search this blog

Monday, September 7, 2015

Copper & Bronze Age genomes from northern Spain


Update: the paper is now available and open access here.

See also: Yamnaya-related admixture in Bronze Age northern Iberia

...

A paper titled "Ancient genomes link early farmers from Atapuerca in Spain to modern-day Basques", by Günther et al., will appear shortly in PNAS early edition. Here's the press release from Uppsala University.

An international team led by researchers at Uppsala University reports a surprising discovery from the genomes of eight Iberian Stone-Age farmer remains. The analyses revealed that early Iberian farmers are the closest ancestors to modern-day Basques, in contrast previous hypotheses that linked Basques to earlier pre-farming groups.

The team could also demonstrate that farming was brought to Iberia by the same/similar groups that migrated to northern and central Europe and that the incoming farmers admixed with local, Iberian hunter-gather groups, a process that continued for at least 2 millennia.

The study is published today, ahead of print, in the leading scientific journal Proceedings of the National Academy of Sciences of the USA, PNAS.

Most of the previous studies about the transition from small and mobile hunter-gatherer groups to larger and sedentary farming populations have focused on central and northern Europe, however much less in known about how this major event unfolded in Iberia. This time, the research team investigated eight individuals associated with archaeological remains from farming cultures in the El Portalón cave from the well-known Anthropological site Atapuerca in northern Spain.

“The El Portalon cave is a fantastic site with amazing preservation of artefact material,” says Dr. Cristina Valdiosera of Uppsala University and La Trobe University, one of the lead authors.

“Every year we find human and animal bones and artifacts, including stone tools, ceramics, bone artefacts and metal objects, it is like a detailed book of the last 10,000 years, providing a wonderful understanding of this period. The preservation of organic remains is great and this has enabled us to study the genetic material complementing the archaeology,” Dr. Cristina Valdiosera continues.

From these individuals who lived 3,500-5,500 years ago, the authors generated the first genome-wide sequence data from Iberian ancient farmers and observed that these share a similar story to those of central and northern Europe. That is, they originate from a southern wave of expansion, and also admixed with local hunter-gatherer populations and spread agricultural practices through population expansions. The authors noticed that although they share these similarities with other European farmers, this early Iberian population has its own particularities.

“We show that the hunter-gatherer genetic component increases with time during several millennia, which means that later farmers were genetically more similar to hunter-gatherers than their forefathers who brought farming to Europe,” says Dr. Torsten Günther of Uppsala University and one of the lead authors.

“We also see that different farmers mixed with different hunter-gatherer groups across Europe, for example, Iberian farmers mixed with Iberian hunter-gatherers and Scandinavian farmers mixed with Scandinavian hunter-gatherers.” Dr. Cristina Valdiosera adds.

The study also reports that compared to all modern Spanish populations, the El Portalón individuals are genetically most similar to modern-day Basques. Basques have so far – based on their distinct culture, non-indo-European language, but also genetic make-up – been thought of as a population with a long continuity in the area, probably since more than 10,000 years ago.

“Our results show that the Basques trace their ancestry to early farming groups from Iberia, which contradicts previous views of them being a remnant population that trace their ancestry to Mesolithic hunter-gatherer groups,” says Prof. Mattias Jakobsson of Uppsala University, who headed the study.

“The difference between Basques and other Iberian groups is these latter ones show distinct features of admixture from the east and from north Africa.” he continues.

These findings shed light into the demographic processes taking place in Europe and Iberia during the last 5,000 years which highlights the unique opportunities gained from the collaborative work of archaeologists, anthropologists and geneticists in the analysis of ancient DNA.

“One of the great things about working with ancient DNA is that the data obtained is like opening a time capsule. Seeing the similarities between modern Basques and these early farmers directly tells us that Basques remained relatively isolated for the last 5,000 years but not much longer,” says Dr. Torsten Günther.

Source: Ancient genomes link early farmers to Basques

403 comments:

1 – 200 of 403   Newer›   Newest»
Alberto said...

This should be interesting if the dates of the 8 individuals sequenced span from 3500 BC to 1500 BC (but we'll see if that info is correct).

A bit surprising what they say about farmers mixing with local hunter-gatherers. Not in itself, because that's a rather obvious thing, but because they can only compare to La Braña in Iberia, and up until now all ancient and modern samples show less affinity to it than to Loschbour or KO1. I wonder if they have the higher quality sequence from La Braña that was said to be coming soon some months ago. Maybe the new sequence does show higher affinity to Iberian samples (and would contradict the tests from the recent paper about the Iberian Cardium sample). But let's wait to see the paper.

It does look suspiciously similar to this one:
http://www.sciencedirect.com/science/article/pii/S1040618215006631

Krefter said...

I like the article(http://phys.org/news/2015-09-ancient-genomes-link-early-farmers.html). Whoever wrote it took good notes and made sure he understood the authors of this new paper. Most news articles I've read on Ancient DNA had some wrong information.

@Ablerto,

Also, the article quotes someone saying Swedish farmers are closest to Swedish hunter gatherers. We know that isn't true.

"This should be interesting if the dates of the 8 individuals sequenced span from 3500 BC to 1500 BC (but we'll see if that info is correct)."

I'm most interested in the guy from 1500 BC. If he shows no ANE, maybe the Celts and other non-Celtic people brought ANE to Iberia.

Krefter said...

Hopefully some of the 8 are male so we can get Y DNA. Everyone would probably guess I2a and G2a. If R1b-L11 pops up before Beakers, I'll give Maju 100$.

la señora bibiloni said...

There was a paper on one sample from El Portalon in 2013 (Sverrisdottir) concluding that "The Neolithic Portalón individual is genetically most similar to southern Europeans, similar to a Scandinavian Neolithic farmer and the Tyrolean Iceman" (That's from Dienekes webpage). That paper said nothing about Basques, so I wonder if there is some kind of novelty in this last paper... Suspense!

Mike Thomas said...

@ krefter

"Hopefully some of the 8 are male so we can get Y DNA. Everyone would probably guess I2a and G2a. If R1b-L11 pops up before Beakers, I'll give Maju 100$."

Ha ha but yes, in all seriousness, this'll seal the deal

Maju said...

@Krefter: "Also, the article quotes someone saying Swedish farmers are closest to Swedish hunter gatherers. We know that isn't true".

Why do we "know" that? We don't "know" what you say we "know".

Actually I recall myself arguing recently in this blog's comments section for a possible SHG admixture all across the board based only on PCA data. There is nothing against that AFAIK. In fact Swedish farmers have not received the attention they deserve and are, excepted the sequencing studies from Uppsala University (also involved in this study), have been almost totally ignored as a marginal note. Only recently (within open discussions rather than being the focus of any study) we have (well, at least I have) began thinking that maybe there is a Portalón-Gokhem particular relevance in the making of modern Europeans and a connection among them of some sort... but everything was ill documented, mostly on PCA data.

This study probably does something more than just PCAs. At least I do expect so based precisely on claims such as the one you feel so skeptical about:

"We also see that different farmers mixed with different hunter-gatherer groups across Europe, for example, Iberian farmers mixed with Iberian hunter-gatherers and Scandinavian farmers mixed with Scandinavian hunter-gatherers."

That's a very revealing information bomb, and something that so far we were in the dark about except for some blurry PCA-based "intuitions".

And that's also something that may well explain excess "ANE" (at least in part) in NW Europe.

Krefter said...

Some updates are being posted here. One update says "Copper and Bronze age" were lactose intolerant. So, far there's a few Neolithic sites in Iberia, Sweden, and Germany with T-13910, and one in Yamnaya.

http://www.anthrogenica.com/showthread.php?1646-Genome-of-a-late-Neolithic-Iberian-farmer/page11

There's a title to the paper:
Günther et al. (2015) Ancient genomes link early farmers from Atapuerca in Spain to modern-day Basques, PNAS.
http://www.uu.se/en/media/press-releases/press-release/?id=2816&area=3,8&typ=pm&lang=en

Chad Rohlfsen said...

I do show EHG in the Swedish farmers, to a greater extent than Esperstedt or Baalberge. Spanish MN samples really lack it. As far as excess "ANE" in NW Europe, that's really not an issue here. It's very little EHG compared to modern Euros, in the MN samples. Way, way, way less. BTW, Basques are certainly not the same as the people from 5k years ago. That I promise. Even F3 stats agree. Spain_MN and EN plus Samara gives a better F3 and Z-score than WHG. I should have that calculator ready to go here in the next day or two. I'm about done trying to max WHG.

Krefter said...

Remember a Remedello sample in Northern Italy from around 1900 BC had I2a-M26 and a Neolithic autosomal-makeup. The same may have been true for the El Portalon sample(s) from 1500 BC. If so, this could mean ANE(Steppe derived) arrived in Iberia and Italy after 1500 BC.

We need genomes from France and the British Isles ranging 2500-1000 BC to know when they went from EEF/WHG to EEF/WHG/ANE. It could have been recent in those regions to, like after 1500 BC.

Mike Thomas said...

@ Chad

", Basques are certainly not the same as the people from 5k years ago. That I promise. Even F3 stats agree...."


Of course theyre not .. because "Basques" did not exist 5000 years ago. The formed around the Roman era , or later ...

Chad Rohlfsen said...

I'm talking genetically, not culturally.

rozenfag said...

@Krefter

Regarding males:

"The interred skeletons—four female and four male including a male child—were directly radiocarbon dated to between 5500 and 3500 years ago, a span bridging the Chalcolithic (Copper Age) period and the Bronze Age."

http://news.sciencemag.org/archaeology/2015/09/unusual-relic-language-comes-small-group-farmers-isolated-thousands-years

Krefter said...

@rozenfag,

Thanks for the info and link. Can't wait to see the Y DNA results. Some interesting info not in the other articles.

"Modern-day Basques do not show genetic mixing from these later migrants, whose genes contribute between 10% and 25% of the genomes of non-Basques living in the Iberian Peninsula of Spain and Portugal today. “The El Portalon individuals, very similar to the Basques, do not show this additional genetic material,” Jakobsson says."

"But today’s Basques turned out to be more closely related to the El Portalón farmers than to any other group in the study, including early hunter-gatherer genomes."

Mike Thomas said...

Chad
Yes I know :)
Are you putting up your calcs anywhere ?

Davidski said...

The sequence data is here...

http://www.ebi.ac.uk/ena/data/view/ERP010927

Matt said...

An El Portalon sample from 2000 BC looked close to Tuscans in a previous PCA. However seems possible some degree of compression linked to this, *or* these samples are earlier, and I think the suggestion is they are between 1500BC to 3500 BC.

Looking forward to see if these are closer to Basques than the La Mina samples from the Haak paper are. These should be a little younger. It will also be cool to see if these are the same samples who were identified as carrying early lactase persistence alleles, as that might inform our ideas on these, that perhaps the allele was moved around in early stages by long range population movements that didn't make much demographic impact on the whole autosome plus selection.

