of various origin (Arabia, Northern Africa, Iberia etc.).
Also for the haplogroups, what I studied above all, the Samaritans demonstrate that a group, isolated for so long, tends to have only one Y and only one mt, and, above all, extinguishing itself.
If you permits, I'd want answering Agamemnon for what he said on Anthrogenica: I didn't think that he were "Italian" for his R-U152 of a grandfather of his, but just for his Y, that, even though J, may be introgressed, in this case recently, after the diaspora, being the same hg. J not originated in Middle East but expanded from the Caucasus and before very likely from WHG.
He denies being Yaniv Erlich (which wasn't an hypothesis of mine but of a friend of mine: I noted only that the haplogroup of Yaniv Erlich, known from Ysearch, and that of Agamemnon (whose data and name weren't ever published, except the final SNP, seemed to me very close). He denies being him, also for that I aslked Daviski to do a plot with Mediterranean Islands, so we could compare his plot and DNA.land.

Davidski said...

You lost me after the first sentence there.

Helgenes50 said...

@ David

About French_East

By using the coordinates, latitude, longitude, they're really in East France, i.e. in Meurthe et Moselle
But the province indicated is not the same, Morbihan is in Brittany, in NW France ???
I sent a message to Lazaridis about this problem, no reply.

French24433 M French French_East France Morbihan 49,047 6,615 Genomic_from_blood Fully Public
French24434 F French French_East France Morbihan 49,047 6,615 Genomic_from_blood Fully Public
French24437 F French French_East France Morbihan 49,047 6,615 Genomic_from_blood Fully Publi

Davidski said...

Yeah, there are clearly some issues with the geographic origins of the new French samples in the Human Origins.

The French_East do look eastern relative to the other French, but what I found was that the French_West were more northern than the French_Northwest, which was strange.

So I removed the French_West and French_Northwest from the K7 spreadsheet. French_East and French_South are still there.

Ryan said...

@Mom

"Historically, Samaritans are said to come from Iraq -- nothing to do with most Lebanese Christians."

No. Historically they're said to come from ancient Israel, and were the largest ethnic group in the area in the late Roman era. They claim to be the descendants of Israelites who didn't get sent into exile by the Assyrians and Babylonians. They have a very similar version of the Torah to that of Jews, but with Mount Gerizim near Nablus as their centre of worship, as opposed to Jerusalem (which was the capital of the southern kingdom of Judea instead).

They seem to have been mostly genetically isolated from the various migrations to the region, whereas Palestinians, while probably mostly descending from a similar stock of ancient Samaritans, Jews and Pagans from ~2,000 years ago, but were more affected by relations with the Bedouin too. From what I understand there are a lot of Muslim Palestinians in Nablus with Samaritan names still though.

@Gioiello - Ashkenazi R1a is R-M582, which is from the Middle East, not Europe. The tree is Z93 > Z94 > Z2122 > M582. Proto-Iranian or some related group.

There's a pretty obvious vector for this too - the earliest mentions of Jerusalem are from letters to an Egyptian Pharaoh by his governor in Jerusalem - a governor with a Hurrian name. There are quite a few Indo-Aryan names among Hurrian/Mittani elite too. Jerusalem was conquered by the Israelites from the preceding "Jebusites" - and if that group had a Hurrian/Aryan elite, that's a pretty obvious source of R1a. Probably via the Zadokites. Christ, Solomon's mother was a Hittite supposedly, and though it's not all that clear that the Biblical Hittites were actual Hittities, it does show Israelite elites intermarrying with locals extensively.

So you're just wrong in looking to that as a marker of European admixture. More fruitful would be looking at mitochondrial DNA.

The Tuscan connection with Ashkenazi Jews is pretty well documented through IDB sharing, which the stuff about "refugium" can't explain. And Davidski's 50%+ Near Eastern estimate for Ashkenazi Jews is pretty much in line with what scientific literature seems to suggest, at least to my eye. Not 30% or less as you suggest.

There's definitely some screwy connections between southern Italy and the Middle East / Near East, but it's not a matter of Jews being south Italian, but rather southern Italy having a strange affinity to the Middle East.

Ryan said...

@David - "You lost me after the first sentence there."

He's saying you have Ashkenazi Jews as too Middle Eastern and not European enough, and that Ashkenazi Jews are really more Sicilian or Cretan or Cypriot and not related to Israeli Arabs or ancient Near Easterners.

Never mind that Sicilians and Cretans and Cypriots all show high affinity to modern populations in the ME too.... :/

M. Myllylä said...

We have no Western Finns in this data set. The selection is essentially weighted eastward from the Finnish average.

Gioiello said...

@ Ryan

The reference paper about "R1a Levites" is the homonymous paper of Behar et al. (2010). Unfortunately has been the same Behar to say at the FTDNA conference of the last year that he was wrong about an origin from Middle East and Old Jews. But it seems that lastly he retracted the retraction and promised a new paper. Thus we'll wait for this possible new paper, without forgetting that the same Behar promised a new paper about mt K1a1b1a, I haven't seen so far.
Of course these scholars may ignore all what I wrote in these last ten years about the argument, but there is also the paper of Costa et al. 2013, etc etc. As to me I think having abundantly demonstrated that my K1a1b1e was born very likely not only in Italy, but just in Tuscany.
About the R1a-M582, unfortunately, also in this case we have a MRCA 1450 years ago, and, if what Ryan says were true, it is difficult to accept that in 1450 years (2900 ya is the age of R-Y2619) of its supposed presence in the Jewish pool no other sample has survived and has had descendance. This is another case like many others in the same situation I spoke about.
Of course I never denied that many "European" hgs may have entered the Jewish pool before the diaspora. We are verifying them, even though it is more and more clear, at least for the R1b haplotypes, that the oldest samples are rooted in Europe with their oldest subclades.
But I don't understand why, if one is a "Jew", it isn't enough to him to declare his faith, and he is searching for a link in the DNA. I understand that for many of them, who have many clear mt and Y of undoubted "European" origin, perhaps a supposed Y may demonstrate that link (say R1a, many J-s or many other hgs.), but their faith should be strong, protected in a fortress, as many cantatas of Bach after Luther (thus the Bible) say, even though to-day we couldn't say that those fortheress are so solid, at missile proof someone could say.

Ryan said...

@Gioiello - No, it's from Rootsi et al. http://www.nature.com/articles/ncomms3928 You are badly misinformed on this subject. Whether you find the author credible or not, the conclusions of the paper are solid, and the position in the phylogenic tree is clear. David blogged about this. It's pretty obviously Indo-Aryan in origin.

You're changing the subject by bringing up mtDNA - I'm the one who first made the point that mtDNA is where the signature of admixture is clearest.

"About the R1a-M582, unfortunately, also in this case we have a MRCA 1450 years ago, and, if what"

That's only for the Ashkenazi specific sequences, and it's for pretty obvious reasons - Ashkenazi Jews as a group are only 1500 years old. The TMRCA for M582 including the non-Jewish Assyrian sequences is 4400 years - which seems rather early to my eye, but isn't that far off from the Marjannu chariot riders showing up as an Indo-Aryan elite among the Hurrians around ~3,500 years ago.

If you're trying to just go by haplogroups though that's not going to give you very accurate results.

Gioiello said...

@ Ryan

I quoted everything by memory, and of course I was wrong in the surname of the first author and the year, but I have that paper, both downloaded and printed in many copies, above all the page with the SNPs, I have studied deeply. I was the first on Anthrogenica to say that the Tuscan from 1KGP was an R-L277. Believe me, I think that very few or no one may compete with me about STRs. Anyway Behar wasn't the first but the second author (Siiri Rootsi, Doron M. Behar) and I have written tons of letters about who are the captains and the soldiers in the papers.
In that paper was Behar the captain, and that he has said that he was wrong is very meaningful. We'll see if he will publish another paper.
The oldest in that haplogroup is an Iberian from 1KGP, and we'll see if some haplotype older than the Ashkenazic one will appear. Certainly it isn't amongst Jews, deeply tested, both in the known results and the hidden ones.
R-CTS6 Z2473/S3408 * S3409 * Z2476/CTS3605/S3414+11 SNPs formed 3400 ybp, TMRCA 2900 ybp info
R-CTS6*
id:HG01617 IBS

Ryan said...

@Gioiello - And? It's pretty conclusive. That's not the European branch of R1a. Just admit it and move on. Sticking to these antedeluvian ideas in the face of contradictory evidence is just not credible. And it seems to be motivated by some dispute with a Jewish commentor on Anthrogenica who you dislike that absolutely 0 people here care about.

Grey said...

Gioiello

Yaghnobi have a mix of R1a and J2 and apparently used to be much more widespread before Turkic expansions so if you think some of J comes from the general vicinity of the Caucasus to Tien Shan region then it wouldn't be surprising imo if some clades of R1a do also.

https://en.wikipedia.org/wiki/Yaghnobi_people

Open Genomes said...

David and all, there is an interesting issue with one of the Arab Christians from Galilee:

One sample, "CArab2" (GS000016885-ASM) from Abu-Snan is R1b-PF6656 (under R1b-U152), and derived for PF6656 and PF6659 but ancestral for the other R-PF6656 equivalents PF6645 and PF6655.

Abu Snan was a Crusader stronghold of the Teutonic Knights:
https://en.wikipedia.org/wiki/Abu_Snan#Crusaders

Of course we know that R1b-U152 was found in a Bell Beaker sample from Osterhofen-Altenmarkt, Germany. It's obvious that the aDNA context of shows that R1b-U152 originated in Europe. R1b-U152 is found in Greece, but this particular sample is ancestral to a clade which so far only includes Sardinians and a few English. However, there are no closer matches to this particular Y haplotype than c. 2500 BCE:

YFull R-PF6656 tree formed 4500 ybp, TMRCA 4500 ybp

We also know that the Sherden (from Sardinia) formed part of the Sea People who invaded Canaan during the Bronze Age Collapse, c. 1206-1180 BCE:
https://en.wikipedia.org/wiki/Sherden

Given the tMRCA at c. 2500 BCE, it's also possible that this Y haplotype arrived with the Sea Peoples. However, we know that I2-M26 is modal among Sardinians. If the Sherden Sea People settled in Canaan, we would also very likely expect to see I2a-M26 in the Levant. We don't.

The Romans of course were present in Roman Era Palestine, and they settled veterans of the Legio X Fretensis in Flavia Neapolis (Nablus), although most of these were probably not Italians but local Levantines.

So the question arises if this particular Galilee Arab Christian sample has Roman Era or more likely Medieval Crusader European admixture. From the various PC plots in all dimensions, (1-2-3, 4-5-6, 7-8-9, 1-4-7) this particular Galilee Arab Christian GS000016885 if anything in some dimensions is slightly closer to Assyrians than the others, but shows no evidence of added European admixture.

Perhaps there are some small segments that are relics of the Medieval European admixture?

Open Genomes said...

As far as Ashkenazi Southern European admixture goes, aren't the "East Sicilians" and "South Italians" a better fit than the Tuscans?

There are Jewish Medieval traditions that indicate that male prisoners from the First Jewish War were settled in the towns of Calabria, Apulia, Naples, and Rome. Archaeological evidence from Jewish catacombs and inscriptions also indicates this to be the case.

While the Northern Lombards expelled the Jews from the Kingdom of the Lombards in 680 and again in 755, upon converting to Catholicism, and the Byzantines expelled the Jews from their domains in 712, the Arian Lombard Duchy Benevento had a treaty with the Byzantine Catepanate of Italy which basically demilitarized the Byzantine towns of Southern Italy while leaving the Lombard Duchy in control of the interior.

The Lombard Duchy of Benevento
The Catepanate of Italy

The Arian Lombards, unlike the Orthodox-Catholic Romans/Byzantines, did not have much of a problem with Jews living in their domains. The Jews of Southern Italy in fact helped the Arian Goths resist the [re]conquest of Italy by the Roman Byzantines under Justinian in 526.

"According to tradition, the settlement of the Jews in southern Italy is connected with the captives brought there by Titus after the sack of Jerusalem (Jerahmeel, in Neubauer, "Medieval Jewish Chron." i. 190; Ahimaaz, "Chronicle," ib. ii. 112, line 4; "Jewish Quart. Rev." iv. 623). If there is any truth in this tradition, then together with Naples, Venosa, Otranto, Taranto, and Oria, Bari must at an early date have become a seat of Jewish influence."
(Rome itself should be included in this list, but Rome was almost depopulated during the Lombard Era.)

Jewish Encyclopedia - Bari
Jewish Encyclopedia - Apulia
Jewish Encyclopedia - Otranto

On the other hand, until the start of the Muslim conquest of Sicily, there seems to have been a hiatus in the Jewish presence in Sicily, because the Byzantines exclusively controlled Sicily and were therefore able to entirely exclude the Jews from residing there.

The Kalonymos rabbinical family from Bari apparently was brought north to Mainz after the siege of Bari in 987 by the Holy Roman Emperor Otto III. Apparently the Jews of Bari were key to negotiating the surrender to the Holy Roman Empire rather than the Antisemitic Byzantines who formerly ruled the Catepanate of [Southern] Italy.

Jewish Encyclopedia - Kalonymos

So I think we need to look for the specific source of Ashkenazi Southern European admixture in the territory of the Duchy of Benevento and the Catepanate of Italy, in other words, mainland Southern Italy, rather than Tuscany or Sicily.

David, can you include your South Italians and East and West Sicilians in this nMonte run, if they aren't included already?

Gioiello said...

@ Ryan

I'll answer you to-morrow. Now it is late here and I am tired, but I never said that R1a-Levites is from Western Europe. Of other haplogroups I said that, and I think having documented. I know very well that the haplotypes downstream Z94 are more Eastern European and Asian than Western European (even though the same Italy, which has a little R1a, has R-M420* and also the oldest R-Z93*), and I have supported many times that R1a was either in the Italian Refugium with R1b1 or in Nortrhern Western Europe, and we'll see where the oldest haplotypes will be found in the aDNA. I said only that there is no proof that R1a Levites is old in the Jewish pool, and we are trying to understand where it came from. We spoke a lot about Khazars or other peoples, not discounting an old introgression pre-diaspora.
To Ted Kandell I would say that I respect his researches and his competence, but that he is trying for his "Jewish" Y too, and that isn't at honour of his open and not racial (at words at least) Weltanschauung.

Open Genomes said...

Rome was an important primary place of Jewish settlement in Italy.

We can see that there was a considerable population of Jews in Rome during the Empire, and perhaps that many of them were young men. (8000 Jewish youths were said to have been expelled to Sardinia by the Emperor Tiberius in 19 CE, but they were allowed to return in 31 CE. The Jewish community of Rome rapidly increased in size after the First (67-70 CE) and Second (132-135 CE) Jewish Wars, partly through the deportation of young male prisoners to Rome as slaves, and partly from refugees arriving in Rome:

Jewish Encyclopedia - Rome - Expelled Under Tiberius.