Re: regional HG admixture or not, the stats to test are whether the later farmers are particularly closer to the local farmers than the farmers who came before them. So you would go D(Spain_MN,Spain_EN;Loschbour,La Brana), for instance, or maybe D(Sweden_MN,LBK_EN;Swedish_NHG,Loschbour). Are they moved towards the local HGs compared to the farmers that came before them, even if they are in an absolute sense closer to, say, KO1? Of course, European Neolithic farmers moved around, so I don't expect strong regional connections to regional HGs.

For Sweden_MN, I did note in the "Pre- and Post-Kurgan Europe" comment thread that D(Ju_Hoan_North,SHG)(BedouinB,X) graphed against D(Ju_Hoan_North,WHG)(BedouinB,X) and D(Ju_Hoan_North,EHG)(BedouinB,X) indicate that Sweden_MN may be closer to SHG than would be predicted by its closeness to WHG and EHG, compared to other West Eurasian / European populations. So regional admixture has some feasibility from that.

andrew said...

The dates are Copper and Bronze Age dates, not early Neolithic as the lede suggests.

Aram said...

"Modern-day Basques do not show genetic mixing from these later migrants, whose genes contribute between 10% and 25% of the genomes of non-Basques living in the Iberian Peninsula of Spain and Portugal today. “The El Portalon individuals, very similar to the Basques, do not show this additional genetic material,” Jakobsson says."


Oh 10-25% of difference is significant between Basques and other non-Basque Iberians.
It can't be called an elite dominance difference. I don't know is there any definition for 'elite-dominance' but I think that some linguistic shift can be called an elite-dominance if the genetic difference is less than 7-10%.

Mike Thomas said...

I bet eventually we might realize that the difference of Basques resides in more recent phenomenal drift- their relative cultural iaoliation and endogamy. If this goes back to the iron age, then iberia might have been non- IE until the celts arrived (as per the traditional Halstatt-Celts model).

Krefter said...

I think the media may be miss interpreting what the authors said. Remember the media said Bronze age Armenians were *exactly* the same as modern Armenians and it turned out untrue, I think the same is true for El Portalon.

El Portalon are probably EEF with a lot of WHG and maybe shared drift with Basque. We've known Basque have more Neolithic/Mesolithic ancestry than other Iberians for a long time. I don't know where the "10% and 25%" post-Neolithic numbers in non-Basque Iberians and 0% in Basque comes from but it doesn't make much sense.

Basque have more than 0% that's for sure. I think Haak estimated around 25% Yamnaya in Basque.

Gioiello said...

@ Maju

Unfortunately, after that Anthrogenica banned me, they don’t permit me to read their posts, only the “Activity stream”. I am reading that in a post of an Italian (really an anti-Italians):
with your right map, we can easily envisage that ancient Tuscans and Basques are the "same tribe"..............do we have any similar markers?
I’d want to say to Maju that I am R1b1a2-L23-Z2110, with 94 private SNPs from my Full Genome of the FGC. I have 16 of these SNPs in common with Barnum/Winterowd but also with a French Basque. After smal (Sergey Malyshev) stole me 66 SNPs (I explained the possible reason), but the link with this Basque remains. I have written in the past that I thought that the Basque language (of Caucasian stock as Alfredo Trombetti demonstrated in Le origini della lingua basca, 1925) had been brought to Iberia from Sardinians, but it is possible that it had been brought by Tuscans. The same Trombetti thought that Etruscan were intermediate between Indo-European and Caucasian. Many linguists think that there were a link between Sardinian/Nuragic and Etruscan, for instance Massimo Pittau.

Simon_W said...

It's not really a surprise that Basques are not predominantly WHG, nobody who followed the last couple of years of research would have expected them to be. Also it's not really news that early Iberian farmers were very similar to other EEFs. EEFs are unlikely to offer any additional big surprises. But thanks god this apparently won't be just another paper about the earliest farmers and the Mesolithic-Neolithic transition. The time span from 5.5 - 3.5 kya is highly interesting and completely unknown as far as western/southwestern Europe is concerned. The yDNA results are definitely going to be fascinating, no matter what they will show.

Graham Little said...

Depends if they have ANE or not.

Karl_K said...

"Basque have more than 0% that's for sure. I think Haak estimated around 25% Yamnaya in Basque."
-Krefter

Mike Thomas said...

@ Simon
Yes, well put

Krefter said...

R1b1a2!!!! According to Geneticker. 2/3 are I2a2a(Already found in Megalithic Spain) and 1/2 R1b1a2.

https://genetiker.wordpress.com/2015/09/08/y-snp-calls-from-copper-and-bronze-age-spain/

Romulus said...

"If R1b-L11 pops up before Beakers, I'll give Maju 100$."

Looks like you might be giving Maju $100 lmao. Need date for the R1b sample.

Krefter said...

@Romulus,

I'm still confident in a Steppe-origin. I think the R1b1a2 guy has ANE ancestry. Wish we could test it for DF27 to see if it is ancestral to modern Iberian R1b.

Chad Rohlfsen said...

The 3500yo sample probably is R1b.

Mike,
My calc will be on Anthrogenica and maybe here.

Romulus said...

@Krefter

I consider myself origin-theory agnostic :). Strange that there is more I2a2 instead of I2a1a1 M26, consider the latter is what is commonly found in modern Basques. Same situation with the Iberian Haak samples...

bellbeakerblogger said...

@ Andrew
"The dates are Copper and Bronze Age dates, not early Neolithic as the lead suggests."

Yeah, I'd call that an oversell! The oldest individual is 3,500 B.C. which begins a problematic time in Iberia's history. So I don't know how they are so confidently making conclusions.

Anyway, if R1b1 is indeed among the results, this is probably the Iberian Beaker that has been rumored to be in the queue.

Alberto said...

The paper is up now. In it, the samples claimed to be R1b has no Y haplogroup assigned, though it's a male. But he's from 5,466-5,312 Cal YBP! And no ANE. Actually it seems not even the youngest sample (ATP9: 3700-3568 cal YBP) has any ANE from the admixture graph.

Shaikorth said...

The study itself didn't call any R1b SNP's in the samples

R1b-M269 call comes from Genetiker who was able to find a single SNP from ATP3 (genome coverage 0.03) associated with the clade. The SNP in question is PF6518, found also in many non-R1 haplogroups, including C-V20*, G-FGC5672 and J-CTS130.

Alberto said...

Well. actually it's hard to say. Basques also appear to have very slight or no ANE in that admixture graph, and in the PCA the most modern sample does seem to cluster with Spaniards, so slightly shifted east. Maybe she does have ANE (she's female, mtDNA U5b1b).

I hope David can get the samples soon.

Roy King said...

Also amazing is another H2-M282 in ATP2! H2 is very rare, found in Iran, Armenia, Cyprus and among Assyrians. A French HGDP sample is also H2.

Krefter said...

@Alberto,

That's shocking.

Lets wait for Davidski to analyses the data. Geneticker has shown a clear agenda for R1b to be of Mesolithic origin, so he may have falsified the results. If you've read his posts, he goes from calm to insane when R1b comes in subject.

Colin Welling said...

Our analysis using admixture graphs suggests different sources for the admixture into Iberian, Central European and Scandinavian farmers

The highest proportion of Mesolithic ancestry in the Portalón individuals seems to be related to central European hunter-gatherers (KO1, Loschbour) and not to the geographically close LaBrana

Interesting. So they have evidence for famers mixing with HG at multiple sites, resulting in differences between the farmers. However, the spanish farmers don't [appear] to lean towards brana and the scandinavian farmers don't lean towards motala.

Graham Little said...

K2-K15

http://www.pnas.org/content/suppl/2015/09/02/1509851112.DCSupplemental/pnas.1509851112.sd01.pdf

Matt said...

Interesting paper. Skim read it. Seems very pre-Haak and Allentoft! Very limited comparison as no analyses with EHG, LNBA.

There were a few mentions of differences in relatedness to present day Europeans between Ajvide and Motala populations. I wonder if that is preparing the grounds for the other paper described previously where they discuss differences between PWC and the Mesolithic population.

PCA analyses are seem pretty odd in the placing of the HGs (e.g. Fig S7). Procrustes fits. Be interesting to see what David makes of them, and also how they fit in terms of D(Ju_Hoan_North,*HG;BedouinB,Pop).

The tactic of using an African free ADMIXTURE to generate a stable North African cluster is also an interesting thing. Lack of membership of South Asian cluster in the El Portalon samples.

Maju said...

@Krefter: Thanks for the info. Big LOLs about that presumpt R1b-M269 (I knew some would show up sooner than later, there should be more by the Atlantic coasts, someone test ancient British farmers pls).

@Gioello: try cleaning up cookies. I have the same "problem" and I usually don't care about it (it's so petty on the side of Anthrogenica!), but in this case I was too intrigued, so I deleted the anthrogenica cookie, clicked on the link again and could read everything (as expected).

@Mike: ""Basques" did not exist 5000 years ago. The formed around the Roman era , or later ... "

WTF! How can you know? The background signal is clearly Neolithic/Chalcolithic with continuity. True that it's hard to take them apart from other pre-IE groups before Artenac culture and I have argued that Basque distinctiveness from other Vasconic groups probably only arose with the encroachment of Indoeuropeans c. 1300 BCE (Urnfield culture) and thereafter, which caused Basques and nearby peoples to become geographically detached from other ethnic relatives such as Ligurians, Iberians, etc. Much of the same happened surely to Ligurians around the Western Alps' backbone. Of course the "Basque miracle" (survival of the most rebel) is late Roman or early Medieval and was caused by the Bagauda or anti-feudal revolution, which seemingly only succeeded over here. Everybody else in Gaul and Hispania was definitely romanized... by Germanic rulers (ironically). But that's not the genesis, just a phase.

@Karl: In Alentoft 2015 (which is peer-reviewed, unlike Haak's pre-pub) Basques have zero Yamna signal and in general the Yamna signal is much much weaker and clinal than in Haak's. I would rather not cling to any particular study and results but when two papers contradict each other so much, one is obviously doing something wrong. And judging on a wider, holistic, view of diverse data, that looks like it is Haak's.

@BellBeakerBlogger: 3500 BCE cannot be Bell Beaker: it is way too old. Oldest Beakers (south only) are from c. 2800 BCE!

Krefter said...

@Maju,

R1b may have been in Britain during this time but there's little room for it to be ancestral to modern British R1b. The British Isles were pretty much repopulated in the Bronze age. No matter what the ultimate origin of R1b is, their R1b-L21 probably arrived in the Bronze age.

Krefter said...

@Maju,

Basque have ANE and Steppe ancestry, no doubt.

Shaikorth said...

Maju, Haak et al. has been peer-reviewed and published for a long time, and Allentoft results do not contradict it re: Basque Yamnaya. The method is based on formal stats, not some ADMIXTURE run.

Maju said...