The Jewish deputation which petitioned for the deposition of the royal house of the Idumeans was joined by 8,000 Jewish residents of Rome. Several Romans adopted Jewish customs, and some, as the rhetor Cilicius of Kalakte, a friend of Dionysius of Halicarnassus, even embraced Judaism (Müller, "Fragmenta Historicorum Græcorum," iii. 331). The reign of Tiberius (until the removal of his minister Sejanus) was fraught with misfortune for the Jews. When the cult of Isis was driven out of Rome (19 C.E.) the Jews also were expelled, because a Roman lady who inclined toward Judaism had been deceived by Jewish swindlers. The synagogues were closed, the vessels burned, and 4,000 Jewish youths were sent upon military service to Sardinia. After the death of Sejanus (31) the emperor allowed the Jews to return.

The emperor Claudius was not unfavorably disposed toward the Roman Jews in the beginning of his reign, but in 49-50, in consequence of dissensions among them regarding the advent of the Messiah, they were forbidden to hold religious services. The leaders in the controversy, and many others of the Jewish citizens, left the city. A considerable number of Roman Jews who had become Christians received the apostle Paul in Puteoli (61) and Rome with due formalities (with regard, however, to Peter's sojourn in Rome, compare Jellinek, "B. H." iii. 60 et seq., and Güdemann, "Gesch." ii. 44 et seq.). Under Nero the Jews of Rome had a comparatively peaceful time, owing to the favorable attitude of the empress Poppæa Sabina; but this was followed by the terrible wars and the conquest of Judea under the emperors Vespasian and Titus. Judaism at Rome was now put on the footing of a privileged religion, instead of its adherents being treated as a separate nation, and the fiscus Judaicus was now levied for the benefit of the temple of Jupiter Capitolinus. A "procurator ad capitularia Judæorum" was empowered to collect this tax, and only those who had abandoned Judaism were exempt from paying it.

Prominent Families.
After the war the Jewish community in Rome increased rapidly; among the prominent Jews resident there at that time, besides Josephus, King Agrippa, and his sister Berenice, are said to have been members of the four families from which the De Rossi, the Degli Adolescentoli, the De Pomis, and the Degli Piatelli families are descended. The pressure of taxation rendered the condition of the Jews very unfavorable under Vespasian and Titus; and it grew worse through the increasing number of those who abandoned, or professed to abandon, Judaism to escape the payment of taxes. These defections at last became so numerous that the emperor Domitian, in the beginning of the tenth decade, found it necessary to adopt stringent measures. Every suspect was examined individually, and if the suspicions entertained were confirmed he was severely punished (Dio Cassius, lxxvii. 2). Among those sentenced to death or banishment for various reasons were the emperor's nephew Flavius Clemens and his wife Domitilla. Rabbis Gamaliel, Joshua, Eleazar, and Akiba preached in the synagogues in Rome during their brief stay, and engaged in disputes with the Judæo-Christians.

The Jews do not appear to have been affected by the severe decrees issued by Hadrian after the Jewish uprising.

Chad Rohlfsen said...

Gioiello,

Give it a rest. I'm sure I'm not the only one tired of your Italo-centric babble. Jews are 50% Near Eastern, R1b didn't originate in Italy. Neither did J1, G2, I1, I2, L1, or any of the others you spout off about. Nor are Natufians "black", as you once said. Do us all a favor and drop the STR shit and read everything after 2010. It'll save us all from having to constantly correct you.

Open Genomes said...

Interestingly, the proportions Ashkenazi from the nMonte run are almost identical to those found in Xue et al.(2016):

"We found that the best fit to the AJ data using a 4-way model (Middle-Eastern,
Southern EU, Eastern EU, and Western EU with proportions 50:35:12:3 (%), respectively) was obtained."

nMonte
49.25 Samaritan + Galilee Arab Chrisitan
30.45 Italian Tuscan
11.75 Polish

vs. Xue (2016):
50% Near Eastern
35% Southern European
12% Eastern European
3% Western European

However, these calculations leave out a non-trivial contribution from the Berbers of North Africa, which we see in the Y-DNA (G2a2b2a1a1c1a1b1a-Z6523 and E-M183) and mtDNA L2a1l2a A143G T14180C (with the closest match among the Pana of Burkina Faso) both of which are found in substantial percentages among Ashkenazi Jews.

The History of African Gene Flow into Southern Europeans, Levantines, and Jews - Moorjani et al. (2011)

A striking finding from our study is the consistent detection of 3–5% sub-Saharan African ancestry in the 8 diverse Jewish groups we studied, Ashkenazis (from northern Europe), Sephardis (from Italy, Turkey and Greece), and Mizrahis (from Syria, Iran and Iraq). This pattern has not been detected in previous analyses of mitochondrial DNA and Y chromosome data [7], and although it can be seen when re-examining published results of STRUCTURE-like analyses of autosomal data, it was not highlighted in those studies, or shown to unambiguously reflect sub-Saharan African admixture [15], [38]. We estimate that the average date of the mixture of 72 generations (∼2,000 years assuming 29 years per generation [30]) is older than that in Southern Europeans or other Levantines. The point estimates over all 8 populations are between 1,600–3,400 years ago, but with largely overlapping confidence intervals.

FrankN said...

@Ryan: To your - otherwise very interesting - interpretation of Samaritian - Hurrian - Hittite interaction, please note that the name Jerusalem itself is clearly Semitic, meaning "Foundation/ House of Peace" (c.f. "Dar-es-Salam" with the same etymology, Hebr. "Shalom" etc.).

@OG: You shouldn't trust "official" YFull TMRCAs, especially when it comes to R-P312 and subclades, including U152. Their calculated TMRCA for R-PF6656 (check the link you provided) is 5 kya, with a calculated age of YF0916 under R-PF6656 of 5.5 ky. Both figures have been revised downwards, as upstream R-Z192 has been estimated at 4.5 kya. That estimation is also arbitrary, yFull's calculation yields a TMRCA of 4.8 kya for R-Z192. The oldest subclade under U152, Y22162, has, per yFull's own calculation, a foundation age of 5.5 kya, corrected downmwards to 4.5 kya in order to bring it line with the estimation for U152 itself, which in turn has been derived from R-P312 upstream, as also the yFull calculation for U152 yields a higher age.
Essentially, the phylogeny below R-P312 is poorly resolved. yFull has it simultaneously splitting into 30(!) subclades - technically impossible, thus meaning that we still lack a number of SNPs to define the phylogeny in appropriate detail. Things like that happen, of course, and I assume a number of people are working on better resolving the phylogeny. Unfortunately, however, instead of treating each of those 30 P312 subclades as potentially having formed at different points in time, yFull just averages acrosss all of them, thereby lumping together, e.g., DF19 (calculated age 5.6 ky) with YF06626 (calculated age 1.9 ky). Moreover, they quite arbitrarily correct formation times up- or downwards in order to create the impression of a consistent phylogenic history. In addition to the a/m "corrections", e.g. they have "corrected" the formation of P312 upwards to bring it line with L21, estimated at an age of 4.5 ky. However, their calculated age of L21 is in fact 5.2 ky. It is complete mess!
You can check out for yourself by clicking the "info" boxes for the yFull formation/TMRCA estimates (pay attention to the final notes following the list of individual samples underlying their caluclations):
https://www.yfull.com/tree/R-P312/

Open Genomes said...

About any so-called "Jewish" Y-DNA:
G-Y12975 is tied for the second largest "Ashkenazi" Y-DNA clade (with J-L556). As far as "Ashkenazi" G-Y12975 (under G2b1-M377 goes, there are only two instances of this clade among Southern Europeans who are not known to be of Jewish ancestry:
• 1 Sicilian from Palermo, who appears to be within one of the two "Ashkenazi" branches of G-Y12975
• A cluster of related haplotypes found among "Mestizos [Admixed]" from Merida Yucatan in YHRD and a Mexican-American with a Mayan surname. These appear to be in the sub-subclade G-Y16169 of G-Y12975. Since all three haplotypes differ slightly at 17 STRs, it would appear that the common ancestor lived hundreds of years ago, and was probably among the Spanish Conquistador founders of Merida in 1542.

The tMRCA of G-Y12297 which includes the "Ashkenazi" Y-DNA G-Y12975 clade and the G-M283 Pathan clade is 5,500 years, and the tMRCA of G-Y12975 alone is c. 1234 years:

YFull G-Y12975 tree

The three other known South Italians are quite distant from G-Y12975, and on the median-joining networks either cluster with the Pathan G-M283s (under G-M377) or are more distant. Two South Italian G-M377s from Salerno and Bergamo appear to be in a clade with Rushani and Shugni Pamiri Tajiks and a Han Chinese from Liaoning Manchuria, rather than the Pathan G-M377 clade. One remaining South Italian from Naples is much further distant than even the Levantine G-M377s with a tMRCA of 9,000 ybp, and and there is an equally distant Armenian / Dagestani G-M377 clade. The tMRCA of these with the remainder of G-M377 may be as far back as 14,700 ybp.

Understand that 5,500 ypb(3,500 BCE) predates the earliest Yamnaya R-L23* samples and is quite close to the tMRCA of all of R-L51 at 5,700 ybp. The earlier South Syrian / Lebanese Christian G-M377 clade has a tMRCA with the "Pathan-Ashkenazi" clade at 9,000 ybp. Clearly,l this branch of G-M377 dates to the arrival of the PPNB cattle herders from the north to South Syria (Tell Aswad and Tell Ramad).

Since earliest branches of R-Z2108 are found among Iraqi Arabs, Gujaratis, and Pathans, we might as well also say that R-Z2108 originated in Iraq or South Asia. Of course we know that the earlier R-Z2103 branches were found in Yamnaya samples from Samara. However, we have never seen any ancient Y-DNA haplogroup G samples from the steppes. The tMRCA of all of R-M417 is precisely the same as G-Y12297 at 5,500 ybp. Since no G2b1-M377 was found in the steppes or Europe, unlike R1b-Z2103 or R1a-M417, we cannot say that G-Y12297 arrived in Eastern Afghanistan, the NWFP and Punjab from the steppes, or anywhere near Europe.

Given that the closest outgroup matches to the "Ashkenazi" G-Y12975 clade are in Eastern and Northen Afghanistan and Punjab (but not further north) and not anywhere near Europe, we more plausibly say that this Y clade originates with the Indus Valley Culture (IVC) rather than anywhere west of that. What we can say for sure is that there's no evidence at all of a European or Mediterranean origin, but equally, no evidence of a Levantine origin either.

WC1 from Wezmeh Cave in the western part of the Central Zagros of Iran (9,500-9,100 BP) was G2b2a-Z8022 (vs. G2b1-M377), but the tMRCA of G2b-M3115 itself is 20,000 years. That's at least 5,000 years before humans entered the Americas from Beringia.

I hope everyone can see that whether such a Y clade that has a tMRCA solely with Pathans is "Jewish" at 5,500 BP (3,500 BCE) is utterly meaningless. The precise origins of G-Y12975 at this point are totally unknown. It's a fascinating question, but given that the tMRCA is right at the start of the Early Bronze Age, there is nothing "Jewish" per se about it.

a said...

Eurogenes request.

"Focusing on "European" population genetics and modern physical anthropology."

Many ydna projects have very high ratios of either "R1a" and or "R1b" as members, or both. Since we have many ancient "R1b" and "R1a" samples and the prospect of many many more in the next few years. It is beginning to paint a picture of various paternal R1a/b groups as autochthonous/homogeneous in various regions[Yamnaya/Corded Ware etc.....]. Can Eurogenes make an all inclusive "R1a/b" autosomal calculator[is it possible?], in order to avoid confusion[and offending anyone of a different ancestry], in distinguishing those with ancient European heritage, Yamnaya/Corded Ware etc.....

Davidski said...

I have no idea what you're saying there.

Richard Rocca said...

@Frank said "It is complete mess!"

I checked the U152+ German Bell Beaker sample against SNPs from 300 samples and no matches. I did the same thing with the German Middle Bronze Age Tumulus Culture (~1610-1450 BC) U152+L2+ sample and again, no matches. If these subclades of P312 were much older than say 2,800 BC, we'd find downstream SNPs in these samples, but we don't.

a said...


@Davidski said...

"I have no idea what you're saying there."
Sorry I don't know if this makes sense so I'll give it another go.
Since R1a/b are native to Europe, and have been for so long time, aren't most R1a/b going to be a simple mix/combination of Villabruna WHG-Karelian EHG and some Armenian/and or Els Trocs. Is it possible to construct a basic Eurogenes Admixture Proportions calculator using only R1a/b as reference samples, eliminating any speculation in ancestry?

Gioiello said...

"Gioiello,
Give it a rest. I'm sure I'm not the only one tired of your Italo-centric babble. Jews are 50% Near Eastern, R1b didn't originate in Italy. Neither did J1, G2, I1, I2, L1, or any of the others you spout off about. Nor are Natufians "black", as you once said. Do us all a favor and drop the STR shit and read everything after 2010. It'll save us all from having to constantly correct you".
(Chad Rohlfsen)

Is it that the thought of the owner of the blog too? If so, it is enough to ask. Anyway is "Chad Rohlfsen" in line with the new rules of the blog?

FrankN said...

As concerns the history of Ashkenazi:

1. I wouldn’t rule out early medieval Eastern European, i.e. Chasarian roots. IIRC, Jews of Lithuanian descent show some genetic pointers to the Caucasus. Statistical analysis, comparing Ashkenazi with, e.g. Georgian or Iranian Jews (many of which have been deported from Georgia under Persian rule during the 16th/17th century) might be interesting.

2. Nevertheless, it seems that most Ashkenazi had their origin in Germany and moved to East Europe with the high/late medieval "Ostsiedlung". Aside from traditional Jewish centres along the Rhine, Jewish expellation from Erfurt (by that time one of Germany's top four cities population-wise) in 1349, or from Mecklenburg in 1492, e.g., could have contributed. The Laz 2016 "German" sample, exclusively from Leipzig, might capture a bit of genetic contribution to Erfurt Jews. Otherwise, Germany is essentially unsampled, including the Rhineland, or Mecklenburg, so we can't tell whether Ashkenazi collected genetic contributions there. [Older, uniparental studies point at a considerable "Roman" element within Rhinish populations, while Mecklenburg, unsurprisingly, seems to entail a substantial W. Slavic (Polish-like) component.]

3. Ancient Jewish communities along the Rhine seem in part to go back to Roman times. This concerns especially Cologne and Trier, which display archeological indications for such a continuity, but possibly also the sizeable early medieval Jewish communities of Worms and Mainz. Especially the latter, however, might have had little incentive to partake in the "Ostsiedlung", as nearby Frankfurt/ Main provided a prosperous "safe haven" to Jews until 1938

4. As already pointed out by OG, Arian Germanics were much more tolerant towards Jews than their Roman Catholic counterparts. This concerns not only Lombards, but - for the sizeable early medieval Jewish communities of Tolouse, Narbonne, Montpellier etc. - first and foremost Visigoths. In fact, Visigoth rule seems to have been strongly based on Jewish political/financial support. Frankish (Neustrian) conquest of Visigoth lands north of the Pyrenees resulted in substantial Jewish resettlement, the directions of which are unclear. Catalonia and Lombard-controlled parts of Italy look most likely, but East Frankish (Austrian) lands along the Rhine might also have served as refugia. Whatever the case - it might be a good idea to include DNA from SW France into models for Ashkenazy ancestry, especially as concerns the "Atlantic" component often found therein.

a said...

@Blogger Open Genomes said
"Since earliest branches of R-Z2108 are found among Iraqi Arabs, Gujaratis, and Pathans, we might as well also say that R-Z2108 originated in Iraq or South Asia. Of course we know that the earlier R-Z2103 branches were found in Yamnaya samples from Samara."

OPG I have always admired your comments and contributions[dedication and effort]. There is a problem with the logic you are using in the above example[Z2108-Z2106* Tabassaran[EHG rich line]. This becomes evident when studying the R1b-L584 line. So as to not embarrass two distinct ethnoreligious groups, who both practice endogamy, study the R1b-L584* from top to bottom in snp formation. It will become evident that 3 [possibly 4] completely different groups[endogamy] carry this branch.

FrankN said...

5. A well evidenced high medieval trajectory leads from South Italy to SW Germany. The Staufen (Ghibelline) rule over S. Italy was characterised by a high degree of religious tolerance - too much for the Papacy that consequently banned Emperor Frederic II. The demise of the House of Staufen, and subsequent falling of the "Two Sicilies" to the House of Anjou, and afterwards the Crown of Aragon, resulted in a series of progroms and expellations that had S. Italian Jewish communities resettle to N. Italy, and former Staufen lands in Suebia and Switzerland.
However, it didn't take long for Jews to be also expelled from SW Germany/ Switzerland - more for political than religious reasons: The Emperors of the HRE used their traditional right to tax Jews as a pretense to raise tax revenue from cities, holding them responsible for collecting and transferring taxes from Jews. To defend themselves against such claims, many cities (e.g. Zürich, with a sizeable Jewish ciommunity since the mid-13th century, in 1436) officially expelled Jews, thereby sending some of them just to the nearest suburb, but causing many others to move on towards Eastern Europe.
Here, I wouldn't expect much "Palatinian" "E. France" or "Swiss" genetic contribution. We are probably not talking about more than 150 years (5-6 generations) stop-over. But S. Italy / Sicily, where many of the migrants might have originated, and which used to have a far longer Jewish tradition, is a different issue.

Gioiello said...

@ Richard Rocca
"If these subclades of P312 were much older than say 2,800 BC, we'd find downstream SNPs in these samples, but we don't".

Of course that isn't true. You demonstrate that this line hasn't so far descendants, is a "dead line", but not its age, except the fact that it has in common with the survived lines the SNPs U152, but you don't know when it was born. And also the age, that of YFull and any other, is questionable (what I am doing from so long),and even though you were able to count all the private SNPs, seen that their number changes in each sample, you wouldn't be able to do any calculation.
That what you say is a demonstration that U152 isn't older that 4800 years is only a presupposition. The question will be resolved from other aDNA, anyway it seems that Germany is at this point a peripheral place of R-U152 and not its source, as I think having demonstrated for the Samara samples (abundantly).

Open Genomes said...

I think the problem with some people's understanding of this is that northern South Asians, Iranians, and Near Easterners don't fit into the so-called "Racial" categories of the current political climate. (People of the Caucasus don't fit, either, which is ironic because the term "Caucasian Race" is still quite widely used.) These so-called "Racial" categories are social constructs, and biologically meaningless. I don't see them clearly on any PCA and the ancient population isolates like WHG, EHG, CHG, Anatolian Neolithic, Levantine/Natufians, etc. certainly don't fit into any of these so-called "Racial" categories either. Many people (partcularly the "HBDers" can seem to get their minds around the idea that one group from what they perceive to be a certain "Race" is a mixture of or more closely related to groups from another very different "Race".

Early Europeans were part of a WHG isolate that vanished in its pure form millennia ago. Early European farmers were more similar to Levantines and Turks. If Arabs and Turks are not "White" and Somalis have some substantial "LBK-like" admixture, then does that mean that the Turks and Kurds in Germany today and the Somalis in Italy and Sweden are more like the pre-Steppe Europeans than Europeans today? And if the Andronovo of Central Asia and the Afansievo people of South Siberia are closer to the present-day Europeans than the First Farmers, what does that make them?

It certainly makes many people uncomfortable.

Hence, the endless discussions about everyone's "favorite pet / crank theory" about the origins of this or that group (in fact, mostly Y-DNA clades) everywhere but where they actually originated. These endless ruminations (and accusations) amounts of words but not only add nothing to our knowledge about the actual origins of these Y-clades and their members today.

Open Genomes said...

These are the facts not "wishful thinking" and "Racial identity-based speculation":
R1b-M343 (xM269) Y-DNA tree. v 4. July 29, 2016

R1b-M269 (P312- U106-) Phylogenetic Tree. v 30. R1b Basal Subclades FTDNA project, September 12, 2016

In fact, Villabruna from Italy, Epigravettian, 12 200-11 800 BCE is R1b-L754, but so is I0122 from Khvalynsk II, Volga River, Samara (Russia), 5200-4000 BCE. However, I1635, Kalavan (Armenia), Armenia_EBA (Kura-Araxes), 2619-2465 calBCE is R-L389*, and I0124, EHG, Lebyazhinka IV (Russia), 5640-5555 calBCE is R1b1a-M478* under R1b-P297. Zhong et al. (2010) found an R1b-M478 (R-M73) in Japan, and this haplotype as certainly not Portuguese or Dutch, but rather, Central Asian in origin.

R1b1c-V88 is found all in Cameroon, Nigeria (4% of Yoruba) and even Zaire. "R1b" is just as "African" as it is "European". Does anyone think that I0410 from Els Trocs, Spain, the Iberian Early Neolithic 5295-5066 calBCE is the ancestor of these West and Central Africans? Given the distribution of R1b1c-V88 and the fact that this was an Early Neolithic sample, it seems just as likely or more so that R1b1c-V88 arrived in Iberia via North Africa. R1b1c-V88 just as "African" as "European" R1b-L51, or even more so.

R-CTS1010/V3949 under R1b-L389* is found among a few Mediterranean people (including Hispanics) but also among Armenians and Georgians, as well as I1635 from Kalavan Armenia,the Armenian Early Bronze Age (Kura-Araxes), 2619-2465 calBCE. It's pretty much guaranteed that Kura-Araxes I1635 did not come from Europe. The earliest branch of R1b-M415, R-PH155, is found in the Ganges valley, Tajikstan, Uyghurs from China, and Bahrain. Villabruna notwithstanding, neither the very early nor the late branches of R1b are European. Obviously, Mal'ta from Siberia was R*, and P1 is found among the Agta (Negrito people) of the Philippines.

So R1b is not ... "White" ("Caucasian" doesn't quite work here anymore), despite what racist websites say?

Everyone needs to put aside their preconceived notions and prejudices about who they "could and could not be" related to. The data is what it is, and that's called "Science". No one is who they think they are, or who they think they "should be". The sooner we stop obsessing about things like "Race" and "religious mythology" that have no basis in the data, the sooner we'll start figuring out what the data really shows. I think most people here can agree on this, even if most people "out there" don't.

And if this makes anyone nervous, uncomfortable, or even angry .. just ask yourself WHY and be honest with yourself.

Ariel said...

Open Genomes

You revaled your lunacy there, r1b is as African as European? ...my goodness.

Grey said...

Open Genomes

"Does anyone think that I0410 from Els Trocs, Spain, the Iberian Early Neolithic 5295-5066 calBCE is the ancestor of these West and Central Africans?"

personally i'd guess at the iberian and african samples having ancestors from the same original source

going purely by the distribution map

http://cdn.eupedia.com/images/content/Haplogroup_R1b_World.png

the African R1b looks to me like some people went down the Nile valley and traversed the Sahel to West Africa.

(if so there should be ancient R1b in Egypt that disappeared with the Arab (or some other) conquest nb the controversy over Egypt hiding tut's ydna)

Why might people travel all that way?

Gold.

(same reason they might end up in Iberia)

Open Genomes said...

@a - Thank you. :)

Please let me say that No DNA result is "embarassing".

(a, the following is not referring to what you personally said, but the general attitudes that are current that make people want to "tiptoe" around DNA results for "fear of giving offense".)

I'm addressing this in depth, of course, so as to try to put to rest all this endless nonsense and just discuss the actual data, without constantly having to fend off these distractions.

"It's just SNPs." Just a base that mutated ... it is what it is. There's no need to say "Here's this group" (say, the Cossacks, or Spaniards) and "below" here's this "endogamous" group ("the Jews"). If anyone thinks that DNA results are somehow an "insult", or have "political implications".

How can DNA have "political implications", when we're talking about this remote in time, say before the Angles and Saxons invaded Britain, and more recent tragic events, like the Holocaust, which took place without any DNA testing?

I wasn't suggesting that R-Z2108 was actually from northwest South Asia. Of course, it would be from somewhere close to where R-Z2105* was found in the Yamnaya Culture. The point is that using these criteria of finding odd samples in this or that place (say, in Europe) we could make a better (but likely erroneous) argument that the the same clade is found in someplace "very exotic" which would make that person's "favorite" Y-DNA clade originate in a region that they really don't want to be associated with (because of concepts like "Race" or present-day "religion").

I think we can say that given R-Y16852 under R-L584 was found in F38 from Tepe Hasanlu from Iron Age Armenia, it was probably found among the Hurrians or even Assyrians, although it almost certainly originated further north.

I of course see that there is a Cossack who shares a SNP with F38 who is parallel to the mostly "Ashkenazi" R-Y11410 clade. The tMRCA of R-Y11410 is 1100 ybp (900 CE). Notice that there's a Spaniard (Mexican) just above the "Ashkenazi" portion of R-Y11410, which is R-FGC14600, with a tMRCA at 650 ybp.

YFull R-Y16852 tree (including F38) formed 4700 ybp, TMRCA 3900 ybp (1900 BCE)

R1b-M269 (P312- U106-) Phylogenetic Tree. v 30.

There's something I think almost none of you know, and it isn't just "TMI" or irrelevant. It's more than you dreamed or would ever want to know about R-L584, but somehow, it has to be discussed here. You'll see why:

R-Y11410 includes the Ashkenazi "Rabbinical" Spira family. Yes, the early Hasidim claimed that this family were the "pure Israelite (non-Cohen, non-Levite) lineage" because this is (falsely?) claimed to include the Rebbe Elimelech of Lizhensk (1717-1787), and even more importantly, the early Lurianic Kabbalist popularizer Natan Nata Spira (1585-1633), Chief Rabbi of Cracow, and his great-grandson Rabbi Jonathan Eybeschuetz (1690-1764), rabbi in Prague and Chief Rabbi of Hamburg. It seems that Eybeschuetz engaged in a bit "more" than Kabbalah, and he may have been the secret leader of the Messianic Sabbatean community in Europe.

Open Genomes said...

In fact, astounding (astoundingly false that is) claims have been made that the Spira family "Descends from King David". Such fairy tales were used by self-appointed "Tzaddikim" ("Saints") to set up hereditary "dynasties" which would demand money and resources from their poverty-stricken Eastern European Jewish followers. Traditional Rabbinic Judaism (i.e. Elijah of Vilna, the Vilna Gaon) forbade discussions of "Yichus" ("pedigree" or "lineage") because such ideas went counter to the concept of a "scholarly meritocracy" among rabbis and judges, who were the administrative and judicial elite among pre-Modern European Jewry.

I'm bringing this up for a reason:
Right now, we have an effort to use DNA testing to "validate" such claims. (Convicted and jailed child pornographer) Dr. Neil Rosenstein has been writing an update to his trash "The Lurie Legacy: A House of Davidic Descent" with the help of DNA testing. This is the sort of ridiculous garbage that he's been promoting:
From King David to Baron David

The real aim here is not exactly what you see. It's to "validate" the "Davidic" claims of various "Hasidic Dynasties" in Israel and the US.

How ridiculous is it that claims of descent from ancient mythological or near-mythological figures in religious scriptures have a place in what is called "Genetic Genealogy", and these ideas are promoted by companies that sell DNA tests?
Yes, sell, and lots of money changes hands for so-called "tests" and "services" that have no business being sold.

Would anyone like to publicly and not anonymously state that one or another Y clade "cannot be the lineage of the Successors to the Prophet Muhammad"?
(I guarantee you that one or another group will be seriously "deeply offended" by your statement, and a certain proclaimed "Caliph" is not in the lineage you may think he "should" belong to. Sometimes, these things go a little bit beyond just "embarrassing" or even "being offensive".

Of course, that never stopped certain companies and their "co-owner/academics" (like one who is a "Senior Researcher" in the Estonian Biocentre and collected many of the EGDP samples - how appropriate!) from making these claims:
DNA could illuminate Islam's lineage
King David's Genes

So it's worse than you thought. You can see that there are "other" unscientific motivations involved here, aside from just "Race". Yes, various religious beliefs also interfere in scientific inquiry, and here I'm giving an example that probably none of you have ever heard of but I know about, yet we could bring up many other examples from various different religious beliefs. (Not just "Abrahamic Faiths" - ever hear of the "Divine Y Chromosome" of the Emperor of Japan? I bet plenty of people here can give "Hindutva" or even "Confucian" examples, too.)

Do we get a nod from homophobic "White Gods of Chachapoya"? Funny how we seem to tolerate that stuff too, don't we?

Open Genomes said...

The point is, "It's not Science." We shouldn't even be talking about this here.

Doesn't every know that Racism and Religon can KILL?

Why do I look at comments on online newspaper artcles and news stories about "Race" and "Politics" and see things like "Yamnaya" mentioned without even an explanation and we see "Alt-Right" and Neo-Nazi sites where people discuss what it memeans that someone has "1.6% Sub-Saharan" in the autosomal DNA?

For people who are obsessed about "Race" and Religion, the things we're discussing here, "mere mutations", a change in a DNA molecule from one base to another, govern all of human history and how the world works. It's even a matter of life and death to them.

We here have a responsibility not only to analyze and discuss these things, but to also make sure that everyone reading understands what it means and what it doesn't mean. There's too much misuse and misinterpretation about all this, and just ignoring it doesn't help the problem, it actually makes it worse. It's up to us to do something about it.

No one has to be bullied or embarrassed into not discussing the actual results of the data, whether it upsets people's religious beliefs - or there deeply held notions about "Race".

And again, it often seems that much of the discussion surrounding population genetics and ancient DNA results "steps on someone's toes" regarding "Race and Religion" and so much energy is wasted on dealing with this nonsense.

There's an easy way to put a stop to it: Deal with it head on, and expose it. Show everyone what it really is. Don't just sit "sit back and take it".

It's a bit like a "debate" where someone is asked some basic detailed questions, and can't even begin to think of an answer. Everybody sees it for what it is, has a laugh, and gets back to real business.

So where do all of you stand? Remember: "Silence is consent."

Ryan said...

Re: V-88 - I think the simplest explanation is that whatever population movement that brought R1b to Italy with Villabruna (probably from the steppe) also dispersed it in the Near East, North Africa and possibly Iberia at various frequencies, and the R1b populations in North Africa were in the right place at the right time to take advantage of the improved climate conditions to expand massively into the Sahara. Is any further explanation really required?

I think Italy could be a pretty plausible vector for this too, with R1b-rich groups winding up in the Savana lands present in parts of Tunisia at the time. If it was further east, I'd think R1b would have been a more common lineage in the early expansion of agriculture into Europe and more common in Semitic groups, no?

Davidski said...

Here's my take on things.

I don't think we should interpret which scientific data and (legitimate) theories might be socially or politically harmful.