Some further comment:

It seems that all these samples were lactose intolerant (re. T-13910) and modern Basques are almost universally lactose tolerant. So in my understanding these people are not quite proto-Basques but rather a Westerly "Neo-European-2" group that is related to Basques (and proto-Basques, whatever that means) and other Western or "Atlantic" farmers (for which we only have very few samples so far, mostly Gökhem). I must recall here the most important issue of two populations (relative to T-13910) co-existing in the "Basque" Upper Ebro in the Chalcolithic, being almost unmixed yet. IMO this means that only some of the Western "Neo-European-2" peoples, probably the "most Atlantic" ones had the T-13910 allele, which may be a founder effect of some sort or an introgression from Western Paleo-Europeans (who would have it fixated by pure randomness).

A previous sample of the same site and chronology (c. 4000 BCE) was studied by Dasakali 2014 (doctoral thesis) and in her PCA it plotted near modern Tuscans (while Gökhem overlapped modern Basques, as they did in the almost simultaneous Skoglund & Mälstrom 2014 study). Why these differences with other analyses (mostly from the German academy)? I think that it is because their PCAs emphasized European-specific tendencies rather than letting West Asians influence them too much. The samples used by Dasakali and Skoglund were very different: Dasakali used an almost Europe-only sample (with only a few Levantine and North Caucasians, who still allowed the "European cross" pattern to form) while Skoglund used instead a much larger West Asian sample but also a huge Swedish one, which indirectly minimized the spread of the West Asians along the PC2 axis. So I strongly suspect that some of the contradictions we are seeing in all these studies derive from using too many West Asians in the PCAs (and maybe also other statistical analyses) and not focusing enough on the European specifics, which should provide much improved precision in the relationships between ancient samples and modern ones. After all, we are trying to figure out Europe, not West Asia (not yet, for lack of local ancient samples).

Maju said...

@Krefter: I strongly disagree. I see no clear evidence of the British islands (except maybe Ireland in the Bell Beaker period?) being "repopulated" at all. There is at least one striking piece of evidence, for lack of sufficient aDNA, which is that early British farmers shunned fish and favored instead meat and milk (lactose tolerance likely therefore) and that this pattern of food preferences remained the same all the way to the Danelaw, when Vikings reintroduced the liking for fish. You could argue, I guess that all those imaginary invaders, had exactly the same culinary preferences as the early British farmers (would be MN/Chalcolithic from a pan-European chronology) but it looks a bit unlikely.

Anyhow, your claims clash with the data of Alentoft, whose ADMIXTURE analysis is very deep (K=15 to K=19) and gives very strong relevance to ancient samples and their peculiarities. Corded Ware, Yamna and other late Chalcolithic/early Bronze samples that are purportedly Indoeuropean, show various shades of teal. The strongest one across the board (dark teal, looking a Caucasus component) is variedly found also among EEF Neolithic samples (Hungary, Stuttgart) but not in Gokhem (nor among Basques). This "Caucasus" component may be a confounding factor.

The other lighter teal components found in Yamna, CW, etc. look pretty much "Siberian" and could probably be assimilated (roughly) to your mystified "ANE" thingy. These are very weak, almost nil at most K values, in modern Europeans (Russians excepted). K=18 would be a partial exception but this one would also make Naqab Bedouins partly "Siberian", so disregard.

The "Kurgan crowd" is totally ignoring this key study and clinging to Haak's instead (which is so far only a "draft"). I won't say that Alentoft's data is the last word either but it is clearly a peer-reviewed contradiction to claims based on Haak's pre-pub.

We have to understand that autosomal analysis is not "rocket science" but rather closer to sociology, opinion polls, market studies, etc. And I say that because the tools are all statistical analysis and statistical analysis is not an exact science. When various opinion polls contradict each other, we should be wary to make predictions about what the result of the elections will be, right? Instead if they all agree pretty much, then the prediction is easy and probably quite accurate. Here we are in the first case: contradictory data suggesting confounding factors and quite probably systematic bias. Problems that must be addressed seriously and not in a militant dogmatic and cherry-picking way.

Maju said...

@Saikorth: "Maju, Haak et al. has been peer-reviewed and published for a long time".

Really? I thought it was still in the pre-pub phase. That's the only link I have and there is no notice whatsoever of it having been formally published yet. In any case the issue of contradictions with Alentoft (and I would dare say internally) stands. Personally I don't think Haak 2015 is at all at the level of quality of Lazaridis 2014, there's a lot of info missing from the supp. materials and there is not even a Europe-only PCA in all the study.

Maju said...

@Saikorth: "Allentoft results do not contradict it re: Basque Yamnaya. The method is based on formal stats, not some ADMIXTURE run".

Can you direct me to the exact location of those formal stats. I have forgotten...

Simon_W said...

PF6518 is not on the ISOGG site. I suppose this means that it's not considered a reliable marker? Genetiker is probably an R1b male, though I don't want to jump to conclusions there.

bellbeakerblogger said...

@ Maju

"3500 BCE cannot be Bell Beaker: it is way too old. Oldest Beakers (south only) are from c. 2800 BCE!"

To clarify for for the benefit of others, I do not suggest a Beaker identity for anyone prior to 2800 B.C. (I assumed the presumptive R1b to be from a more recent individual)

I understand your comment through my assertion that R1b-P312 spread in Western Europe via an out-of-Iberia migration, which remains to be seen.

Simon_W said...

What the paper tells is that there is no R1b in two males living in El Portalon between 3000 and 2600 BC. This is quite a crucial period. At this time the earliest Iberian Bell Beakers occured and started to expand. Of course this doesn't prove that R1b-L11 is from Yamnaya. Maybe it was in El Portalon, still in the minority back then, maybe just at around 33% frequency. Or maybe it was only in other parts of Western Europe, in Britain or in Portugal, or whatever. Or maybe not. At least, these results are perfectly in line with my expectations!

And the ADMIXTURE analysis shows that even as late as inbetween 1700 and 1568 BC there was a female there with no Yamnaya admixture. Still, it's kind of borderline. Maybe the invaders had just arrived back then, in form of males. Or maybe they arrived around 1600 BC and the female lived shortly before this. But still, essentially the evidence suggests that the ANE/LNBA/central European admixture in Iberia arrived after the middle Bronze Age.

Open Genomes said...

"Matajo" (ATP12-1420) is Y-DNA I2a2a-PF3858* CTS9183-, but RISE552 Y I2a2a1b1b2-S12195 Yamnaya Ulan IV Russia , RISE247 Y I2a2a-L368 Vatya Szazhalombatta Hungary, and RISE479 Y I2a2a1a2a2-SK1247/Y4915 Vatya Erd 4 Hungary are all I2a2a1-CTS9183+. Why are the First Farmers of Spain seemingly *earlier* in the Y phylogeny than the Hungarian Vatya and Yamnaya samples? Was there a "Vasconic" influence among the Proto-Indo-Europeans of Yamnaya?

UCSC Browser showing I2a2a1-CTS9183 for "Matajo", RISE552 Yamnaya, and RISE247 + RISE479 Vatya

Chad Rohlfsen said...

In other words, the claim of one being R1b looks dubious?

Shaikorth said...

Maju, Haak et al. published version is here:

http://www.nature.com/nature/journal/v522/n7555/full/nature14317.html

The methodology of their fits is explained in Supplementary Information 9 (page 94), and supplementary S9.27 on page 125 has the best (lowest resnorm) fits.


I don't think there is much difference in quality between Lazaridis and Haak papers. There isn't an European PCA but we already can figure from Lazaridis et al. what it would look like when done on the Human Origins set. Anyway, back then I did ask David to make a PCA with more Europeans but no West Asia/MENA, while adding Eastern HG:
https://drive.google.com/file/d/0B9o3EYTdM8lQNXlTSDIxS0NuN2M/view

Gioiello said...

MY VICTORY IS COMPLETE NOW

The paper of Guenther et al, published to-day, about the samples of aDNA from Portalon (Spain), not only confirms the previous mtDNA I linked with Italy (K1a2 (b in this case) and X2c), but, beyond the notorious Western hunter-gatherers U5b1 and U5a1 and the unequivocal European H3, also the mt haplogroup U5b3, which was the only one that Italian researchers demonstrated some years ago being born in Italy, when I said that they were too much shy, because I claimed to Italy also all the others.

Maju said...

@Aram, Mike: From other data (Matt just shared these graphs in another thread, for example: clusters, tree, but there are other reasons, including data produced by Davidski), my growing impression is that there is an important pre-Kurgan divide between "first farmers" (EEF, EN) and "Atlantic farmers" (Gokhem, Spain_MN, one of the Spain_EN samples even) that is actually the main structure shaping modern Europe, regardless of whatever Kurgan impact. Hence I'm likely to disagree (awaiting for the data anyhow) with the assessment that the Basque-Iberian distinction is only due to extra input in the latter because that "extra input" looks nothing but extra EEF affinity (and mostly not kurgan stuff) and that EEF stuff was already in Iberia since the very arrival of Cardium Pottery.

However the EEF affinity seems not just structured in time but also in space. Iberia is a very large region spanning more than 1000 km in any direction (with rugged terrain, different ecosystems...) and archaeologically we can also distinguish some gradations and differences, whose genetic implications are hard to gauge without enough samples (same for France). Atapuerca is far to the West compared with most Iberian EEF samples and therefore it makes sense that they were more admixed with local Paleo-Europeans and fall in the Atlantic Farmer or "Neo-European 2" cluster instead, much like Gokhem and modern Basques.

As we know (or we should), this Atlantic Farmer cluster very dynamically expanded in the Chalcolithic before the main Kurgan immigration took place in Central Europe with Corded Ware. We know that Basques and North French show modern-like mtDNA pools since at least those Chalcolithic times. We don't have yet all the data but the one we have is very strongly supportive of an Atlantic Neolithic distinct cluster setting the basis of the distinctions of Basques vs Spanish, French vs Italians and English vs Bulgarians. The general theme is exactly the same for all these pairs. We even can see in at least some PCAs a relatively thick PC1 "empty zone" buffering between the Atlantic and the EEF (now "Mediterranean") strips.

It is indeed possible that EEF-loaded populations expanded in Iberia at some point between the Chalcolithic and the Iron Age, reinforcing those distinctions but so far it's impossible to gauge the timing of that re-expansions. It could be even partly with the Celts, judging on the little we know about ancient Portuguese mtDNA, and, if so, it could explain why Kurgan and EEF autosomal genetics go hand on hand in Iberia (not clear enough yet anyhow).

Krefter said...

@Maju,

A shared trait doesn't equal genetic continuation. One ADMIXTURE from Haak isn't good evidence, because Davidski, Chad, and others have ran Haak genomes through many ADMIXTURE tests. Genome bloggers have done much more work than Haak or Allentoft. They have the same tools and genomes. Evidence from blogs is greater.

Anyways, academics supports what genomes bloggers say. Allentoft didn't go in as much depth as Haak did. Table S9 in Haak models Northern Europeans(including, from British Isles) as about 100% Bell Beaker/Unetice. Davidski's analysis have supported this. Such people didn't live in the British isles in the Neolithic. It makes sense such people migrated there and repopulated the Isles.