A lot of people try and do this, and it's a horrible mistake, because no one has any idea what might happen in 50, 20 or even 10 years from now, and thus how this information might be taken advantage of.

If people want to kill each other, which is usually because of competition for resources with racial, religious and/or political undertones, then they will find a way to do it and to dehumanize each other to make it easier to stab, shoot and hang each other.

This has happened before and it'll happen again, and the only way to minimize the damage is to make sure everyone has all the facts so that it's more difficult to take advantage of and twist reality.

Open Genomes said...

@Rob

Yes, I'm trying to be "ironic" of course.

1. It does seem to appear that R1b-V88 branched off from a European parent.
R1b-M343 (xM269) Y-DNA tree. v 4. July 29, 2016

The basal R1b-V88 branch R1b-V35 appears to be "Sardinian" ... except that a 17th century African slave, STM1, from the Caribbean Island of St. Martin was R1b-V35:

Genome-wide ancestry of 17th-century enslaved Africans from the Caribbean - Schroeder et al. (2014)

Within R1b, the STM1 lineage was ancestral for all R1b1a-M269 SNPs and derived for 4 of the 5 R1b1c-V88 SNPs for which sequencing data were available. The one anomalous SNP, at hg19 coordinate 18,685,985, had been assigned to this clade on the basis of just two observations amongst 29 individuals, thus rendering the SNP itself suspect.
Therefore, we could definitively classify STM1 as a descendant of R1b1c-V88, despite the fact that no sequencing data were available for the V88 SNP itself.

Downstream of R1b1c-V88, there are two main subgroups in (42). A cluster of 18 individuals represents a lineage we refer to as “R1b1c-V88*,” and the remaining 11 represent R1b1c3-V35, which ISOGG currently labels “R1b1c2.” We observed 7 ancestral alleles amongst 7 R1b1cV88* sites for which STM1 had data, and we observed 5 derived and 9 ancestral alleles on the R1b1c3-V35 branch. Upon diverging 5/14 of the way down this branch, the STM1 lineage parts company with fully-sequenced Y chromosomes of the current literature.

Divergence time of the STM1 Y-chromosome lineage

Francalacci et al. (42) report a cluster of 29 R1b1c-V88 lineages from Sardinia. Though the terminal branch lengths from this study must be viewed with caution due the low-pass sequencing approach, the internal branches had high effective coverage due to the superposition of multiple sequences. Consequently, we can regard internal branch lengths with confidence. Specifically, they report 30 SNPs on the R1b-M343 branch shared by R1b1a-M269 and R1b1c-V88 (branch 1), 51 exclusive to the 29 R1b1c-V88 lineages (branch 5), and 62 SNPs on the branch shared by 11 R1b1c3-V35 lineages (branches 6 and 7). Based on the subset of 14 sites for which we have
sequencing data, we infer that 5/14 of these 62 (~ 22.1) were carried by STM1 (branch 6).
Consequently, approximately 103.1 (30 + 51 + 22.1) SNPs accumulated between the emergence of R1b and the time when the STM1 lineage diverged from R1b1c3-V35. Because this study was based on 8.97 Mb of sequence, whereas that of Underhill et al. (37) analyzed 10.35 Mb, we must scale the mutation period by a factor of 1.154. Thus, we estimate that this interval corresponds to 14.5 ky (103.1 SNPs · 1.154 · 122 years/SNP). Consequently, we conclude that it was approximately 8.5 kya that the Y-chromosome lineage carried by STM1 diverged from that carried by the 11 Sardinians.

Samuel Andrews said...

@Open Genomes,
"I think the problem with some people's understanding of this is that northern South Asians, Iranians, and Near Easterners don't fit into the so-called "Racial" categories of the current political climate."

You're falling for the lies that, 1. That because racial concepts don't express human diversity perfectly they're completly false and 2.that racial concepts are lies that were created for political reasons. The motivation for you and others to believe this isn't facts but emotion and an inappropriate strategy to end racism(which comes from good intentions). This strategy is to say race doesn't exist and anyone who acknowledges it in science has a racist agenda. We can't have this believe in science, especially genetics, because it's false and judgmental(assumed anyone who acknowledges race is a racist).

Our American or Western Culture's concept of race derives from THE FACT that people from different parts of the world look different. These concepts originated during the European age of exploration. People used names; white, black, red; to express the REALITY of human phenotype diversity which is a reflection of the REALITY of human genetic diversity. As early as the 1600s in what would become the United States, you see the terms Negro, White, and Indian. You can distinguish Africans, Asians, and Europeans from each other with DNA not just phenotype, and that sense RACE DOES EXIST. Our culture's racial concepts can't explain human genetic diversity perfectly because colonial Europeans weren't geneticist with access to DNA. Don't put them under undeserving scrutiny. This doesn't mean race doesn't exist.

Our racial concepts aren't political they're an expression of REALITY. Politics have been applied to this REALITY. Our racial concepts have often put Whites/Europeans at the top of racial hierarchies for an obvious reason; during the age of exploration, the time our racial concepts originated, Europeans were subjugating non-Europeans all over the world.

It doesn't matter whether or not the age of exploration occurred, if Africans, Europeans, Amerindians, Asians, etc. encountered each other for different reasons we'd have the same racial concepts because those concepts are an expression of REALITY.

Open Genomes said...

The tMRCA that YFull gives for R1b-V88 (including R1b-M18 and R1b-V35 samples from Sardinia) is 10,200 ybp, and the tMRCA for R-Y7777, the parent of R-V35, is 7,600 ybp, which predates I0410 from Els Trocs Spain, Iberian Early Neolithic, 5295-5066 calBCE, in R-Y7777*:
R-V88 YFull Tree

R1b-M18, which is the most basal R1b-V88 branch, is not only found in Sardinia, but also in Lebanon. There is also an R1b-V88 found among Greek Cypriots. Is this a back migration from Europe, or is this a basal Near Eastern branch that spread to Cyprus and Sardinia with the Phoenicians?

There is also an R1b-V88* (not tested for M18) Armenian and a Georgian.

So STM1, who is clearly West African with no European admixture, is also basal to the "Sardinian" R1b-V35 subclade.

Here, in Cruciani et al. (2010), we can see that there are many R1b-V88* (x R-M18,R-V8,R-V35*,R-V7,R-V69) haplotypes among Berbers and Chadic speakers:
Frequencies (%) of Y chromosome R1b subhaplogroups in Africa and Eurasia - Cruciani et al. (2010)

It's not clear if these are R-Y7771 which is ancestral to R-V69 and is found in West and Central Africa and among Berbers.

It's not impossible that the West African STM1 descends from an Early European Farmer, but there's no evidence for R1b-V88 earlier than c. 7,300 ybp (5,300 BCE).
We have to consider that R-Y7771 under R-Y8447, which is parallel to R-V35, is not found in Europeans, and in fact R-Y8447, the parent of R-Y7771 is found in Saudi Arabia, and clade with Ashkenazi Jews and a Peruvian (with a tMRCA of 900 ybp).
The Saudi of course could be of African descent, and while the Ashkenazi Jews would appear to originate in Spain, they have a tMRCA of 3600 ybp (1600 BCE) with the Saudi. The Ashkenazi Jews in R-FGC20980 could ultimately be of North African origin. It seems much less likely that the Saudi is of European origin (from Spain) than he is from a common North African origin with the Ashkenazi Jews and the Peruvian.

So while there are basal R1b-V88s in Europe, there are equally basal R1b-V88s in Africa and the Near East. Given the different basal "pre-Sardinian" R-V35 and R-Y7771 haplotypes from Africa, and the presence of R-M18 both in the Near East (Lebanon) and Sardinia, if the origin of R-V88 were in Europe, we would have to propose an origin of the African Afro-Asiatic Chadic and Berber speakers in Europe, and the back-migration of R1b-M18 from Sardinia to Lebanon, rather than the other way around.

The most parsimonious explanation is that R-V88 spread across the Mediterranean with the First Farmers after about 7,600 BCE, while spreading across the Sahel and ultimately to Northern Cameroon with the Chadic speakers, and across North Africa with the Berbers.

The most basic questions anyone has to ask here is:
Why are there no R1b-V88s in Europe before Els Trocs, in the Early Neolithic?
and:
Why are basal R1b-V88 clades spread out in parallel in different continents, Europe, Africa, and West Eurasia?

Samuel Andrews said...

Open Genomes,

Sorry David for off-topic discussion. I think it's important Open Genomes understands how he's wrong.

"I think the problem with some people's understanding of this is that northern South Asians, Iranians, and Near Easterners don't fit into the so-called "Racial" categories of the current political climate. (People of the Caucasus don't fit, either, which is ironic because the term "Caucasian Race" is still quite widely used.) "

We also have "concepts"(reality) of Chinese and Russian and etc. and our ancient genomes can't be fit in those categories, just as they can't be fit into racial categories.

The first reaction people have to 50% of Irish's ancestors being from people who lived in Russia 5,000 years ago and another (close to)half from people who lived in Turkey 8,000 years ago, is Irish are part Russian and Turkish, partly descended from actual Russians and Turkish.

The issue isn't that people can't understand genetics because of a believe in race(or ethnicity for that matter) but because their minds are set on the modern world(made up of races and ethnicities), and you can't blame them or society for this.

What people have to learn inorder to understand human genetics, is that most modern ethnic groups and genetic clusters(we call a lot of them races) are often a mixture of older ones and therefore humans in the distant past can't be placed somewhere in modern human diversity.

Open Genomes said...

What people may not know is that "Neolithic" haplogroup G2a is also found among Egyptians, Sudanese, Libyans, Berbers, Fulbe (Fulani), Mende from Sierra Leone, and among Cameroonians from the highlands of Southwest Province. The 1K Genomes Mende shows no sign of European admixture, and yet he is in the Sardinian-Berber-Ashkenazi G-Z6028 clade with a tMRCA of 3,700 ybp. The only other non-African haplogroup found among the West Africans is R1b-V88. The same Cardial culture which Els Trocs in R-Y7777* belonged to was found in Morocco and Tunisia as well as Italy and Iberia. It may be that R1b-V88 co-migrated with various G2a2-L30 clades, but G2a2 generally had a more northerly route and R1b-V88 (or R-Y7777) had a more southerly route around the Mediterranean.

There's a simple test that can falsify a migration from one continent to another:

If R1b-V88 originated in Europe, then why don't we also see I2a clades in West and Central Africa? Did the Sardinians do a DNA test to check to see who was R-V88 and G-Z6028 vs. I2a-M26 before they allowed them to take the boat to Africa? ;)

We can say that about any hyptothetical R1b "back-migration from Europe":
Why don't we see I2a alongside R1b outside of Europe?

Again, it's not what we'd "expect" because it's "R1b" but given the distribution of R-V88 and the aDNA evidence from Els Trocs and STM1, it makes sense.

Open Genomes said...

What people may not know is that "Neolithic" haplogroup G2a is also found among Egyptians, Sudanese, Libyans, Berbers, Fulbe (Fulani), Mende from Sierra Leone, and among Cameroonians from the highlands of Southwest Province. The 1K Genomes Mende shows no sign of European admixture, and yet he is in the Sardinian-Berber-Ashkenazi G-Z6028 clade with a tMRCA of 3,700 ybp. The only other non-African haplogroup found among the West Africans is R1b-V88. The same Cardial culture which Els Trocs in R-Y7777* belonged to was found in Morocco and Tunisia as well as Italy and Iberia. It may be that R1b-V88 co-migrated with various G2a2-L30 clades, but G2a2 generally had a more northerly route and R1b-V88 (or R-Y7777) had a more southerly route around the Mediterranean.

There's a simple test that can falsify a migration from one continent to another:

If R1b-V88 originated in Europe, then why don't we also see I2a clades in West and Central Africa? Did the Sardinians do a DNA test to check to see who was R-V88 and G-Z6028 vs. I2a-M26 before they allowed them to take the boat to Africa? ;)

We can say that about any hyptothetical R1b "back-migration from Europe":
Why don't we see I2a alongside R1b outside of Europe?

Again, it's not what we'd "expect" because it's "R1b" but given the distribution of R-V88 and the aDNA evidence from Els Trocs and STM1, it makes sense.

Samuel Andrews said...

@Open Genomes,
"And if the Andronovo of Central Asia and the Afansievo people of South Siberia are closer to the present-day Europeans than the First Farmers, what does that make them? "

What you're seeing here is what I just posted, ancients often can't be placed somewhere in modern diversity.

I see Europeans and almost anyone else as pre-historic Latinos. Latinos are multi-racial like them but the mixture occurred in historical times not pre-historic, and so people are aware of the fact they're mixed. Despite this hispanics are seen as a race by the masses.

5,000 years from now people might be asking "Were Spanish people Mexian or were Mayans Mexican? It's the same as asking "Was EEF white or was Yamnaya white?"

Gioiello said...

@ Open Genomes

I am seeing that someone answered your posts as I would want having done, for instance Rob. Of course what you said about hg. R1b is false, and there are 10000 letters of mine to demonstrate that.
1) R1b-V88 (both M18 and V35) is older in Sardinia/Italy than Middle East, Africa etc.
2) The oldest sample is that of Sexton, linked to Italian Marchesi.
3) R1b1a (we'll see at the P297 level) has been found in Italy and not elsewhere. Also the samples found at Samara are younger of many thousands of years.
4) I have written tons of letter about the trees of smal (Sergey Malyshev) and why the Italian (or sometime Western European) samples are the oldest in any subclades.
5) Of course you may think that they descend from Jewish prisoners, but very likely you know a little Roman history and the ergastula.
6) I have written a lot about the R-L584 from Samara to F38 to Iberian Grijalba and the Jewish Spyra family. You cannot be sure where an Y comes from, expecially of a people who in the history has had rapes pretty any year.
7) I think to be a "scientist" so that the truth above all, for that I am open to your analyses and your hypotheses, and I read your writings always and with interest, but I don't believe in your humanitarism.
8) The world will go to a nuclear war very soon, and each of us had to choose whom he is with.
9) Were you with the agents in the fortress or with Russian missile?

Davidski said...

Back on topic please.

The topic is the Estonian Biocentre Human Genome Diversity Panel (EGDP).

Open Genomes said...

@David and all,

I'm certainly not calling for any kind of "censorship". Rather, I would like people to challenge others with the data when these sorts of unscientific ideas come up.
The issue of which Y clades are "really Jewish" popped up, when this actually had nothing to do with original discussion at hand.

Too often the people who continue to perpetuate these ideas are the "loudest" and no one wants to challenge them. We saw it right here, for example, when all you have to mention is "The Jews" (or rather, Ashkenazi Jewish origins) and "everyone has a theory" - except what the data actually shows, where the Ashkenazi Genome Consortium and your nMonte generally agree.

The reason that this is so "hot" when you merely listed some nMonte results for a certain population isolate is that for religious and "socio-political" reasons, anything that has do with a group that is the focus of the Hebrew Bible, New Testament, and Quran, and whose ancestry is related to a current major world conflict, is bound to make many people "agitated" about their "true origins".

The problem is the notion of "purity".
Being a mixture of earlier populations doesn't "delegitimize" anyone, except for people who believe that "purity by descent" is a must. (FYI, blood descent is not a Jewish religious concept, but there is a tremendous obsession with "who are the Real Israelites" to the point where several different religions are founded on it, including the Worldwide Church of God and various "Black Israelite" groups.)

Are Jews a "Race"? Are Jews "White"? Somehow, this nonsense became critically important in the 19th century, and deadly in the 20th century. It's still discussed:

Jews Are a ‘Race,’ Genes Reveal by Jon Entine

How did a crank, a former sportswriter without a degree who uses the title "Dr." and who claimed in a talk at the first so-called "FTDNA Conference" (a sales/marketing seminar, that is) that "Judaism is Blood Religion" and "Jesus came and changed the human genome from blood to faith", get to be the arbiter of dialog about Jewish ancestry?
And how did Eran Elhaik get a position at the University of Sheffield, while pushing his "opposite" crank theories about the "100% Khazar" origins of Ashkenazi Jews?

There's something that none of you know. The following paper
A substantial prehistoric European ancestry amongst Ashkenazi maternal lineages - Costa et al. (2013)

Was submitted for review and rejected six times and finally published on the seventh try. Some "anonymous" peer reviews had totally irrelevant nonsense questions and criticisms, but some had racist statements - so much so that the editors of respected scientific journals refused to show these reviews to the authors.
Are these the journals "Science" and "Nature" or "American Renaissance"?

Does that surprise any of you? Have you heard of such a thing going on in other "boring" genetics studies?

Just think about it ...

Open Genomes said...

This is how "academics" argue:

'This is when Elhaik chimes in. For him, Entine has revealed his own prejudice in his comments: “I would like to thank Jon Entine for disclosing his scientific guidelines for studying Judaism as believe in God (though it is ok not to), patriotism (though living afar is also okay), and the purity of the blood line…Not surprisingly, the last two scientific principles of Entine share a common ground with the Nazi ideology. While this may makes sense to some people and may fit with their belief, for those of us who actually practice science this is mere nonsense."'

Did anyone ever think that this "common ground with Nazi ideology" applies to all of them, except the authors of the mtDNA paper?

From the same academic group, that did the above mtDNA paper:
A genetic contribution from the Far East into Ashkenazi Jews via the ancient Silk Road - Tian et al. (2015)

"Ashkenazi" M33c1 from Guangdong, and "Ashkenazi" L2a1l2 from Burkina Faso.
There are unique "Ashkenazi" clades from everywhere in between, including K1a4b which was found in I0867 from Motza, Israel, in the Levantine PPNB, and K1a4b1 is also found among Druze, Negev Bedouin, and an Armenian.

Are there no signs of this "exotic" non-Near Eastern non-European admixture among Ashkenazi Jews? The North African admixture is in fact detectable, even through IBD. It may be some East Asian admixture is not just "noise". Ashkenazi Jews do in fact have some small segments that appear to be East Asian.

There is no East Asian Y-DNA among Ashkenazi Jews, but in fact there is a small West African E1a1-M44 clade, and a tiny West African E1b1a1a1d1a-U290 clade as well.

Even though we can see that these findings were published, somehow, their implications for Ashkenazi autosomal ancestry are ignored.

The LRRK2 G2019S (rs34637584=A) Parkinson's disease risk allele accounts for 15 to 20 percent of Parkinson's disease cases in Ashkenazi Jews and about 40 percent of cases in North African Arab Berbers. The G2019S mutation is also common not only in Parkinson's disease patients but also in controls at a frequency of 2.2% in North African Berber-Arabs (13/597, [PMID 18666856]), 1.4% in Sephardic Jews (5/361, [PMID 18666856]) and 1.3% in Ashkenazi Jews (4/317, [PMID 16436782]) while it is found at a very low frequency of 0.06% in Europeans (1/1550, [PMID 18666856]) and not found in sub-Saharan Africans.

Here's an example were ignoring so-called "exotic" ancestry from the "wrong continent" can actually be damaging to the health of individuals, because it's thought that it's "impossible" that certain IBD risk alleles are shared between two seemingly unrelated populations.

Now ask yourself if ideas about "Race" interfere with scientific research. What we have right here is nothing compared to what really goes on in the academic world.

Dospaises said...

I suspected that even with the new rules I was going to read a lot of posts that lack logic and already my suspicions have been confirmed. Not only that but there are very long ramblings. sigh.

Nirjhar007 said...

You guys should do a bit more yoga ;) . You need to learn to relax and focus . Anyway, if any bud is interested -
http://www.nature.com/hdy/journal/vaop/ncurrent/full/hdy201679a.html

Ariel said...

You can take 100 europeans today and they will have more R1b variations that the rest of the world combined.

http://b2.ifrm.com/67/29/0/p510366/R1b_sub_structure.png

(Also note that pontic steppe almost complety lacks all the western variations. And yes they have same L23 non western stuff as Yamna)

That's why R1b and at least most of his subclades came from Europe/Russia (if not western Europe).

Plus Villabruna was upstream to both L23 and V88, what a coincidence.

MaxT said...

@Open Genomes

In context of Iranians/Western Asians/Jews - They are predominately "Basal-Rich" population. R*, R1a, R1b Y-DNA lineages did not** originate among Basal-Rich populations.

1. If we do find R1b or R1a in West Asia, it will be due to ANE-related migration.

2. R*, R1a, R1b, R2 Y-DNA descends from ANE-high population. Both ANE and WHG in turn descends from ENA (Eastern Non-Africans). ANE-high population lived in steppes and W.Siberia.

In context of South Asia - We do not have any aDNA from South Asia yet but even if we do find R1a, R1b & R2 aDNA there, it will be due to ANE-related migrations, whether it's paleolithic, neolithic or bronze age migration. They are ANE-rich population compared to West Asians.

MaxT said...

@Ariel

Villabruna's Y-DNA R1b is from ANE. He had 15% ANE (AG3 related) ancestry. David also says Villabruna inherited his Y-chromosome from ANE.

http://eurogenes.blogspot.com/2016/05/villabruna-cluster-near-eastern-migrants.html

epoch2013 said...

Why Estonian? Does Estonia have such a good DNA lab? I would applaud that, btw.

Ariel said...

MaxY

Yes R came from ANE, nobody is question that. I wonder if R1b came from WHG/ANE mixed populations since it was in western Europe already 14.000 years ago.

Nirjhar007 said...

Just call him Max ;) . on R1b some Mesolithic Anatolian samples will be nice ..

epoch2013 said...

@Max & Ariel

The only issue is that American-Indians - the persons with the most ANE - do not seem to have native R1b or R. EHG, 2/3 ANE, 1/3 WHG, was patently R1b/R1a, though. From ANE we *only* have so far: 1 Y-DNA: R*, 2 mtDNA: U and R3. We could do with more, and more importantly far better ANE samples.

capra internetensis said...

@FrankN

Please explain your objections to the TMRCA methodology more clearly, they aren't making sense.

The different lineages under P312 cannot have 'formed' at different times, unless the MRCA of P312 was a time traveller, or Methuselah. Nor is it technically impossible for P312 to have 30 different primary branches.

Mutation is a random process, not a steady one. While a lineage of a given age will have an *expected* number of mutations (and all descendants of one ancestor will of course have the same expected number of mutations), the actual number of mutations accumulated varies randomly but with a greater probability of a number closer to the expected value than of one further away. Hence each TMRCA is assigned a confidence interval - typically a rather large one - and the true age of the lineage will not usually coincide with the central estimate (by definition 5% of the time it will fall outside of the 95% CI entirely). Somewhat comparable to radiocarbon dates, where there is an inherent range of uncertainty because the amount of carbon-14 remaining in the sample after a certain time is probabilistic and not fixed.

So the calculated age of most lineages under a node will be substantially older or younger than the real age, and some will be *much* older or younger. This is the *expected* distribution. It doesn't mean that the lineages 'formed' at a range of different times, peaking around a certain period. A more accurate TMRCA estimate can be obtained by averaging the number of mutations on each *independently descending* lineage from a node, which is what Y-Full does.

Now with subclades only slightly younger than the parent clade, as is the case with most of P312's branches, it is expected that some of them will have estimated TMRCAs older than the parent clade's. If you had an independent date for a subclade, like from ancient DNA, this could be used to set a lower bound. However, the age of the subclade is just an estimate, based on only a subset of the data used to estimate the age of the parent clade, and hence intrinsically less accurate. If the estimated age of the subclade is older this does not mean that the TMRCA of the parent clade is an underestimate, but that the TMRCA of the subclade is an overestimate (probably). The estimated ages of subclades with many descendants are more accurate than those of individual isolated lineages, but they will still randomly deviate from their actual ages. Laymen (including me) often overinterpret these things, forgetting the high degree of inherent randomness and uncertainty.

In reality a reliable SNP only comes along every three or four generations on average, so the branches descending from the P312 node do not represent the 30 sons of the P312 patriarch, but the surviving paternal lines from his clan over a number of generations (they only become distinguishable from the node once a mutation occurs in the line). The P312 patriarch himself could have had only 2 sons. This means that there are almost certainly invisible subclades under the P312 node, i.e. some lineages are actually more closely related to each other than others, but we can no longer tell from the DNA. For instance, it is entirely possible that some early lineage under P312 went for several centuries without a SNP, then expanded, but this would be undetectable.

Hopefully in the future we will have lots of high quality ancient genomes which will allow us to directly calibrate Y haplogroup ages.

Now I am not saying you should take Y-Full's TMRCAs as gospel, you certainly should not. But from your posts I think you are misunderstanding the methodology.

MaxT said...

@epoch2013

Native Americans carry Q Y-DNA which is sister-clad of R* Y-DNA.

In context of EHG - Lazaridis et al (2016) says "(EHG) derive ~3/4 (75%) of their ancestry from (ANE)." It seems like WHG in EHG was most likely female-mediated.

Matt said...

@ Davidski, while I don't really doubt that analysis is plausible, do you think using the Basal Rich K7 which is based on the ancient populations, might not be less accurate than a model based on recent people?

It seems like surely there is some degree of recent drift beyond what is captured by the K7 three, that could be informative as to which populations contributed to the AJ population. Unless there is zero useful modern drift, the K7 should be lacking some information about modern populations (assuming the K7 clusters aren't incorporating any drift from recent populations, which they shouldn't be). Even is small because much less time and drift proceeds slower in large population units.

Basically, I understand why you wouldn't want to understand ancient people in the terms of Eurogenes K15 modern people, but surely the Eurogenes K15 should be better for this sort of purpose? Or a modification of Eurogenes K15 based on newer and more complete datasets (possibly without supervision).

@ OpenGenomes: Early Europeans were part of a WHG isolate that vanished in its pure form millennia ago. Early European farmers were more similar to Levantines and Turks. If Arabs and Turks are not "White" and Somalis have some substantial "LBK-like" admixture, then does that mean that the Turks and Kurds in Germany today and the Somalis in Italy and Sweden are more like the pre-Steppe Europeans than Europeans today?

This might come across as selectively quoting your wider post, but as a technical the Early European Farmers actually weren't more similar to present day Levantines and Turks, judging by the D-stats (double) or f3 stats measures.

For example, Lithuanians share more drift with the Boncuklu population than do the Levant Neolithic, or present day populations from Palestine or Turkey, who are way down the list.

That's pretty abundantly clear in both the D stats Davidski's run and the outgroup f3 stats from the original paper that covered the Boncuklu

There's a lot of substructure in the ancient Middle East, enough that genome wide it looks like the EEF actually are more related to the Europeans who are their descendents (despite them being heavily admixed with Yamnaya related populations and WHG) compared to other Middle Eastern people who are not descended from them.

Now, one day when someone does a study on phenotypically active mutations, it may be that the phenotype of the ancient EEF actually is closer to present day Middle Eastern people than present day Europeans.

But for now, I think we can't say that they were more like people like Levantines or Turks than current Europeans (even the Europeans who are most extremely *un*like the EEF, Lithuanians and Finns and Estonians and Russians). The Turks and Kurds in Germany and the Somalis in Italy and Sweden are *not* more like the pre-steppe Europeans than current Europeans are, by essentially any measure we have.

Slightly conversely to your point, it would actually be interesting to see an outgroup f3 od D stat run for the Natufians against both recent Southern Europe and the Near East (Palestinians, Jordanians, etc.)

Gioiello said...

@ capra internetensis

I completely agree with what you said above. It is what I've been saying from many years. All who dislike me and my theory will realize how I was at least 5 to 10 years in advance.

epoch2013 said...

@MaxT

O yes, I do think Q was a ANE Y-lineage, simply because it is so abundant among American-Indians. It may be that the eastern ANE that mixed with proto-east-Asians were slightly different from MA1 and both Afontova Gora samples or the ANE that became part of EHG.

Richard Rocca said...

@Gioiello,

"Of course that isn't true. You demonstrate that this line hasn't so far descendants, is a "dead line", but not its age, except the fact that it has in common with the survived lines the SNPs U152, but you don't know when it was born."

Yeah, I check the other L51 samples as well, so at some point we can't assume that we were so unlucky that all of those samples that were tested were dead ends. Of course, they weren't. And if they were much older as you claim (want), there would be many many branches detected, and there aren't.

Grey said...

"So where do all of you stand? Remember: "Silence is consent.""

I think the bad stuff you list is generally the result of resource competition which people rationalize in whatever way is most convenient.

The solution is minimizing resource competition.

#

"If R1b-V88 originated in Europe, then why don't we also see I2a clades in West and Central Africa? Did the Sardinians do a DNA test to check to see who was R-V88 and G-Z6028 vs. I2a-M26 before they allowed them to take the boat to Africa?"

Personally I don't (currently) think R1b originally came from Europe. I think it was originally mammoth steppe and drifted south after the mammoth died out - although that's been a while.

However on your main point IIRC the I2 in Sardinia is mostly in the mountainous interior protected until recently by malarial swamps with the other haplogroups around the coast.

I think it's in this one

http://www.ajol.info/index.php/ijma/article/viewFile/60356/48592

Grey said...

"I think it was originally mammoth steppe and drifted south after the mammoth died out"

hence V88 for example could end up in the middle east somewhere and spread from there

(among other possibilities)

Gioiello said...

@ Richard Rocca

Are you saying that you checked a sample positive for L51 but negative for all the known SNPs downstream?
If we take the YFUll tree:
R-L51 L51/M412/S167/PF6536 * PF6535 * CTS10373/PF6537/FGC39+2 SNPs formed 6300 ybp, TMRCA 5700 ybp info
id:ERS257000 ITA [IT-CA]
R-L51*
we see that R-L51 had a subclade in
R-L151 P310/S129/PF6546 * YSC0000191/PF6543/S1159 * L52/PF6541+9 SNPs formed 5700 ybp, TMRCA 4900 ybpinfo
from which all the descendants
and a minor subclades in
R-Z2118 Z2118/PF7589 * Z2113 * Z2112+5 SNPs formed 5700 ybp, TMRCA 5100 ybp info
Apart the fact that ERS257000 ITA [IT-CA] seemed to me a phantasm clade (and I wrote about that, and evidently neither smal nor YFull agreed with me) and that against my interests, because an L51* in Sardinia would be in favour of my theory of an Italian Refugium,
anyway we see that L51* emerged from a little bottleneck of 5 SNPs, and L151 after a bottleneck of 12 SNPs. Thus what do you mean for your R-L51? Has it all the other 4 SNPs at its level or not? If it hasn't all the other 4 SNPs recognized from YFull, it is a dead line, or, at least, a new subclade whose we haven't so far descendants but we cannot exclude that someone may be found in the future. Anyway a minor line.
You said that Villabruna was a dead line, and it is possible that it isn't in the line which brought to P297* ancestor of we all. Anyway it does mean to me that he was part of a tribe of R1b1 from where the R-P297* line was the unique to survive and is our ancestor.
To think that it were a single case isn't parsimonious. Better to think to a "clan Villabruna". So the R1b found in the aDNA from Germany could be the "couche" of all the R1b descendants, but also some migrants from the center of R1b, as I think were the Samara samples, and you should read what I wrote about the fact that they aren't the ancestors of the Western European subclades, if not in rare cases, like perhaps R-CTS7763. All to discuss the rest, and also about R-M73 I think having demonstrated that Western European samples don't derive from Samara, and are older.

jparada said...