With Allentofy we can see the same happened in Sweden. In 3000 BC we see Gok2, in 2800 BC we see Corded Ware, in 2000 BC we see people there like modern Swedes. Gok4 had Y DNA I2a1b1, Corded ware+2000 BC guys had R1a-Z284/R1b-U106/I1 the main lineages of modern Swedes. In Poland, Lithuania, Germany we see a similar processes occurred.

Maju, i don't understand why you're stubbornly against huge movements of people from the East. SHG, WHG, EEF can't explain what we see today. Laz 2014 discovered this, and that's why they went to Yamnaya and Corded Ware to find the source of ANE in Europeans.


Gioiello said...

The U5b3 from Portalon is U5b3f. These are the samples on “Ian Logan spreadsheet”:
U5b3f G16129A

65. GQ129179(USA) Pala U5b3f 23-JUN-2009 A73G C150T G228A A263G 309.1C 315.1C A750G A1438G A2706G T3197C A4769G C7028T G7226A A7768G A8860G G9477A A11467G G11719A A12308G G12372A T13617C T14182C C14766T A15326G G16129A C16192T C16270T T16304C
66. GQ129180(Italy) Pala U5b3f 23-JUN-2009 A73G C150T G228A A263G 309.1C 315.1C A750G A1438G A2706G T3197C A4769G C7028T G7226A A7768G T8469G A8860G G9477A A11467G G11719A A12308G G12372A T13617C T14182C C14766T A15326G G16129A C16192T C16270T T16304C
67. GQ129181(Italy) Pala U5b3f 23-JUN-2009 A73G C150T A263G 309.1C 309.2C 315.1C A750G A1438G A2706G T3197C A4769G C7028T G7226A A7768G A8860G G9477A T10493C A11467G G11719A A12308G G12372A T13617C T14182C C14766T A15326G G16129A C16192T C16270T T16304C
68. GQ129182(Spain) Pala U5b3f 23-JUN-2009 A73G C150T A263G 309.1C 315.1C A750G A1438G A2706G T3197C A4769G C7028T G7226A A7768G A8860G G9477A A11467G G11719A A12308G G12372A T13617C T14182C C14766T A15326G G16129A C16192T C16270T T16304C
69. GQ129183(Italy) Pala U5b3f 23-JUN-2009 A73G C150T G228A A263G 309.1C 315.1C A750G A1438G A2706G T3197C A4769G C7028T G7226A A7768G A8860G G9477A A11467G G11719A A12308G G12372A T13617C T14182C C14766T A15326G G16129A C16192T C16270T T16304C

Gioiello said...

The sample from Portalon has 16129G, thus it is the original U5b3*. Also better.

Krefter said...

@Gioiello,

Thanks for pointing the U5b3f out. The only examples of U5b3f in my data come from Spain and Italy.

@Everyone else,

A single R1b1a2 from Spain some 5300 years ago, doesn't take away all other evidence for a Steppe origin in West Europe. A theory of local origin has too many holes in it.

I'm waiting for Davidski and Felix to analysis ATP3's Y DNA, to legitimatize Geneticker's work.

Anyways, ATP3 adds to the mysterious history of R1b-L151.

la señora bibiloni said...

R1b-M269?! WHAT?! I'm laughing out loud. Completely unexpected. Still, let's wait for confirmation.

Krefter said...

Also. All other Y DNA from El Portalon is under typical Neolithic clades unheard of today. There was some-type of expansion of R1b-P312 into Central Europe by at least the Bell Beaker period which was 1,000 years later, that is directly related to P312's dominance today. In-between 3500 BC and 2500 BC something happened.

Colin Welling said...

didn't the study throw out some of their results due to contamination concerns?

Mike Thomas said...

Krefter
What's yr email ?

Maju said...

@Saikorth: Thank you very much. What I notice is that the resnorm improvement is very small and anyhow subject to the a-priori first analysis, which is {EHG,EN,WHG}. There is no attempt to use for example {MN,Yamna}, let alone something specific as Gokhem or alternative "HG" influences like SHG, which I suspect important in Northern Europe. So I stand on the previous contention: statistical analysis it's not "rocket science", and different assumptions, sampling strategies, choices of component populations, may heavily slant the results. As there is no rocket that will fall if you do something wrong, evidencing the error dramatically, it's sometimes very difficult to nail it... unless you try really really hard.

The EHG Europe-only PCA is not quite what I wanted but it clearly shows three stripes of high, low and null EHG influence along the horizontal axis. Now: is it EHG or is it SHG or something in between? This is not any "burning nail", if you scroll around in the supp. materials, you continuously see Motala_HG scoring high everywhere. For example, in fig. S9.5, Baalberge can be seen as {83% LBK + 17% WHG} (optimal) but also very good fits are {89% LBK + 11% SHG} or {88% LBK + 12% Yamna}, or also {91% Hungary_EN + 9% SHG/Yamna}. In this particular case at least Motala (SHG) and Yamna can be exchanged without problem, with marginal preference for the former.

If you stick to the minimal resnorm, you have to accept a Spain_EN migration to Esperstedt, Corded Ware and German Bell Beakers, and simultaneously a reverse LBK_EN migration to Atapuerca, what does not make any sense. Unetice people themselves would be made up of Spain_MN (i.e. El Portalón people discussed here) and Motala_HG with no influence from Yamna or LBK or local MN whatsoever. That's not quite acceptable, really. So there has to be another less optimally fitting result that is however closer to reality, in this last case it could be Baalberge+Motala, Esperstedt+Yamnaya, Esperstedt+Motala, etc. (again we see that often Motala and Yamna are mutually exchangeable often).

So careful...

Maju said...

@Krefter: "Maju, i don't understand why you're stubbornly against huge movements of people from the East".

Because they just do not fit. You are just now claiming a blank slate 100% replacement of the population... I can laugh or cry but I cannot accept it. In fact Haak's data does not support that in any solid way anyhow: it's an artifact.

PF said...

@Krefter,

Where can we find your aforementioned blog? Interested in your analysis of mtdna H (in general, and of course partially for personal reasons, as my dad seems to be H1as2). Thanks!

Krefter said...

@PF, Mike

Blog: H1+H3+H5 are the main forms of H in Europe. All look like Neolithic-markers(based on ancient mtDNA). H1 and H3 are very diverse. High frequency doesn't seem like a recent thing. Although it's totally possible some is of pre-Neolithic origin.
mtdnaatlas.blogspot.com

Email:
sammyisaac107@gmail.com

@Maju,
"Because they just do not fit. You are just now claiming a blank slate 100% replacement of the population... I can laugh or cry but I cannot accept it. In fact Haak's data does not support that in any solid way anyhow: it's an artifact."

100% replacement is crazy. Although something like 80% replacement is very reasonable.

Krefter said...

@Maju,

Just wondering. Can you speak Basque fluently?

Mike Thomas said...

Maju

I'm not sure why you believe that Atlantic and Scandinavian farmers are being excluded ? I've seen your take on the PCAs, with your added vector lines - from EEF to EHG, etc.

But its quite simple, IMO. Bronze Age Europeans are a fusion of MNE and CWC-types. In turn, MNE is the result of final assimilation of EEFs and residual WHG-foragers. I don't think its from an as yet unaccounted -for "Atlantic' mystery population.

Shaikorth said...

The resnorm differences are minimal for some fits in those two-way cases for ancient genomes, but more visible with modern populations as the number of populations is increased (until 4/5) and that's where I'm focusing. The ancient samples have varying degrees of quality and so on which may account for minor differences between the fits so Spain_EN giving slightly better fit than some other EN pop (they are almost identical anyway) in some situation is not that important, the main thing is that EN+WHG+Yamnaya is better than EN+WHG or EN+SHG/Motala for modern populations.

As for the PCA, perhaps you wanted to see Basques define dimension 2 more clearly? That happens if there are less Northwest Euros relative to Basques, like in the HO set, but here a methodological issue with PCA comes into play. Basques are very homogenous and drifted, and the more of them there are relative to other samples, the greater the chance that they will take over the PCA dimension. Some Northeast Italian villagers can be much more distinctive than Basques or even Sardinians on an European PCA if there's a comparable amount of them, even with West Asians included.

Chad Rohlfsen said...

result: SwedenSkoglund_MN WHG Basque 0.008738 0.001499 5.830 161187
result: SwedenSkoglund_MN Spain_EN Basque 0.013948 0.001198 11.646 160201
result: SwedenSkoglund_MN Motala_HG Basque 0.007753 0.001333 5.817 159670
result: SwedenSkoglund_MN Spain_MN Basque 0.013066 0.001247 10.478 159402
result: SwedenSkoglund_MN Samara_HG Basque -0.001006 0.002239 -0.449 97520
result: SwedenSkoglund_MN Yamnaya Basque 0.001189 0.001079 1.102 160882
result: Spain_MN WHG Basque 0.009148 0.000874 10.468 230610
result: Spain_MN Motala_HG Basque 0.009341 0.000720 12.975 228161
result: Spain_MN Samara_HG Basque 0.001568 0.001258 1.246 136931
result: Spain_MN Yamnaya Basque 0.003070 0.000645 4.761 230363
result: Spain_MN Georgian Basque 0.003777 0.000478 7.908 230905
result: Spain_MN Balochi Basque 0.002087 0.000427 4.886 231006
result: Spain_MN Tajik_Pomiri Basque 0.001351 0.000476 2.841 230852
result: SwedenSkoglund_MN WHG Basque 0.008738 0.001499 5.830 161187
result: SwedenSkoglund_MN Samara_HG Basque -0.001006 0.002239 -0.449 97520
result: SwedenSkoglund_MN Yamnaya Basque 0.001189 0.001079 1.102 160882
result: SwedenSkoglund_MN Georgian Basque 0.003686 0.000712 5.175 161301
result: SwedenSkoglund_MN Balochi Basque 0.000608 0.000710 0.857 161387
result: SwedenSkoglund_MN Tajik_Pomiri Basque 0.002239 0.000796 2.811 161303
result: Spain_EN WHG Basque -0.000438 0.000868 -0.505 233658
result: Spain_EN Motala_HG Basque 0.001500 0.000749 2.004 230344
result: Spain_EN Samara_HG Basque -0.004272 0.001298 -3.292 137906
result: Spain_EN Yamnaya Basque 0.001538 0.000619 2.484 233333
result: Spain_EN HungaryGamba_EN Basque 0.015076 0.000700 21.539 233715
result: Spain_MN HungaryGamba_EN Basque 0.010772 0.000742 14.522 230626
result: Spain_MN HungaryGamba_CA Basque 0.007986 0.001340 5.961 141998

Samara HG plus Spain_EN and MN, is about all that works. Count out another farmer component, Motala, or other MN pops.

Open Genomes said...