A question, do the Arab Israelis differ from Palestinians, and if so, which populations do they resemble most?

Richard Rocca said...

@Gioiello,

No, what I said was that I checked all ancient DNA samples that were at least L51+ and none belonged to subclades below DF27 nor U152.

a said...

Blogger Open Genomes said...

@a - Thank you. :)

Please let me say that No DNA result is "embarassing".

Thank-you kindly,
This is simply a rhetorical question.No need to reply.
I don't want to detract from the thread[All the new Biocenter samples]. One of my heavy-weight favorites in the field of early genetics-Karl Landsteiner. Until today I did not realize both Landsteiner and Harry Hamilton Laughlin passed away in 1943.

So here is the rhetorical question. With a mix of Italians, Slavs, Jewish diaspora etc as the audience[include many posting] .
Could you imagine a debate between Karl-Landsteiner versus Harry Hamilton Laughlin. Mediated with a [personality, passionate but neutral-well versed in human nature ] someone like Edwin Black?
Again only a rhetorical question.No need to reply.

FrankN said...

@capra: I don't want to do a full review of yFull's Age/ TMRCA calculations, but will instead focus on P312 and subclades, which will nevertheless raise some issues of general concern:

1. Their method yields a calculated P312 age of 4.176 ybp. However, that age is revised upwards to 4.5 ky, as their age estimation for L21 (4.5 ky) is higher (see the note at the end of the P312 "info" box). So, in this case they themselves deviate from the general approach that you have well described in your post.

2. Their stated rationale doesn't look sound. There are several subclades for which yFull calculates a higher age than the 4.9 ky they atribute to P312, including:
- YF05889 (singleton) 6.6 ky
- DF19 5.6 ky,
- L21 5.4ky (my recalculation, their table says 5.2 ky)
- Y18209 5.3 ky,
- Z2244 4.9 ky,
So, why are they assuming L21 to be the oldest subclade, instead of DF19? And why are they correcting L21's age downwards to 4.5 ky, instead of accepting their own age calculation of 5.2 (5.4) ky? Why do they estimate the age of P312 as 4.9 ky, instead of the 4.5 ky they attribute to L21 (and, btw, also to U152 [RISE563]?

I suspect that the real reason for their corrections is Quedlinburg BB I0806, P312, 2431-2150 calBCE - obviously older than their calculated age of P312. I furthermore suspect that assuming L21 to be the oldest P312 subclade has something to do with Rathlin1, 2024–1741 BC, L21->DF13->DF21. Moreover, RISE563 (U152) hasn't been C14-dated, but Heyd (2000) places the Osterhofen-Altenmarkt gravefield into BB Phases A2a/b, ca. 4.5-4.3 kya.

This means there are pretty good reasons aDNA-wise to fix the ages of L21 and U152, and consequently the P312 TMRCA, at at least 4.5 kya, but those reasons don't have anything to do with yFull's standard methodology. That methodology is neglected (L21 age calculation overrides upstream P312) or bent (disregard of own age calculation for L21), without proper explanation.
Hmm.. does such practice install trust in the appropriatenes of their overall approach?

Well, and then we have the Poznik e.a. 2016 study, yielding different, and generally older age estimates (e.g. 5.6 ky for P312, 5.5 ky for U152). I tried to figure out reasons for the differences from Poznik's SM 4.5 & 8:
1) Poznik e.a. use Fu's (2014) mutation rate of 0.76 per bill. SNV, derived from calibrating UI, which is lower than the yFull rate of ~0.82 per bill SNV.
2) As the likelyhood of detecting mutations depends on the sampling quality, Poznik e.a. discard most terminal branches from their age calculation. Instead they build a skeleton of datable split nodes by only using sequences of 5X or greater coverage, and subsequently traverse branches downwards to infer the remaining nodes. In contrast, yFull uses all terminal branches, even those of rather poor quality (>0.9X), to construct the network bottom-up. This should generally result in lower age/TMRCA estimates - the more terminal branches included, the greater the underestimation. P312, with 22 terminal branches (singletons) considered for age estimation,
seems especially vulnerable to such underestimation. Interestingly, for upstream L151/P311, Poznik (5.9 ky) and yFull (calc. 6.1ky, stated 5.7ky) more or less agree on the age.
http://www.nature.com/ng/journal/v48/n6/extref/ng.3559-S1.pdf

Gioiello said...

@ Richard Rocca

"@Gioiello,
No, what I said was that I checked all ancient DNA samples that were at least L51+ and none belonged to subclades below DF27 nor U152".

Then you should do what I said above: to check all the SNPs at the L51 level, at Z2118 level and at L151 level... so you'll understand if that line is a dead line or not.

CroMagnon said...

Frank is not alone to suggest that some estimates are underestimates by, say, 20%
If so, and looking at L51 as a whole, not just P312, gives us 5700-> 7125 yBP for TMRCA and 7800 yBP for "time of formation".

One might thus argue that L51 expanded c. 5000-4500 BC, although from where this took place is hard to say.

Moreover, the strathlin samples were already DF27 and L21.

No hard conclusions but food for thought.

Simon_W said...

jparada, I might be wrong, but I suppose Arab Israelis are just Palestinians with an Israeli citizenship, which they have because they are from within the official borders of that state.

MfA said...

Arabs in the David's database is the Christians from Israel. They're a different bunch than Muslim Arabs in Israel/Palestine.

capra internetensis said...

@FrankN

Yeah, I expect in this case they have adjusted the age due to ancient DNA.

The rho statistic is known to be biased downward - there was a paper on it but I can't find an accessible version. One problem (as alluded to before) may be that subclades with no mutations along the branch from the parent node are treated as multiple independent branches, so shorter branches end up being weighted more heavily. MCMC estimates should be more accurate than rho statistic in principle but of course they are also a lot more computationally demanding.

I don't know what you mean by sequences with 0.9x coverage? Y-Full uses BigY samples with 50x+ coverage, and corrects for differences in coverage breadth. However, I don't know if they correct for the higher coverage of internal branches over terminal ones - I don't see anything about it in the FAQ. If not, that might be why the paragroup samples are decreasing the average branch length.

epoch2013 said...

@Open Genomes

So where do all of you stand? Remember: "Silence is consent."

I like to remember you that this thread did not discuss the numerous suggestions you brought up, nor did it suggest such things. You hijack the thread with numerous straw puppets and then suggest we need to take a stand. Please stop that. We're not your private army, as the saying goes.

@MfA

Are you sure? Christians in Israel (Nazareth for instance) are often Palestinians. George Habash of the infamous PFLP is a Christian Palestinian, for instance.

Simon_W said...

@MfA

Well, I checked the sheet, the new Israeli Arabs from the EGDP are three Christians and two Muslims, so a mixed bunch. And the Palestinians David usually used are all Christians, really?

Simon_W said...

@ Matt, re: Europeaness of EEF

Good point. If we take modern Sardinians as a proxy for EEF, you can see they even look very European. These are native Sardinians from the interior of Sardinia (courtesy of Andreas Vonderach): https://jpst.it/NTTa
Upper row #1 could even pass for British Islander imho, upper row #3 and bottom row #2 could easily pass for French. More exotic looking types do occur on Sardinia, but I guess they're more common on the more admixed coast.

epoch2013 said...

@CroMagnon & FrankN

If those dates are underestimated by 20% we are bound to find mutations in ancient DNA that can't have existed yet, according to those underestimations. Do we already have such samples?

Gioiello said...

@ epoch2013

"If those dates are underestimated by 20% we are bound to find mutations in ancient DNA that can't have existed yet, according to those underestimations. Do we already have such samples?"

Can you show us which is rationale reasoning as to why if a SNP should be worth more it should multiply the number of the SNPs?

Ariel said...

Simon_W

From a(my) central italian perspective Sardinians are very "dark", especially for the almost complete lack of blonde hair and blue eyes, you will find way more central europeans looking people in southern Italy, way more. Even if they have no ANE they still can look like caucasians and east- med, the reason its kind of a mistery for me. Generally the pics on the internet of sardinans are very (very) cherry picked. I'll post some crowds.

http://lanuovasardegna.gelocal.it/polopoly_fs/1.13864782!/httpImage/image.jpg_gen/derivatives/gallery_978/image.jpg

http://lanuovasardegna.gelocal.it/polopoly_fs/1.13424214!/httpImage/image.jpg_gen/derivatives/gallery_978/image.jpg

MfA said...

Simon_W

Yeah but David only used the Christians: Arab_Israel_1:GS000016206,Arab_Israel_1:GS000016217,
Arab_Israel_1:GS000016885

https://docs.google.com/spreadsheets/d/1tFAa7oxWpcNN-OdMMjBdb4NeWKG7EkpKMzZJVW2_MME/edit#gid=1581589557

MaxT said...

@Ariel

ANE has nothing do with "Caucasian". If that was the case then Native Americans and Siberians should look Caucasian by your logic.

ANE and WHG both descends from ENA (Eastern Non-Africans) and more related to say Australian Aboriginals for example than to "Basal Eurasian" component.

Sardianians are predominantly basal Eurasian, Basal Eurasian is basically Caucasian.

Ariel said...

I was referring to the lack of CHG, and the typical look of people from the caucasus, you know the caucasus? The region with mountains, never heard of it? Georgia, Armenia... those places...

MaxT said...

@Ariel

Thank you for clarifying that, you seem to have typed ANE instead of CHG in your previous post.

Matt said...

@ Simon_W, no way I really have a clue about the phenotype of Sardinians vs the Near East. What you say could be correct though.

There was a paper last year about how Europeans seem to get more ancestry from the WHG than they should do, inside genes, relative to outside genes, so Europeans may appear more WHG descended in appearance than they are. May apply to / against Sardinians.

@ MaxT, in the latest numbers from Lazaridis , the estimate for the Anatolian Neolithic is around 25% Basal Eurasian, while the Middle Neolithic farmers are at 20%, having picked up 20% WHG (and similar to the Yamnaya) -
http://www.nature.com/nature/journal/v536/n7617/images/nature19310-f2.jpg.

The Sardinians should be more like the Middle Neolithic based on Davidski's methods, so not predominantly Basal Eurasian (about 20%) on the latest estimates. It may be more like 30% at the end of the day, but should be not predominantly Basal Eurasian, rather predominantly WHG.

Shaikorth said...

@Matt
Yup, that's roughly it, while Caucasus pops are less "basal" than Satsurblia which is about 30%. Predominantly they, as is typical of West Eurasians, are ANE-related and Villabruna (WHG) related. Reich lab's newest trees used Kostenki14 as the descendant of the actual West Eurasian clade IIRC.

MaxT said...

@Shaikorth

In context of Native Americans, can they be considered partially "West Eurasian" when it comes to ANE? ANE seems like intermediate component rather than West Eurasian.

MaxT said...

@Matt

Thanks for the link, Matt. That makes sense.

Davidski said...

Quick K=8 ADMIXTURE run with EGDP, HO, most ancients and various other samples.

https://docs.google.com/spreadsheets/d/1XswOpaOowUkH-Io2sjEN1dN9HeLFBKcKPd1K3dxIidc/edit?usp=sharing

ArtemisVentus said...

it's interesting in your spreadsheet that the Yamnaya are roughly 40-45 % Iran Neolithic and 55-60% West Eurasian HG, but Corded ware is about 20-25% Iran Neolithic and 55-60% West Eurasian HG

if CWC is derived from Yamnaya shouldn't the ratio be the same?

Davidski said...

if CWC is derived from Yamnaya shouldn't the ratio be the same?

Nope.

CroMagnon said...

As far as reconstructing moderns, we should stick to Admixture K's based on actual prehistoric groups (EHG, ANF, WHG, Iran-N) instead of reconstituted proxies like ANE or "basal". Otherwise it can confuse people , as above

Davidski said...

ANE was an actual prehistoric group. Another name for it was used in the Fu paper though: Mal'ta cluster. But they also called WHG the Villabruna cluster.

In my K7 I just use AG3-MA1 and Villabruna.

FrankN said...

@capra, CroMagnon, epoch:
The current practice of estimating hg formation/ TMRCA by applying an average mutation rate to the number of observed mutations has a number of theoretical problems:

1. It is well documented that certain STRs mutate more frequently than others. This should also apply to the SNPs underlying those STRs. Hence, regarding the accumulation of mutations as a purely stochastic, probability-driven process, which is the key assumption underlying the a/m current approach, is, nicely put, quite a simplification.

2. We know that genes (SNPs) are under selection, positively or negatively (and such selection may be one of the reasons for different STR mutation rates as addressed above). Again, not a purely stochastic, random-driven process but instead one dictated by external factors, e.g. nutrition, viral/bacterial pressure, social (mating) advance.

3. An implicit assumption underlying all probabilistic age/ TMRCA determination via an average mutation rate is constant population size. Of course, that assumption is unrealistic. In fact, bottlenecks are likely to enhance positive selection, purging deleterious mutations, thereby reducing the number of accumulated mutations. Conversely, periods of population growth reduce selective pressures and allow non-sense and even deleterious mutations to survive, thus increasing the observed mutation rate. These mechanisms are regurlarly used to identify phases of population bottleneck and Expansion, respectively.

4. There is quite some evidence that the mutation rate hasn't been constant over time, but accelerated over the last millenia. Possible reasons include population growth, and increased selective pressure (e.g. Neolithisation->Zoonoses).

As such, I generally think that age/ TMRCA determination via applying a constant mutation rate to the number of accumulated mutations is a theoretically poorly founded, rather crude way of approximation. However, it is the best method we have so far. Holding a Bachelor of Business Mathematics degree, I am well aware of the methodological problems involved in coming up with more precise estimates, and certainly don't feel myself prepared to provide any, but I shall welcome anybody, who improves age/ TMRCA calculation by developping theoretically and stochastically well-founded models.

Having said that - for the time being I will stay with the Poznik e.a (2016) approximations, however crude they are. According to them (Supp.Tab 17, 19), R1b started to expand in earnest under P311/L11 some 5,9 kya, with in average 1.2-2.1 sons per father over 18 generations (540 y), after which (i.e. some 5.4 kya) there were some 2.3k L11 males, growing to 3 million by 0 CE.
As stated before, that growth pre-dates any "Steppe" influence into Central/ Western Europe. However, it corresponds well timewise with the second wave of Neolithisation (Alps & northern forelands S. of the Danube, British Isles, North European Plain, Scandinavia), and the "secondary product revolution" (dairying, wool, etc.). Subsequent expansions (P 312 5.6 kya, DF27 and U152 5.5 kya, the two latter acc. to Poznik showing a massive burst over just 1-5 generations) correspond to de-population of the Michelsberg core (Paris Basin and areas along the Rhine), and appearance of Michelsberg traits over a wide area streching at least from Ireland to Central Sweden (not sure about NE Iberia in this respect).

Capra: I will follow up on your remarks and P312 in a separate post.

CroMagnon said...

Davidski

"ANE was an actual prehistoric group. Another name for it was used in the Fu paper though: Mal'ta cluster. But they also called WHG the Villabruna cluster."

Yes but whilst Amerindians do indeed derive their ANE from the Mal'ta/ Afantova cluster, Europeans do not. Their source for ANE was whatever was present in Mesolithic Eastern Europe, the Copper age Caucasus, etc
There was only limited direct from from the Altai- seen by the odd haplogroup Q and mtDNA C.

Davidski said...

Yes but whilst Amerindians do indeed derive their ANE from the Mal'ta/ Afantova cluster, Europeans do not.

Europeans derive their ANE from both EHG and CHG. How can you be sure that neither EHG nor CHG have any direct ancestry from the late UP Afontova Gora population?

CroMagnon said...

# Davidskiy

As stated, some gene flow is possible, seen by Y DNA hg Q (Khvalysnk) and mtDNA C (Karelia Meso, Ukraine Neo). But this could have come from Botai.

a few weeks back a study sampling several the Altai cemeteries of Neolithic and Bronze Age were all N (xN1c "Tat"); and Q.

So whatever the major source of ANE in Europe was, it came from elsewhere; or was in likelihood multilayered, with minimal additions from Altai. I'm sure the Harvard team will soon amend their models

Davidski said...

So whatever the major source of ANE in Europe was, it came from elsewhere.

AG3 is by and large a good proxy for the ANE in Europe, so I doubt that this "elsewhere" was far away from Afontova Gora, and in fact I can't actually see any evidence as yet that it wasn't Afontova Gora.

And the problem with modeling EHG as an unmixed population is that it's obviously mixed, and more obviously mixed than ANE, which is also probably mixed, but everyone's mixed if you go back far enough.

CroMagnon said...

Maybe if we define "near" as 1000 km due west ;)

Davidski said...

It's hard to say, but it's difficult to imagine a fragmented landscape with glaciers and then lakes like Ice Age Siberia being home to a homogeneous population. I suspect that it was populated by several fairly distinct meta-populations, most of which went the way of the Dodo during the Ice Age, and one or two just happened to survive and formed ANE.

So I think AG3 is a powerful clue as to where the ANE that is found in Europe, much of the Near East and the Americas came from.

CroMagnon said...

# Davidksi

Fair enough

#FrankN

How might you convince us of Y DNA haplogroup continuity through Michelsberg- TRB - CWC - BB cultures, against the obvious cultural and autosomal shifts ?

FrankN said...

@CroMagnon: "How might you convince us of Y DNA haplogroup continuity through Michelsberg- TRB - CWC - BB cultures, against the obvious cultural and autosomal shifts ?"
There wasn't much of a "cultural shift". TRB adopted many Michelsberg traits, especially the enclosures, but also, e.g., the "baking plates" unknown elsewhere during the MN, while CW and BB set forth the TRB Beaker Tradition. The CW burial ritus - crouched sidewards, E-W / W-E orientated (gender differentiated) was taken over from Bernburg aka "Terminal Central TRB".
The famous Bleckendorf grave, ca. 2650 BC that was mentionned earlier by Rob, figures with Furholt (2003, Fig. 26B) as CW, while Heyd sees it as early BB. Müller’s assessment is:
This grave contained a Corded Ware pot, a hammer-headed pin and copper objects (a hilt tang dagger and a metal awl) which have been related to Bell Beakers. This relation was confirmed by the orientation of the skeleton (N-S). The dating supports new typological analyses by Matuschik who considers the assemblage as a Corded Ware grave influenced by the pontic region on the basis of the form of the dagger.
The dagger finds its closest contemporary parallels in Bulgaria, the pin is Yamnaya, and the pot itself (see linked photo) bears the herringbone pattern that later became the trademark or Maritime BB, and is otherwise very similar to pots from the post-GAC Schönfeld Culture (Lower Middle Elbe) that settled nearby. Bleckendorf (in 1953 renamed into Egeln-Nord) is located approx. halfway between the ancient salt towns of Bernburg and Schöningen, and 30 km NE of Quedlinburg.
http://www.museum-digital.de/san/index.php?t=objekt&oges=11152

As illustrated by the Bleckendorf example, cultural distinctions may get quite blurred along the Middle Elbe. The region was multicultural (in addition to CW, BB and Schönfeld, you also find Bernburg and GAC there well into the 25th cBC). We don't know anything about the autosomes of the latter three, and neither about Michelsberg, so it is difficolt to tell whether there have been autosomal shifts. CW bears of course its distinctive pattern, but it is a cultural minority in the area and apparently had its autosomal impact diluted towards Unetice and later periods.

Having said that - Michelsberg is just one of the options to consider when it comes to a possible Central European vectors for initial growth and subsequent dispersion of R1b. We might as well be talking Pfyn/Altheim in SE Bavaria here, which colonised the Eastern Alps down to Trentino, apparently as transhumating pastoralists. Or the SW-German/ Swiss pile-dwelling cultures that drew on strong Mediterranean influence. Or the NMB in Burgundy, with cultural links a/o to Michelsberg and TRB. Until we get aDNA from any of them (plus the Balkans), it is all just guessing.

Chad Rohlfsen said...

R1b didn't spread with Michelsberg, nor will it be any different autosomally than all MN samples that surround it.

Onur said...

@CroMagnon

As stated, some gene flow is possible, seen by Y DNA hg Q (Khvalysnk) and mtDNA C (Karelia Meso, Ukraine Neo). But this could have come from Botai.

a few weeks back a study sampling several the Altai cemeteries of Neolithic and Bronze Age were all N (xN1c "Tat"); and Q.

So whatever the major source of ANE in Europe was, it came from elsewhere; or was in likelihood multilayered, with minimal additions from Altai. I'm sure the Harvard team will soon amend their models


The Altai region and Siberia in general have experienced big changes in genetics since the times of MA-1 and AG3. MA-1 (male) and AG3 (female) are from much before than the Neolithic, and they were autosomally ANE and had West Eurasian haplogroups (Y-DNA R* and mtDNA U*, and mtDNA R1b, respectively).

MaxT said...

@Onur

Y-DNA R* is North Eurasian (& extinct), it's ancestor Y-DNA P1* is still found in Siberia. Basal Y-DNA P* is basically South Eurasian (from Southeast Asia). Basal R* Y-DNA can't be termed West Eurasian.

P -> P1* -> (R* & Q*) migration din't take place through West Eurasia. Keep in mind that R* and Q* split from same ancestor aka Y-DNA P1*.

Karl_K said...

@Chad

"R1b didn't spread with Michelsberg, nor will it be any different autosomally than all MN samples that surround it."

I think you should be more specific. R1b is not an autosomal component, and it has already been found in more than one ancient autosomal context.

Onur said...

@MaxT

Yes, Y-DNA haplogroups R and Q are sister branches. But MA-1, a tested ANE individual, carries R like many West Eurasians today rather than Q, which is today concentrated in Amerindians and some eastern and central Siberian populations. But I guess there were Q-carrying ANE or ANE-like people too and Amerindians got their ANE portion from them. My point was to show that MA-1 or AG3 might well be ancestral to the ANE portion of at least West Eurasians.

jparada said...

@Onur

With respect to Y-dna haplogroup R, i suspect a lot of people have it backwards: It's modern West Eurasians who got it from ANE, but it's unlikely ANE itself got it (or its predecessor) from anything that could be called "West Eurasian" in any meaningful way, if the distribution of its sister clades under K2 is any indication.

Ariel said...

I find interesting that poeple forgot that there is a thing called "indian sub-continent", and probably had a big role on the colonization of west eurasia, ANE included.

Samuel Andrews said...

@Karl_K,
"I think you should be more specific. R1b is not an autosomal component, and it has already been found in more than one ancient autosomal context."

It tends to exist in people of a certain autosomal makeup. For example in EHG as opposed to WHG and in Steppe as opposed to MN.

capra internetensis said...

MaxT

I know of various reports of P1* in older studies, mainly from South Asia, but not of any from Siberia - do you recall where and according to what source?

MaxT said...

@Onur

Same as what @jparada said regrading Y-DNA R*.

@capra internetensis

Y-DNA P1* is found in high frequency in Siberia (Miroslava Derenko et al 2005)

Basal Y-DNA P* is mainly found in Southeast Asia, in Aeta population.

Seinundzeit said...

David,

That K8 is rather impressive. In fact, it doesn't even seem to need more work/development, looks pretty solid.

Ryan said...

Re: ANE in Europe and the Near East - I have a hunch based on the admixture results from the Anzick-1 paper and Davidski's models that the ANE that admixed with CHG was closer to Amerindians than the ANE that admixed with EHG.

Chad Rohlfsen said...

Karla,

I never said it was an autosomal component. When I say R1b, I figure everyone understands that I mean L51+.

Ryan said...

@Samuel_Andrews:

"It tends to exist in people of a certain autosomal makeup. For example in EHG as opposed to WHG and in Steppe as opposed to MN."

Villabruna is our oldest sample of R1b and it's level of ANE admixture is relatively low all things considered. It's also our earliest example of WHG. I don't doubt that the R1b in Villabruna came from ANE and ultimately the Steppe, but I think it's quite possible/plausible that it passed from WHG into Neolithic populations.

Davidski said...

@Ariel

I find interesting that people forgot that there is a thing called "indian sub-continent", and probably had a big role on the colonization of west eurasia, ANE included.

I seriously doubt that.

Anything in Europe that was previously though to be from India, like Y-HG R and mtDNA M, was already in Europe during the Upper Paleolithic, and it's unlikely that it got there from South Asia.

Matt said...

jparada: it's unlikely ANE itself got it (or its predecessor) from anything that could be called "West Eurasian" in any meaningful way, if the distribution of its sister clades under K2 is any indication.

I'm far from an expert on y-dna, so it seems like the question her is whether:

a) the mutation defining K2 happened in the East Eurasian group, after the "Crown Eurasian"* clade split into a East Eurasian (Onge, main trunk of East Asian, Papuan) and West Eurasian (Siberian Paleolithic - ANE - and the European Paleolithic clades) clades. Then K2 reentered the Siberian Paleolithic group via gene flow.

or

b) the mutation defining K2 actually happened within the unified Crown Eurasian clade; it's only preserved differentially in Southeast Asia and in North Eurasia via R and Q by chance, and has gone extinct elsewhere. In an earlier time in history there may well have been lots of K2 subclades about.

From the dating K2 apparently diversified at 40,000-50,000 BP. Then K2b2 / P group directly ancestral to R and Q and the Southeast Asian clades happens with approx 10,000 years after that.

Ust Ishim also apparently had K2 at 45,000 BP and crucially he seems to be a fork off of the Eurasian lineage leading to Crown Eurasian before the split of the West Eurasian and East Eurasian clades, which we know had happened at least by 37,000 BP with Kostenki14. (Though it's possible Ust Ishim's ancestors had split off well before 45,000 BP).

It seems like with the timescales involved, either scenario is plausible. To me it looks like it would come down more strongly on the K2, and probably K2b2, defining mutations happening in the pre-East+West Eurasia split of "Crown Eurasian" and probably even the K2 defining mutation happening prior to the Ust Ishim+Crown Eurasian split.

In either case though, like you say it seems like it's unlikely that K2 would've diversified within the West Eurasian side (or its Siberian subclade) after the split of the "Crown Eurasian".

* Using Ryukendo Kendow's nomenclature for the unified Eurasian group that includes all the main trunk of all Eurasians ancestry except Ust Ishim, Oase1 and the Basal Eurasian.

capra internetensis said...

K2 is only 1 SNP under K, MP 1 SNP more under that, and then there is 2 SNPs more before the divergence of P2 (from Wallacea) and P1 (QR). And 6 SNPs to MS (Sahul/eastern ISEA) on the other branch. So this is all very early in history of modern human expansion, within a matter of centuries, and probably only a millennium or two after the MRCA of F itself.

That is a long time for drift to destroy the evidence of K2 elsewhere, and we don't even know much about what we do have; there's a fair bit of K* around in South and Southeast Asia and so far there's a pre-NO Telegu from the 1000 Genomes and that's about it sequenced from India; Karafet has been plugging away in Indonesia, but the K2* in western ISEA remains poorly known (could be pre-NO also).

@MaxT:

That's P(xR1, Q-M3), not actual P1*, i.e. can be R2 or almost any kind of Q. In Altai-Baikal populations that will be mostly Q1a2 (L53* and L330).

Chad Rohlfsen said...

Matt,

He was talking about mtDNA. One of the samples was R1b.

Rob said...

@ FrankN

I'm still at a loss on how to accomodate the "R1b from the east" and "BB from the west" conundrum . For this to hold, then the German/Czech BB group must be tacit infiltrators from the Carpathian basin or something like GAC, who worked their way into the BB network (= a popular theory going round). Hopefully the next couple on papers clear this up

Rob said...

MaxT

"Y-DNA P1* is found in high frequency in Siberia (Miroslava Derenko et al 2005) "

Nope.
That study only tested for R1 and Q3, thus all their "basal P*" is actually Q (xQ3).
(It's always worth reading the "Methods" section of a paper).

Certainly, a later paper by Karafet featured no "P*" in the same populations.

@ Onur

"The Altai region and Siberia in general have experienced big changes in genetics since the times of MA-1 and AG3. MA-1 (male) and AG3 (female) are from much before than the Neolithic, and they were autosomally ANE and had West Eurasian haplogroups (Y-DNA R* and mtDNA U*, and mtDNA R1b, respectively).'

Exactly. But let's look at it from a scientific & rational way ;)

What does this tell us ? It tells us that the Altai-Baikal region was demographically unstable, hardly a 'refuge zone'

And you're actually incorrect about the lineages, Afontova was Q*, not R, and Mal'ta wasn't 'basal R*', but an R(XR1) parallel (not ancestral) to actual R1, and ultimately an extinct line.

In essence, only 2 R1a (?subclade) have thus been found pre-mid Bronze Age Altai/ Sayan - at Lokhomotif. The other 30 or so samples are Q and N, and we also have Q and NO* from Palaeolithic samples in the vicinity.

On the other hand, we have R1b (pre-M297) in LUP Italy, and all the R1s in Mesolithic Russia.
Combined with the lack of any archaeological link between 14 kya Italy and Siberia, then it doesn't take much effort to realise we need a more sober explanation

Gioiello said...

@ Ryan
"Villabruna is our oldest sample of R1b and it's level of ANE admixture is relatively low all things considered. It's also our earliest example of WHG. I don't doubt that the R1b in Villabruna came from ANE and ultimately the Steppe, but I think it's quite possible/plausible that it passed from WHG into Neolithic populations".