ATP2 is in Y haplogroup H2b1-Z19051+. While H2-M282 was very common among the First Farmers of Hungary and Germany, today it's found among Sardinians in Francalacci (2015), a few Central Europeans and people from the British Isles, and also Assyrians, one Armenian, in Kuwait, the United Arab Emirates, and Bahrain. One place it is *not* found is in Iberia or among Basques. How could Y H2b1-Z19051 be the origin of the Basques, when it's totally absent from the Basque region and Iberia today? How did it die out and get replaced by R1b1a2a1b-P312?

UCSC Browser showing AT2 H2b1-Z19051+

Romulus said...

In Genetiker's K17 analysis ATP3 (the r1b genome) looks very different from the I2a2 genomes, especially Matojo. ATP3 seems more Near Eastern while Matojo scores higher in HG/Farmer components. That would obviously support the r1b haplotype, it will be interesting to see the H2 admixture.

CroMagnon said...

Why doesn't the official paper report the R1b result ? Contamination ? Poor cover ?

Maju said...

The paper is up in any case, so time to read it in full (open access):

→ http://www.pnas.org/content/early/2015/09/02/1509851112

My first impression is that, while the ATP samples are about the most Basque-like in the known aDNA, they are still not quite Basque enough. Rather they seem equidistant between Basques and Sardinians. They clearly have excess WHG admixture (La Braña in one tree, proto-Lochsbour in the other), while Gokhem shows proto-Ajvide admixture (and not Motala).

Notably their mtDNA pool is still way too pre-modern (low H and U), what contrasts with the Neolithic Basque and Portuguese Neolithic ones, suggesting that we still require more specifically "Atlantic" samples to understand what is going on. Similarly they were all lactose intolerant, unlike modern Basques and at least some Chalcolithic Basques, and Chalcolithic Swedes.

The Bronze Age woman shows again Basque-Sardinian intermediate affinities but in the supplemental data it is apparent that she also shows strong British affinities, notably with Cornwall and Scotland.

Someone asked (not sure if here) for a PCA with North African samples. Well, there is one and it only shows a huge gap between North Africans and Europeans, reorganizing European samples in a single line between West Asia and Paleoeuropeans that could be of some interest in order to perceive the most basic gradation. But PCAs are not the most informative piece of data here in any case.

Maju said...

@Cro-Magnon: many samples have very low coverage.

la señora bibiloni said...

Is it possible to compare ATP3 with the confirmed M-269 we have (Kromsdorf Bell Beaker)? Would it bring valuable information? Just wondering

Krefter said...

@Open Genomes,

Geneticker listed all the Y SNP calls he got for ATP3.

https://genetiker.wordpress.com/y-snp-calls-for-atp3/

Krefter said...

@Everyone,

We have 3 examples of pre-Beaker R1b in Central-West Europe:
R1b1a2 Spain 3516–3362 BC
R1b Hungary 2860-2620 BC
R1b1 Poland 2865-2578 BC

We have *40* Y DNA samples from Spain, France, Germany, Hungary, and Sweden from 3000-3600 BC. All is I2a, G2a, E1b-M78, F*, and H2, except for 2. The 2 are R1(xR1a1a, R1b1a) from Germany and R1b1a2 from Spain.

R1b-P312 certainly hadn't made it's expansion in Atlantic Europe by 3500 BC or even 3000 BC(considering what we see in Trellis France). The R1b1a2 in Spain from 3500 BC was certainly a lone wolf.

The first time we see popularity of R1b is with Bell beaker in Germany and Czech around 2600-2200 BC. We know for a fact it is related to the popularity of R1b-P312 today.

So, somewhere between 3500 and 2200 BC there was an expansion of R1b-P312 which is one way or another the ancestor of modern R1b-P312.

Another thing: Saying R1b1a2 in Spain in 3500 BC is prove R1b-L11 has a local origin is nonsense. We have plenty of R1b1a2 in Russia from around the same time period. Who's to say this R1b1a2 in Spain isn't a brother or cousin to R1b-L11 like Yamnaya's R1b was?

We need more Ancient Y DNA. Data from Spain, France, Germany, Balkans, Ukraine ranging 3500 BC-2800 BC will probably find the ancestor of R1b-L11.

Harvard is analyzing the 2800 BC R1b from Hungary. That'll hopefully give us a clue.

Maju said...

Oh, c'mon, Krefter! Treilles (that's the correct spelling) is not by any means in Atlantic Europe but clearly in the Mediterranean (look up "Languedoc" in Wikipedia). I know that the divide gets a bit narrow, complicated and with overlaps over here but anyhow.

Similarly Atapuerca is also pretty much Mediterranean, although it sits on the water basin divide. At the very least it is in the Mediterranean side of the climatic divide. And anyone familiar with local archaeology and old-fashioned anthropometry knows that "the Mediterraneans" came up the Ebro river and formed a mixed population in the Upper part of it while just a bit north of them a proto-Basque ("Pyrenean" type, would be similar to "Keltid"?) population lived.

I'm less familiar with the details north of the arbitrary border but in general terms there is also a first impact coastal area in Provence and Languedoc, a secondary influence area up the Rhone and around Toulouse and an intriguingly novel groups spawning towards the Atlantic, near Bourdeaux and Brittany. And also La Hoguette culture in North France and the Rhine, and other mysterious novel, presumably more mixed groups in Belgium (Limbourg pottery).

All that, as well as other groups like the first dolmen builders of Portugal and what-not are all very intriguing cultures or groups that were probably more mixed with Paleo-Europeans but until we sequence them we won't know. Right now all we have autosome-wise is Gokhem and Atapuerca. It's something but clearly not enough. And definitely not for the clarifying of Y-DNA.

I would beg you, Krefter, to be less dogmatic and also more patient. Rome was not built in one day. If you are as young as imagine, you should have plenty of time in your life to participate of this discovering process, the excitement of which we do share (but little more).

Gioiello said...

@ Maju
Unfortunately I am working with an old PC. I’ll buy another more powerful next. I thank you for your science and for what you said about the autosome. I followed the Y and the mt, and very likely I arrived where no one did. You know that I didn’t agree with you only because I said that what it is in Iberia came before from Italy, this is the meaning of my “Italian Refugium”, and about the Bell Beakers I am waiting that more data will be added from aDNA from Tyrrhenian Italy.

@ Open Genome
Sardinia in Francalacci 2015 has 7 E2-M282 labelled as F3 (samples 139-145), and not only the mt-s of Portalòn , as I think having demonstrated before, but also the T very likely came from Italy.

Gioiello said...

Of course the Y.

Gioiello said...

@ Open Genomes
•H2 P96 (14869743 C->A)
• •H2a L280 (14869743 C->A) FTDNA subgroup [removed by Family Tree DNA prob in early '15]
• •H2b Z19065 (2716568 G->T)
• • •H2b1 Z19045 (2681998 G->A)
• • • •H2b1a Z19039 (16710674 G->T)
• • • • •H2b1a1 Z19033 (6937005 G->A) Sardinians
Thisis your Z19051 (7377286 C->A) and are the samples 141-145,

The H2b1/Z19051 of Portalòn came from Sardinia, where the hg. is old of many thousands of years.

Open Genomes said...

The case for ATP3 being R1b1a2-M269 is as follows:
GHIJK-M3773/CTS12673 => K-CTS9278/PF5501/M2693 => R1-M748/YSC0000207 => R1b1a-L389-P297:FGC46/Y97 => R1b1a2-M269:PF6518
PF6518 [Y:23,099,729 A>C] is a transversion, and it's not a C>T or G>A which could be the result of DNA damage. Therefore, this could be a case for some kind of R1b1a2-M269 being present in the Pre-Bell Beaker/Chalcolithic 5526-5372 ybp (calBP + 60 years):

UCSC Browser ATP3 PF6518+ [Y:23,099,729 A>C] Pre-Bell Beaker/Chalcolithic 5526-5372 ybp (calBP + 60 years)

Open Genomes said...

YFull has the tMRCA for R1b1a2-M269* at 6132 ybp, and R1b1a2a-L23 at 5908 ybp, both old enough for ATP3 at 5526-5372 ybp.
R-P312, which is typical of Western Europe has a tMRCA of only 4600 ybp, so it would be ruled out, and RISE98 from the Battle Axe/ Nordic Late Neolithic Culture in Sweden who is R-U106* was dated to 4285-4042 ybp.
It doesn't seem to be likely that R-U106 originated in Iberia in the Neolithic and migrated to Sweden to become part of the Battle Axe Culture, and R-U152 which was found in the Bell Beaker Culture in Germany at 4500 ybp is also too young.
If ATP3 is in fact some kind of R1b1a2-M269*, it's likely in one of the early non-R1b1a2a2-Z2103 branches, like we find in the Yamnaya Culture which arose shortly afterward, around 5300 ybp.

http://www.yfull.com/tree/R-M269/

Gioiello said...

@ Open Genomes

I thank you for your analysis, thus also this R1b1a2-M269* came very likely from Italy, perhaps it too from Sardinia where it is well known and diffused and one haplotype of them I found in my acquired cousin Fabrizio Fdereighi from Santa Maria a Monte (Pisa) through a Geno 2.0 as you can see in the smal's tree c/o "R1b1a2 (P312- U106-) DNA Project". This would be the definitive proof of my "Italian Refugium".

Gioiello said...

Of course the name is Fabrizio Federighi. I didn't believe when 23andMe found in me one mutation of dyslexia, ahahahahahah

Gioiello said...

@ Open Genomes

Very likely the YFull dates are underestimnated of 1000/1500 years as my R-Z2110 tested from FGC (Full Genomes) demonstrates, in fact it is old about 7100 years and not 6200, and this very likely will be the truth.

Gioiello said...

Big Y has on average 12000 no calls out of 72000 SNPs. I had only about 300, and all the Big Y, which should be named Little Y, lowered the dates.

Maju said...

@OpenGenomes: "It doesn't seem to be likely that R-U106 originated in Iberia in the Neolithic and migrated to Sweden to become part of the Battle Axe Culture"...

U106 has nothing to do with Iberia or Southern Europe in any way. U106 is a Northern-specific "brother" haplogroup of S116/P312, which most likely also did not originate in Iberia either but rather in "France" (Occitania, Aquitaine or somewhere in that zone).

I don't know what's up with this obsession with "Iberia" and this lack of interest on France. It's like France does not exist at all, when historically and paleohistorically is no doubt one of the most important regions of Europe. And all just because we have no aDNA data of relevance from that huge state. In any case, France or Iberia, U106 cannot originate so far South. However S116/P312 is a completely different story.

Both share a common ancestor but we cannot even pinpoint where that ancestor lived: somewhere between the Balcans and the Atlantic Ocean, that's all we can say (with good candidates in the Mid-Danube, North Italy and Iberia judging on paragroup distribution). And it's not likely that we will find him via aDNA (logically). So I'm not really interested on R1b in general nor R1b-M269 even. I just assume, judging on modern data, that they were there like a thin trickle among other lineages in the Neolithic flow (unless a bottleneck and older chronologies are appealed to). They will show up here and there and that's not too useful on its own. The issue is when and where does P312 (and U106 too, but not necessarily related in time or culture) show up. The only dates we have so far are terminus ante quem, which correspond to roughly 2200 BCE in the case of U106 and not much earlier for P312. However while I guess one could argue for U106 relation with Corded Ware (OK, it overlaps somewhat, if we ignore Britain) it is impossible to argue for such relation with P312, for two simple reasons that are one: the Bell Beaker phenomenon flowed from Southwest to the North (and NE) and so did R1b-P312 (at least North of the Pyrenees).