Ryan, I thank you, and, seen that I cannot write on Anthrogenica neither read it beyond the Stream Activity, i.e. only a first line, please, write to some stupid persons likely Richard Rocca, Agamemnon, vettor etc., who are speaking very likely about my "pet theory" (I love pets like cats much more than stupid men) and remember them what Romans did to the "paricidi"...

MaxT said...

@Rob

1. 14 kya Italy Villabruna was 15% ANE (AG3 related). Villabruna inherited his R1b from ANE. ANE is intrusive to Europe.

http://eurogenes.blogspot.com/2016/05/villabruna-cluster-near-eastern-migrants.html

2. Ancient-Indo-European samples are predominantly ANE derived. I don't think at any point in history mainland Europe populations were 50%-60% ANE.

3. None of Ancient-Indo-European samples cluster with Villabruna. They cluster with EHG and have close pull towards ANE.

In context of EHG - Lazaridis et al (2016) says "(EHG) derive ~3/4 of their ancestry from the (ANE)." and it's most likely that WHG in EHG was female-derived.

Onur said...

@Rob

And you're actually incorrect about the lineages, Afontova was Q*, not R, and Mal'ta wasn't 'basal R*', but an R(XR1) parallel (not ancestral) to actual R1, and ultimately an extinct line.

By R* I meant that MA-1 belongs to an undefined subclade of R. I know that his subclade is not ancestral to that of any known living R person, but he may well be one of the autosomal ancestors of the modern West Eurasian ANE portion given that he comes from an early population that has Y-DNA R. As for AG, I meant AG3, who is female, not AG2, who is male, which makes it clear that I had essentially autosomal ancestry in mind rather than uniparental ancestry. I did not mention AG2 because his genome is low coverage and his Y-DNA result is unreliable due to the low coverage.

MaxT said...

@Rob

You answered your own question, why do you think there were such high ANE population in Russian Europe but not in interior Europe? This ANE came from neighboring Siberia.

Yamnaya clusters with EHG, and has close pull towards ANE, that itself shows who their ancestors were basically. Considering Yamnaya were predominantly ANE-derived.

Y-DNA R & it's descendants are ANE derived, not WHG from interior Europe like you're trying to suggest here.

Onur said...

Bullshit. You're not presenting any coherent evidence to explain how Siberian groups reached Palaeolithic Italy, or Europe at large. Where's the link ? Name artefacts, cultures, dates. ? What route did they take (how did they pass the giant Ice Lake between western Russia and Siberia??)

I am not saying that the ANE portion of modern West Eurasians necessarily ultimately descend from Upper Paleolithic populations from around the Altai. It is just one of the possible sources. In all likelihood, ANE or ANE-like populations probably lived in a widespread area during the Upper Paleolithic times and it is now clear that by 14 kya their descendants had reached Western Europe (in the form of WHG). But, like I said, even before the emergence of WHG ANE or ANE-like populations had a widespread living area stretching from Eastern Europe to the Altai-Baikal region of Siberia, so the ANE portion of modern West Eurasians likely have multiple Upper Paleolithic sources.

Secondly, you're trying to explain away the very obvious fact that AFG was haplogroup Q, and all 98% of other Altain ancient samples are not R1

We cant just twist reality to fit our malformed opinions.


It is not me who is saying that the genome and Y-DNA results of AG2 are unreliable due to the low genome coverage, it is geneticists and genetics experts who are saying that. Just ask Davidski. My opinions are based on the genetic results, certainly not any pet theory. And if you pay attention, you will notice that I am constantly talking about possibilities and probabilities, not pushing any pet theory as some people here do.

Rob said...

@ MaxT

"You answered your own question, why do you think there were such high ANE population in Russian Europe but not in interior Europe? This ANE came from neighboring Siberia."

Different Ice Age refugia, that's why.

"Yamnaya clusters with EHG, and has close pull towards ANE, that itself shows who their ancestors were basically. Considering Yamnaya were predominantly ANE-derived. "

Not quote. Yamnaya is 50% non-EHG. And you have no proof that the ANE in Russia came from the Altai.

"Y-DNA R & it's descendants are ANE derived, not WHG from interior Europe like you're trying to suggest here."

Are you illiterate ? Because I'm not suggesting that at all.
Im stating that R1 had been in Russia since Palaeolithic, but did not come from the Altai, because there's no R1 to be found there. Obviously R1 arrived to Russia from elsewhere.
But don't bother arguing, because your ignorance on these matters is self-evident (just refer to your quote about P* in Siberia).

MaxT said...

@Rob

"Different Ice Age refugia"

Even if that was true, the source of ANE is still Siberia.

"you have no proof that the ANE in Russia came from the Altai."

Laughable, that's like saying WHG found in Near East did not come from Europe. ANE came from Siberia -> Russia, deal with it. ANE was intrusive to Europe.

Rob said...

Onur
Enough with the pontificating
State some facts, data and hypotheses
Sketch out the post- glacial history of Eastern Europe ?

Onur said...

@Rob

Onur
Enough with the pontificating
State some facts, data and hypotheses
Sketch out the post- glacial history of Eastern Europe ?


For what purpose? I am not MaxT, I have no strong opinions on the place of origin of ANE or its specific types.

Rob said...

Sorry Onur
Quite right, you two are in different ballparks
Your caveats are accepted, but ultimately change nothing in what I've outlined

Onur said...

Sorry Onur
Quite right, you two are in different ballparks
Your caveats are accepted, but ultimately change nothing in what I've outlined


No worries. My comments came after MaxT's comments, so you had to reply to both of us and coincidentally on a very similar topic.

Seinundzeit said...

These are some nMonte fits, based on the new K8, they look pretty good (my West Eurasian references were the four poles of ancient West Eurasian variation, ANE, WHG, Iran Neolithic, and Levant Neolithic):

Shugnan (tdj126_shugnan)
51.40% Iran_Neolithic
17.50% Loschbour
16.10% MA1
10.90% Levant_Neolithic
4.10% ENA (Ami + Papuan)

Kalash (HGDP00315)
76.8% Iran_Neolithic
14.9% Loschbour
8.3% ENA (Ami + Papuan)

Pashtun (HGDP00218)
66.35% Iran_Neolithic
12.35% Loschbour
8.80% Levant_Neolithic
9.15% ENA (Ami + Papuan)

Brahui (HGDP00023)
75.10% Iran_Neolithic
14.4% Levant_Neolithic
4.25% ENA (Ami + Papuan)
3.25% Loschbour
3.00% MA1

Iranian Mazandarani (GMII54)
60.05% Iran_Neolithic
34.15% Levant_Neolithic
3.00% Loschbour
2.80% ENA (Ami + Papuan)

These look rather accurate.

Also, always wanted to model the Roma, with modern populations:

Roma (GS000014325)

46.70% Armenian
30.25% Punjabi_Lahore
23.05% Romanian

Not too bad, makes sense.

Davidski said...

Two posts deleted.

No more discussions about politics. The topic is the EGDP.

Grey said...

MaxT

"ANE is intrusive to Europe."

I think ANE is connected to the mammoth steppe and the mammoth steppe extended a lot wider than the modern version.

https://cdn-assets.answersingenesis.org/img/articles/tj/v14/i3/fig2.jpg

So yeah it's not "Europe" but it's not, not-Europe either - there's an overlap.

Also worth remembering the post here a while back about mammoth extinction - how they seem to have disappeared west of the Urals sooner than elsewhere. What did those people do:
- lay down and starve
- follow herds east of the Urals
- drift south?

Davidski said...

Here's a K=13 run with a Steppe_EMBA cluster, probably based on a founder effect in Afanasievo.

https://docs.google.com/spreadsheets/d/1XUr5VPrvykkEV8NHsTORs0MdCloS9weB33okkWN9a80/edit?usp=sharing

The Fst distances are here...

https://drive.google.com/file/d/0B9o3EYTdM8lQVWpwN3dnMHN6STA/view?usp=sharing

epoch2013 said...

@Davidski

David, is there a way to check if Villabruna can be modeled as Vestonice (or even better, Ostuni or Paglicci133) plus an unkown HG, not yet sampled?

Gioiello said...

@ Davidski

I thank you for your spreadsheet, but I apologize for my ignorance: how is it possible to compare a sample of 14000 years ago with one of 4000? Isn't it the night where all the cows are grey (Hegel)?
When I compare the Y, I compare millions of single SNPs and their rare mutations. In the autosome I compare SNPs which have all the pound of their history upon themselves. Don't you think that you should compare only SNPs which are so low mutating one to haven't mutated within these last 14000 years?
The facts that you sometime use transversions may be useful for eviting the degradation of the SNPs, but not for eviting the pound of the time.
I understand that my discourse is philosophical, but out of metaphor: Villabruna has given changes to Anatolians, Middle Easterners, Caucasians that, when you find them in present day Europeans seem coming from elsewhere.
This mistake doesn't happen in the uniparental markers, because they are always recognized.

Gioiello said...

Of course "pond" and not "pound"...

Gioiello said...

I wanted to say "weight" of course. Diffcult English!

Gioiello said...

Even though "pound" derives from Latin "pondus" = weight!

Matt said...

Quick dendrogram based on pop averages for that 13 - http://imgur.com/NqKlEUU

Because I thought the SEA were the most interesting "new" populations, I created virtual populations for the East Asian part (I defined as Beringia, Siberia, East Asian [which peaks in Austronesians] components) and non-East Asian part of some of the populations in East Asia who scored between 95% and 77% East Asian, and also added them to the dendrogram.

The non-East Asian portion of the different SEA populations, insular and mainland, falls in different places on the part of the tree between Andamanese and Oceanians (which is probably only an approximation of the true structure). As well as slightly different parts of the East Asian part of the tree. May reflect persistence of Onge like people on the mainland or Indian migrations in varying proportions for different mainland SEA populations.

(With just some of the high East Asian - Siberian, Beringian, East Asian - only together: http://i.imgur.com/CLGOxlB.jpg).

Seinundzeit said...

That K13 looks rather nice.

Here is an nMonte-based model, for a Pashtun sample:

39.48% Steppe (Scythian_IA + Andronovo)
39.00% Iran_Neolithic
14.85% ENA (Andamanese + Ami + Papuan
5.70% Bedouin
1.00% MA1

A Kalash sample:

39.55% Steppe
27.85% Iran_Neolithic
12.35% MA1
10.80% Bedouin_B
9.35% ENA

Pretty good.

huijbregts said...

@ Davidski
Thank you for the K13. It is good to have an up-to-date high K calculator.
I tried Bell Beaker:

Bell_Beaker_Germany
Afanasievo 29.8
Corded_Ware_Germany 24.9
Sweden_MN 22.4
Baalberge_MN 12.75
Iberia_Chalcolithic 10.2
distance% 0.31%

Sisophon said...

Looking at the East Asian populations in the K13 sheet, I was surprised to see Steppe_EMBA show up in the 3 SE-Asian (Burmese, Thai & Cambodia) populations at about 2%, and Iran_Neolithic at about 4% in the same populations. Both these components mostly absent from the Vietnamese, Dai, Miao & Han.

Could this be a sign of the early Indianised state of Funan? Or is it too week a signal to mean anything.

Garvan

Davidski said...

You're probably correct that it's a signal of steppe ancestry coming into Southeast Asia via South Asia. But it is a very weak signal.

epoch2013 said...

@huijbregts

That's interesting. I bet Baalberge + "Swedish MN" could serve as a proxy for German Funnelbeaker. The Iberian MN is very interesting as the BB phenomenon is currently thought the have been originated in Iberia. However, there are also theories that see the origin of BB in Middle-Europe or the Rhineland.

ak2014b said...

@MaxT: "Basal Y-DNA P* is mainly found in Southeast Asia, in Aeta population."

According to a paper from 2012 about a place in southern India (by Spencer Wells and colleagues for the Genographic Project), there is some type of R1a1 and R2, besides some other R that is distinct from both, some Q, as well as some kind of P that is distinct from all of the former.

I looked up Q in the region, and a 2007 study reports low presence of Q1 as well as a slightly higher occurrence of Q*, Q4, Q5 in India. Q4 and Q5 were described as specific to India. And the Q* is not identical to that in Central Asia either:

"In the present study, we screened 630 samples belonging to different regions of India and observed 2.38% (15/630) individuals bearing Y-HG Q. It was interesting to observe that 14/15 samples did not show any of the already known Y-HG Q sub-haplogroups (Q1, Q2, Q3 and Q4), defined by biallelic markers M120, M25/M143, M3 and M346, respectively (Table 1). Only one individual was observed with the presence of M120 polymorphism, representing Q1 lineage."

"It was interesting to find that most of the Indian Q (Q4, Q5 and Q*) associated Y-STR haplotypes were separated from the bulk of Central Asian Q* associated haplotypes."


So there is P (xR,xQ) in Aeta Filipinos and in Indians. And elsewhere on this blog where people discussed M52 found in Maykop, I already brought up this news of a study about mtDNA M52 in the Aeta and Agta of the Philippines, thought to have reached them since the Holocene:
"Two mtDNA sets, M52'58 and M52a, that both originate from Indian populations were found in the Aetas of Zambales and the Agtas of Iriga in the Philippines."

Ancient DNA may shed more light on what actually happened, but it looks like Aeta Filipino and Indian populations both have some kind of P (Y) and M52 (mtDNA).

Simon_W said...

@Ariel

The Sardinian autosomal samples commonly used are exclusively from the interior of the island, I think. But there appears to be more variation on parts of the coast, for example a half-Sardinian / half North Italian genetic cousin of mine got very high CHG/Caucasus related scores in various tests, including the MDLP K23b. (I'm not speaking of Dodecad here.) So high that he cannot be explained as a 50-50 blend of the standard Sardinians with any one of the mainland Italian samples. His best approximations are with 1/4 Sardinian and 1/4 from the southern Balkans.

capra internetensis said...

@ak2014b

The P* reported in India is P1-M45(xQ, R), while the P* in Aeta is P2-B252, which is P-92R7(xP1). So the Indian P* is more closely related to Q and R than to the Filipino kind. Unfortunately none of the Indian P* has been properly investigated. It could even all be due to error, though it comes up too often for that to be likely IMO.

The Indian Q* and Q5 is likely to be Q1b-L275, which is pretty well represented in South Asia (the Y-Full tree shows this well). Q1 is Q1a1a1-M120, which is typically Chinese but widespread; as it was found in Kashmir it may have arrived via Tibet. Q4 is Q1a2-M346, which is everywhere, but there is an old branch of it, Q1a2b2-Y2659, which is found in Tamils.

As for the South Indian R* it might be R2*-M479 or various kinds of R1b, if not something more exotic.

MomOfZoha said...

@OpenGenomes:
I saw late that my previous too long and sarcastic comment has been deleted. I'm happy with that, but I wanted to point out exactly one thing that I hope will remain:

http://www.haaretz.com/israel-news/science/1.681385

"As Ostrer notes, citing the Biblical tale of Cain and Abel, “the fact that people are related to one another doesn’t prevent their developing extreme hostility to one another.”

Notice the awesome photo of the armed Israeli soldier -- supposedly ben Abel -- at the checkpoint shielding himself from the unarmed Palestinian "Cain", with a bunch of "double helix DNA wires" flying magically in the air. Here's the inspiring caption:
"Blood brothers: Palestinians and Jews share genetic roots, though so did Cain and Abel"

Now THAT is science.

(And, now, I shut up.)