Most probably both lineages expanded in relation with previous processes within Megalithism, which are clearly related to important demographic expansions and replacements, especially towards the North of Europe, which is the same reason why we see more "modern" genetics in general (more HG admixture in autosomal DNA, more mtDNA H, etc.) If so, then U106 is probably related to Funnelbeaker, while P312 should be considered in its various subclades, for example M529 is surely related to the Neolithic settlement of Great Britain and Ireland from Brittany and NW France c. 4000 BCE, the expansion of U152 towards Switzerland and Italy may be related to the Chassey-La Lagozza proto-Ligurian cultural fusion across the Alps of similar age. However I do not have a good idea of what could explain in this model the expansion of Z196 towards Iberia.

This is my best Neolithic model. I say: sample Neolithic Britons and Bretons, sample peoples from the Atlantic basin if what you want is to find R1b-S116/P312. If what you're looking is for U106 probably you want to look at NW France, Belgium of NW German Neolithi... or maybe it spawned from Danish early TRBK (although then harder to explain its impact in Britain, unless it is indeed "Viking").

Krefter said...

@Maju,

I said this earlier. Ancient Y DNA doesn't support a Megalithic or Neolithic expansion of R1b-L151. It supports a lone R1b-L151 in Western Europe in the Neolithic that expanded by at least 2200 BC(Central European Bell Beaker oldest example R1b-L151 being popular).

We have 40 Y DNA samples from Western Europe dating 3000-3600 BC. Only one is R1b1a2. It had not become dominate like today yet. The modern regional-specific clades probably hadn't expanded yet. Unless each was rare when they first arrived in their regions and later became popular.

Artur Martyka said...

Isn't PF6518 present in few haplogroups?

I've made a little check and it's in G-FGC5672, J-CTS130, E-CTS1622, C-V20* and N-Y4374, aside obvious R-M269.

E-CTS1622 is only found in some Yoruba sequences, so it's rather not possible for ATP3 to belong to this branch.

J-CTS130 is found in 2 Columbians from Medellin but I have no idea if they particularly fit with their age (even considering that YFull ages are somewhat too low).

G-FGC5672 - it has Oetzi and Punjabi from Lahore, more distant upstream sample belongs to Puerto Rican from Puerto Rico. Age doesn't seem to be the obstacle in that case.

This sample can also very well be C-V20 like La-Brana. It is not unlikely.

So, everybody is still aroused? There are few likely possibilities - G2a,C-V20,R1b-M269.

YFull will take care of ATP3, anyway :).

Gioiello said...

This is the ISOGG tree of hg. H2 with the Sardinian samples from Francalacci et al. 2013 and 2015:

•H2 P96 (14869743 C->A) Sardinian sample 139
• •H2a L280 (14869743 C->A) FTDNA subgroup [removed by Family Tree DNA prob in early '15]
• •H2b Z19065 (2716568 G->T) Sardinian sample 140
• • •H2b1 Z19045 (2681998 G->A)Sardinian sample 141
• • • •H2b1a Z19039 (16710674 G->T) Sardinian samples 142-143
• • • • •H2b1a1 Z19033 (6937005 G->A) Sardinian samples 144-145

Krefter said...

@Maju, About whether Steppe-ancestry is significant.

I don't have time to lay out all the evidence, but I might gather it up at some point. But, in the mean time I think you should read Haak and Allentoft over again. Read Threads made by Davidski over the last 7 months. Read what Harvard professors who sampled the DNA said themselves. Their reliable people. The only legitimate counter argument is the methods used are flawed.

Krefter said...

@Artur,

ATP3 had 3 Y SNPs leading up to R1b1a2 and he was negative for Y SNPs leading up to justa bout every other Y DNA haplogroup found in Pre-Historic Europe.

Artur Martyka said...

@Krefter

OK, I had no idea about the negatives. Thanks for the information.

So it was likely an R1b. A dead end?

Annie Mouse said...

Hmm I am waiting for the actual paper. This press release looks alarmingly inconsistent to me.

These neolithic farmers look like other neolithic farmers AND the Basques?!

Therefore this somehow means that the Basques must have arrived in the neolithic with farming!?

This does not fit with previous data like Behar's assertion that:
"We detected signals of the expansion of these haplogroups at ∼4,000 years before present (YBP) and estimated their separation from the pan-European gene pool at ∼8,000 YBP, antedating the Indo-European arrival to the region. Our results clearly support the hypothesis of a partial genetic continuity of contemporary Basques with the preceding Paleolithic/Mesolithic settlers of their homeland."

It is worth remembering that we still don't know anything about the autosomal genetics of the people who lived around the Mediterranean prior to the neolithic. We have hardly any genetic information at all (a few degraded haplogroups), but the archaeology looks quite different from the north.

Davidski said...

The paper has already been published...

http://www.pnas.org/content/early/2015/09/02/1509851112.abstract

Kristiina said...

I noticed that according to Davidski's K12_10K_SNPs_or_more_results, RISE471 (Corded Ware, Untermeitingen, Germany) is 50% European HG, 26% Afanasievo and 22% Bedouin. Among the CW samples, he/she is the only one to have a significant amount of Bedouin. Rise 471 is mtDNA J1c1b just like one of Portalón samples (ATP7 J1c1b1, age 5295-4894). So, it looks like there is similar gene flow from Neolithic farmers into Corded Ware and Portalón or even from South Western Europe into Corded Ware.

Simon_W said...

Interesting the full ADMIXTURE analysis: ATP9, the Middle Bronze Age female, seems to have a tiny bit of steppe admixture. It starts to appear at K=6 and is most evident at K=9. It's just minute, but it's there. I think this is just minor geneflow from LNBA central Europe with no effect on language.

Kristiina said...

Correction: RISE471 is not Corded Ware but Bronze Age from South Germany. His sex is male but DNA sex not determined.

Davidski said...

In Supplementary Table 9 RISE471 is listed as Corded Ware. But not in Supplementary Table 1.

So it might be an uncertain classification.

truth said...

@ Annie Mouse

Basques have more Hunter-Gatherer than Neolithic Farmers or Sardinians.

Simon_W said...

On a side note, a word @ Maju's comments regarding the Ligurians. I know there is an old scholarly tradition to regard the ancient Ligurians or Ligures as a non-IE population. But if you review the evidence brought forward for this opinion, it turns out to be very flimsy. At any rate there is not a single non-IE text attributable to the Ligurians. The argument rests basically on place names which have been interpreted as non-IE. But even if this could be proven true (I'm not sure there), it doesn't follow that these names are from the time when Ligurian was spoken in the area. They might be from the pre-Ligurian substrate which may have existed once. A couple of very short inscriptions from the Ligurians do exist, and they show a language that's intermediate between Celtic and Italic, for instance according to the linguist Georg Renatus Solta. See also the discussions here: https://en.wikipedia.org/wiki/Talk:Ligurian_(ancient_language)

The direct epigraphic evidence, albeit meagre, is for sure the most important category of evidence regarding the classification of ancient Ligurian. But there are also other hints:

There is a correlation between formerly Ligurian areas in Italy and a higher incidence of light hair colour: http://jpst.it/yK_D
This strongly suggests that the Ligurians came from a more northerly place to the Mediterranean.

Finally there are the opinions of ancient authors which are somewhat more dubious. Seneca the Younger spent eight years on Corsica, and according to him Ligurians settled only along the coast, while the interior of the island was inhabited by Iberians. There is also Thucydides' account, who was drawing upon Antiochus of Syracuse, that the Sicani were Iberians who had been expelled by Ligurians. Both accounts seem to imply that the Ligurians didn't belong to the oldest population layer of the western Mediterranean.

Simon_W said...

Regarding R1b, I'd say, if ATP3 is indeed R1b-M269, then that's a serious blow to the thesis that R1b-L151 is Yamnaya-derived!

What evidence do we have so far? Yamnaya did have R1b-L23, that's an established fact. Then, west of that, there is one (probable) R1b in a Polish Corded Ware male and one in Vucedol, both without any further downstream classification. Although the Corded people clearly had eastern roots, not all of their yDNA was non-local, as the find of a Corded male with G2a had shown. So this R1b might be from locals or not, we don't know. At any rate, the Corded Ware was strongly dominated by R1a, and it seems dubious that this population could be at the origin of the west European R1b predominance. Vucedol on the other hand probably had local roots; according to Volker Heyd it was even oppositional to the steppe intruders.

And then there seems to be this rather early R1b-M269 in northern Spain. Since there was an archeologically attested Bell Beaker movement from Iberia to central Europe, it's perfectly conceivable that this brought about the sudden predominance of R1b there. Autosomal evidence suggested that the Bell Beaker people were intermediate between Corded people and Iberian-like farmers. In Southern Germany there lived very steppe-like and very Iberian-like Bell Beaker people at the same time in the same settlements. This movement from Iberia may have counteracted the Indo-Europeanization for a while, while R1b-L151 was installed. Later movements from the east may have re-Indo-Europeanized the area again, without cancelling the R1b-predominance. As a result, genuinely Indo-European subclades of R1b-L151 arose.

One difficulty remains: according to Yfull, R1b-L23 came into being c. 4400 BC and has a TMRCA of c. 4200 BC. Since virtually all of the West European R1b is L23, the Iberian Bell Beaker people must have had this marker, if West European R1b is descended from them. But as mentioned, R1b-L23 was definitely in Yamnaya. So there must have been a relatively late link between the steppe and Iberia, if this line of reasoning is right. I could imagine that R1b-L23 is from rather eastern HGs which eventually entered eastern farming cultures like Cucuteni-Tripolye. And subsequent contacts between farmers eventually took this to Iberia.

Interestingly ATP3 has quite a lot of shared drift with western Ukrainians, judging from the map with the f3 stats (Mbuti; ATP3, X).

Kristiina said...

I compared Portalón mtDNA and RISE mtDNA:
RISE489 Remedello di Sotto 4185 yBP, X2c1, 47% Bedouin; and ATP16, Portalón, 5211-4866 yBP, X2c
RISE94, Sweden, Viby, Battle Axe, 4025 yPB, K1a2a, 1% Bedouin; and ATP3,Portalón, 5466-5312 ybp, K1a2b
RISE42, Denmark, Nordic LN, 3681 yPB, H3v+16093
RISE71, Denmark, Nordic LN, 3701 yPB, H3b, 3% Bedouin; and ATP17, Portalón, 4957-4821 yPB, H3

With all likelihood, H3 has spread from Iberia to the rest of Europe.

As for K1a2, this citation from Anthrogenica is interesting:
K1a2 is predominantly European but with a single Armenian lineage at its root, possibly indicating a Near Eastern source and an early Late Glacial dispersal into Europe: K1a2a and K1a2b are both wholly European and date to the Neolithic and Late Glacial respectively. There is also a derived Anatolian lineage within the minor K1a2d, suggesting back-migration from Europe" (Costa et al., 2013, Supplements, p. 30).

As for X2c, Anthrogenica also tells that "the table says that one of the remains in the Viking burial site at Galgedil, dated ~1000 CE, was mtDNA X2c.

So, there seems to be plenty of Mediterranean mtDNA that made their way to the North and East.

Kristiina said...

Not only Rise 471, Corded Ware/German Bronze Age (J1c1b), and Portalón ATP7 (J1c1b1), age 5295-4894, are J1c1b, but J1c1b1a has also been found in RISE386, Sintashta, 2.5% Bedouin. So, this should not be a coincidence as we have been told that Sintashta has ENF admixture.

Karl_K said...

Hey Simon.

Who really thinks all that R1b in western Europe is from Yamnaya via Corded Ware? I would say lactose tolerance was highly associated with western R1b from very early on, and independent of language. Only direct male lines have a huge advantage from lactose tolerance when milk comes only from the mother. As soon as that allele came into contact with an animal milk drinking situation in Iberia... Boom! There goes the dynamite.

Aram said...

I am just citing the Maciamo's comment based on Genetiker's admixture run of ATP3.

""
ATP3 (3516–3362 BCE) stands out from other samples thanks to its high Northern Middle Eastern ancestry (31.97%) against 0% for ATP20, 11% for ATP17 and between 0% and 8% for other samples. What Genetiker calls Northern Middle Eastern is what we typically referred on this forum as Caucaso-Gedrosian admixture - the same as in the "Armenian-like admixture" in Yamna samples.

With its 32% of Caucaso-Gedrosian, 14% of Northern European ancestry, 6% of European Hunter-Gatherer and 3.8% of Veddoid, it does indeed look as if ATP3 has about half of Steppe ancestry, which explains why he could be R1b-M269. What is amazing is that ATP3 reached Iberia at the very beginning of the Yamna period !

""

Aram said...

Also this comment of Maciamo

""
Most samples are predominantly Southern European (ATP17=66.7%, Matojo=64.88%, ATP7=63.4%, ATP16=62.6%, ATP9=51.8%, ATP2=46.8%), except ATP3 and ATP20 who who have considerably less (ATP3=32.47% and ATP20=33.03%) as if they had been recently hybridized and only had half of the Iberian ancestry of other samples.

Both ATP3 and ATP20 have a higher percentage of Northern European and European Hunter-Gatherer ancestry, with an combined average of 20.5% and 23% respectively. Only ATP9 has more (32%).

More fascinating is the fact that ATP3 and ATP20 are the only samples with Veddoid ancestry (i.e. Indian subcontinent, but possibly of IE origin), with a remarkable 28.14% for ATP20 and a much more moderate 3.82% fro ATP3.""

Chad Rohlfsen said...

Apoarently, it's negative for L51. So it's still a dead end, at this point.

Karl_K said...

@Aram

ATP3 and ATP20 do seem remarkable. There must have been a direct migration to Iberia (and perhaps other places in Western Europe) from Eastern Europe, which gave rise to the Bell Beakers. So everyone is kind of rightish.

Karl_K said...

@Chad

Exactly. At this point. But it is suggestive.

Alberto said...

The samples are low coverage and apparently very noisy in those tests by Genetiker. I wouldn't take the results too seriously for now, until someone else runs some more tests. The stats in the paper are not too exhaustive to be sure, but I'd be surprised if any of the samples (except maybe APT9 slightly) stand out from regular MN European farmers.

Karl_K said...

@Alberto

Exactly. You can't be too careful.

Karl_K said...

@Karl_K

Exactly. You can't be more exactly.

Gioiello said...

@ Kristiina

When you read Anthrogenica, please read also what I wrote. I was “Rathna” there and wrote more than 1000 letters before being banned, and they dedicated to me a thread “Rathna’s assessment of genetic material” which has been read for more than 53000 times. Don’t read only my opponents, were they Gail T, like Tonnesen, or Jean Manco. I think having brought tons of proofs for my theory. Not last I found, within three years (with a little help of Ian Logan), the most varied mt hg. H with 12 extra mutations, and against what Ian thought, in favour of the Caucasus, she resulted full blooded Italian: Barbara Ann Lewis, alias Di Mele, Balducci, Gabriele, Sampo etc. Her ancestress from Molise Apennine, a very conservative region.

Simon_W said...

@ Chad, yes ATP3 is negative for L51, but so are the Yamnaya males sampled so far.

Simon_W said...

Sound intriguing, these ADMIXTURE results by Genetiker, but I'm still sceptical, let's see if others can confirm it.

If ATP3 was just a regular MN farmer, then the following caveat applies:

The apparently young age of L23 is a serious obstacle for theories about a western origin. If that marker goes back to 6400 BP, and was in Iberia at the time of ATP3, then it had just 1000 years to cross the continent – and this in a context of sedentary farmers. And it wasn't just a minority in Yamnaya, no, the large majority of Yamnaya males even in far eastern Europe were positive for L23. And moreover it would be a very strange coincidence that R1b-L23 had an enormous success on the PC steppe and roughly at the same time in a completely different culture in western Europe, if these enormous expansions were completely unrelated with each other.

Simon_W said...

A link with lactose tolerance as suggested by Karl might be an explanation, but I don't quite see how these different alleles could stay linked, as lactose tolerance is determined by autosomal DNA...

Simon_W said...

In the PCA of the paper ATP3 doesn't look very unusual: he's inbetween Basques and Sardinians, with a slight shift towards Spaniards. The latter might be indicative of eastern admixture, but it's quite slight.

Karl_K said...

@Simon

"A link with lactose tolerance as suggested by Karl might be an explanation, but I don't quite see how these different alleles could stay linked, as lactose tolerance is determined by autosomal DNA..."

Lactose tolerance is autosomal, true. But, it can be selected in an unusual fashion related to gender.

All children with lactose tolerance have a potential advantage, being able to nurse at advanced age. But, any female children have a later disadvantage, as at least half of their own older children could compete with the infants.

Only the direct male line has the advantage at each stage in every generation.

Chad Rohlfsen said...

They've only sampled east of the Don.

Gioiello said...

@ Simon W

Your reasoning unfortunately doesn’t fit. My R-L23-Z2110* is 7100 years old from Full Genome (Big Y, that I’d call Little Y, has 12000 no calls out of 72000 SNPs). The samples found in East Europe are old subclades or dead lines of the first R1b1a2, and may be at the origin of some diffuse R-L23 or nearby in the East, but certainly there isn’t any downstream subclade. You don’t consider that haplogroups may have been contemporaneously in different regions, because come before, and the Eastern R1b may have come from Western European Hunter-Gatherers. Certainly the massive R1b-L23-CTS9219 came from West Europe, where we find CTS7556, brother to my Z2110, certainly from Western Europe (link with Barnum and a French Basque)… But I wrote about this matter infinite times. As we find R.L51 in Italy or nearby and not elsewhere every discussion will stop.

la señora bibiloni said...

My hypothesis, if this M-269 is confirmed: I'd say the origin of M-269 is Northern Iran (too bad there's not available ancient DNA from that area). Some went north and that's how you get Yamna M-269 and subclades leading to Z2103 - some went west via Anatolia and that's how you get the Iberian sample. Coming from Northern Iran, you may have Caucasus/Gedrosia without passing through the steppe cultures. I would not even discard some earlier M-269 from Northern Iran migrating to Sredny Stog or CT.

Krefter said...

@Kristiina,
"I compared Portalón mtDNA and RISE mtDNA:"

Too few samples from Haak and this study to learn anything. There's 100s more though of less coverage which tell us a lot.

Romulus said...

@la señora bibiloni

I think R1b is Sumerian.

Krefter said...

@Simon_W,

Who told you ATP3 is negative for L51? That would make sense because it is obvious he was a lone wolf and R1b-L151 not expanding in the Neolithic. There's still plenty of room for R1b-L151 to be descended of a lone Mesolithic European or Neolithic West Asian R1b-L23 lineage.

I don't get why some people think R1b-M269 was born in the Neolithic. We have pre-R1b1a1-M478 in Russia 7500 years ago, R1b-M269 in SPain 5,500 years ago, and R1b-Z2103 in Russia over 5,000 years ago. R1b-M269 I'd say is at least 9,000, 10,000 years old.

bellbeakerblogger said...

@ Senora Biblioni

"I'd say the origin of M-269 is Northern Iran...[and the rest]"

You got my vote.

Kristiina said...

@Krefter
”There's 100s more though of less coverage which tell us a lot.”

That is quite unspecific.

By the way, I took a look at yfull R1b tree and noticed that Z2106 which is found in RISE547, Yamnaya, 2887-2634 BC and RISE555, EBA, Stalingrad Quarry, 2857-2497 BC, is not under R-L51 where European R1b is almost entirely found. Z2106 has a more easterly distribution as according to yfull it is found in China, India, Germany, Slovakia, Bulgaria, Italy, Colombia, Norway, Spain and Russia. Yamnaya Lopatino I, 3305-2925 BC and Yamnaya Ekaterinovska, 2910-2875 BC are under this more eastern branch that is outside R-L51.

Neither RISE397, LBA, Armenia, 1048-855 BC is under L-51, but instead under the above mentioned more eastern Z2106 branch. However, RISE Bell Beaker, Nordic BA ad LN samples are already under R-L51.

It is not impossible that L23 developed somewhere in the midway between Iberia and Ukraina 6300 (or 7100) years ago and spread from there to the East (Z2103) and West (L51), although there must have been much earlier offshoots of R1b, which are not L269 or L23, from an unknown place to multiple directions such as African lineages and Samara Hunter gatherer lineage, M478, which has a TMRCA of 7300 years.

In my opinion, Yamnaya R1b and Samara R1b were different folks who spoke a different language and had a different history.

Gioiello said...

@ Kristiina

“It is not impossible that L23 developed somewhere in the midway between Iberia and Ukraina 6300 (or 7100) years ago and spread from there to the East (Z2103) and West (L51), although there must have been much earlier offshoots of R1b, which are not L269 or L23, from an unknown place to multiple directions such as African lineages and Samara Hunter gatherer lineage, M478, which has a TMRCA of 7300 years”.

“midway between Iberia and Ukraina”: did you mean Italy? Don’t you know that there is a country named “Italy”, whose name seems coming from a word linked with Latin “uitulus” (calf)?

“6300 (or 7100) years ago”: 7100: are you referring to my hypothesis and to my Full Genome? Gioiello Tognoni: FTDNA: ID H1614 (tested firstly from DNAHeritage, after absorbed as pretty all the rest from FTDNA), FGC: ID LDGHP. The question is wrongly posed. R-L23* doesn’t exist and R-Z2103 and subclade is only a brother clade of R-L51*, which doesn’t exist and has only a subclade like R-PF7589 which is a brother clade of L11 etc etc. We hope, as philosophers said, that: if smoke perhaps fire. It is undoubted to me that Italy has the highest presence and variance of R-PF7589* and has all the pathway from R1b1-L389* to R-M269* to R-PF7589* etc etc but also R-V88 (highest variance in Sardinia), R-M335, R-M73… for that I theorized the “Italian refugium”.

Mike Thomas said...

@ Karl K

" There must have been a direct migration to Iberia (and perhaps other places in Western Europe) from Eastern Europe, which gave rise to the Bell Beakers"

Did they catch Ibiza airlines ?
:)

Karl_K said...

@Mike

I can't say for sure which airline. We need more data.

Chad Rohlfsen said...

M269 is old and all over. Even east of the Urals. If they aren't L23 or more likely, L51, forget about it being an ancestor.

Chad Rohlfsen said...

Genetiker's admixture runs are almost always dubious. 4-7% African? Come on! He doesn't know what he's doing in this area. I'll test them for EHG after I get them.

Mike Thomas said...

@ Chad

Were any Yamnaya tested so far L23?

Chad Rohlfsen said...

I think one was.

Chad Rohlfsen said...

They're all younger than L51 though. I have a good feeling about west of the Don.

Mike Thomas said...

I just checked and I don't think any yet samples Yamnaya are L51. But you're right that L51 arose west of the Don. If the origin of L51 is still East of France, it'll have to be in one of those Baden-Beloraz-Cernavoda areas, or some pre-Yamnaya group on the Dniester

Richard Rocca said...

David / Chad, it would be interesting to find best population fits for Bell Beaker now that we have Copper Age southern European data in the form of Remedello and Portalon. Would Samara_HG plus Spain_MN???

Davidski said...

I don't yet have these gnomes. Trying to get them now.

Krefter said...

@Simon_W,

Who told you ATP3 is negative for L51?

Roy King said...

@Krefter,
According to Ted Kandell, there is a no read at L51 and at L-51 equivalent SNPs. He is confident now however that APT3 is derived for an M269 equivalent SNP (like Genetiker stated). We know nothing about L23 or L51.

Krefter said...

@Roy,

Thanks for clearing that up.

Open Genomes said...

David and everyone, you can get the El Portalon genomes BAM files here:

El Portalon whole genome sequence BAM files and indexes

Chad Rohlfsen said...

When I get them from David, I'll include them in my test. We'll see if they have EHG.

Aram said...

The ATP17 was not assigned any Y Hg by the authors. Genetiker classified it as I2a2a-M223.

ATP17 Pre-Beaker Copper Age 3007–2871 BC I2a2a-M223 calls
https://genetiker.wordpress.com/y-snp-calls-for-atp17/

Looking at the number of positive calls in the haplogroup I/I2 it lefts little doubt that the assignment is correct.

Krefter said...

Detailed analysis of ATP3's Y DNA by Ted Kendell.

http://www.anthrogenica.com/showthread.php?1646-Genome-of-a-late-Neolithic-Iberian-farmer&p=107950&viewfull=1#post107950

I(I2, I1), J2, G2a, T1a, H2, E1b-M78, C1a2, R1b1c, R1b1a2, R1(xR1b1a, R1a1a) are the lineages found in Neolithic farmers dating 5500-3000 BC.

Maybe R1b1a2 was just one of many Eurasian lineages that one way or another made its way into Neolithic Europe. R1b1c existed in early farmers with no ANE, why not R1b1a2?

The finding of R1b1a2 in Neolithic Spain doesn't disprove the Steppe-theory but it definitely leaves room for a local origin of R1b-L151. We have R1b1a2 in both Russia and Spain 5,000 years ago. R1b-L151 could be from either location.

If R1b-P312 is of local origin, R1b-L21 still probably arrived in the British Isles with Steppe-blood and R1b-U152 probably arrived in Italy with Steppe-blood. R1b-DF27 would probably be a differnt story.

Colin Welling said...

"M269 is old and all over. Even east of the Urals. If they aren't L23 or more likely, L51, forget about it being an ancestor."

You are right, M269 is relatively old and apparently it spread quite far and with different cultures by 3k bce. I still await confirmation from the academics since they chose to leave it out. Someone who understands ydna very well should write to the authors!

The steppe origin for modern west european r1b is still outstandingly favored for 3 reason. (Frequency, Timing, and Migration Pattern)

1) The frequency with which we find relevant r1b in the steppe around 7kya to 5kya far exceeds the frequency we find it in Western Europe.

2) Eastern Yamnaya are L23 derived and they lived in a time that was close to the most recent common ancestor of L23 people, i.e. the yamnaya were very close in time to the common ancestor between them and modern west europeans.

3) For L23 to have originated in Western Europe would require that Eastern yamnaya are the descendants of west europeans. Yet there is no known migration to support this notion. We would also need to explain how the yamnaya are associated with PIE a steppe language. Finally, the Eastern yamnaya do not have autosomal admixture from western europe post neolthic but West Europeans are proven to have steppe like admixture post neolithic. Remember, all but one yamnaya results have come back r1b while half of the yamnaya heritage is EHG (a native of the steppe and absent in neolithic West Europe.

This new finding adds some complexity but the overall trends remain the same.

Gioiello said...

@ Colin Welling

Your theory lacks a fundamental presupposition: Eastern Europe and Asia lacks the upstream subclades of R1b. Asia has R1b1-L389- but not R1b1-L389+ like Western Europe (highest variance in Italy, Ashkenazi Jews introgressed from Western Europe). Only Armenia has a R1b1-L389 * haplotype (that with YCAII=23-23), whereas Italy has many: 18-22, 18-23, 23-23. Your theory lacks the fact that probably the Eastern European R1b expanded with Western European hunter-gatherers, like very likely R1a-M420* and the oldest subclades, which only more recently developed in Eastern Europe, where the R1a subclades are rooted.

Mike Thomas said...

Gio
Have you ever considered that the apparently old R1 and J2 groups in italy are due to the persistence of a neolithic tradition rather than Italy being some kind of (otherwise undocumented) mesolitic refigiuum ?

Colin Welling said...

"Eastern Europe and Asia lacks the upstream subclades of R1b"

Not really. Also, I'm was talking about L23 and M269, not r1b.

Alberto said...

@Gioiello

I respect your knowledge of Y-DNA and your hypothesis in which you have worked hard and for a long time. But you have to adapt it to the new findings if you want it to still be relevant.

For example, if R-L23 formed in Western Europe in the Chalcolithic, for Yamnaya to have R-L23 we need to have some prove (archaeological, genetic,...) that there was a migration from Western Europe to the Steppe after 4000 BC. But as Colin pointed out, there's not such proof. Yamnaya have 0% Western European admixture (either from farmers or from hunter-gatherers). Nor any cultural influence, etc...

For your theory to stand, you have to address these issues. You can't just point to modern DNA, because that's no proof in itself. Modern Italians could have arrived from Iran 3-4 thousand years ago bringing R1b with them. The upstream subclades don't exist in Iran anymore maybe because the people who carried them *left* Iran and moved to Italy.

It's just an example, but the main point is that you have to provide better possibilities to explain our current state of knowledge regarding ancient DNA. If you manage to explain that, then your theory can live on until more data can prove it (or disprove it). But right now you're not attempting to explain the current findings successfully. You repeat the modern DNA patterns again and again, with just very general statements about ancient DNA ("R1b in Yamnaya could come from WHGs". After 4000 BC???? How exactly?).

Simon_W said...

@ Karl

Quite an unusual explanation. I've never had the idea that lactose tolerance could have started with adults feeding on their mother's breasts! And is there any evidence that they did? I'd rather think it had something to do with the consumption of cow, goat or sheep milk. And the advantage that is provided by the abiltity do digest large amounts of these sorts of milk is shared by all people, male, female, and hence all the offspring who had inherited the necessary alleles.

@ Gioiello

Well, maybe L23 is somewhat older then, if yours is really that old. That would make it even more possible that it crossed the continent before notable amounts of ANE admixture did.

@ Krefter

I don't know that ATP3 is negative for L51, it was just a misformulation. ;-) Yeah, M269 is probably quite old, it may have crossed the continent in hunter-gatherer times.

Simon_W said...

@ Mike

The large majority of Yamnaya samples is positive for L23. But L23 without any positive downstream markers is only one: I0443.
And yes, as I said, no L51 has been found in Yamnaya tested so far.

Gioiello said...

I answer you both.
@ Mike Thomas, Colin Welling
I have said from the beginning, just for the papers of Zilhao et al. which demonstrated that there has been a migration of agriculturalists by sea from Italy to Iberia 7500 years ago, that hg. R1b-L51 (and others possible upstream subclades) were brought to Iberia from them. It is the Cardium Pottery culture, widely demonstrated by an archaeological point of view. Of course I have also always said that there was the possibility that those agriculturalists had come from East (Asia Minor, Levant etc.), even though I believed from many other considerations that they were autochthonous of Italy. We are waiting that aDNA tests from those places demonstrate whether those haplogroups were there before or not. I think don’t. We’ll see. By a scientific point of view are the proofs to decide which theory was right and which was wrong. What I disagreed was the presupposition that what was in Italy had always come from elsewhere, and Colin repeats that, in saying that Iranians may have come to Italy even though no one of those hgs is found there now. I didn’t accepted that, just because it was a prejudice, beyond any possible demonstration, so, as not being falsifiable, no scientific (see Karl Popper).
But my theory wasn’t based only on hg. R1b, but upon many others, both Y and mt, just those are being found in Iberia now. I cannot read Anthrogenica, but Krefter posted here the analysis of Ted Kandell (I know very well him, from when I always said him that his Jewish G2c was rooted in Italy and he very likely didn’t like that). If these are the hgs found in Iberia

I(I2, I1), J2, G2a, T1a, H2, E1b-M78, C1a2, R1b1c, R1b1a2, R1(xR1b1a, R1a1a) are the lineages found in Neolithic farmers dating 5500-3000 BC.

I bet that they came from Italy. The proof? To test Mesolithic Italy as the person we cannot named here asks from so long (and me too).

Alberto said...

@Chad

"M269 is old and all over. Even east of the Urals. If they aren't L23 or more likely, L51, forget about it being an ancestor."

Yes, I agree. But I hope this applies to all the cases and we can finally also put into context the R1a/R1b findings in EHGs. "R1b is an EHG marker" is no longer viable. And it will be the same for R1a once we start finding it all around Asia.

So this R-M269 at least serves to dispel some myths and open up other possibilities.

Rokus said...

'I've never had the idea that lactose tolerance could have started with adults feeding on their mother's breasts!'
I've heard of Yamnaya warriors carrying around their dolls. To travel the Eurasian plains with their mothers on horseback for a steady supply of milk may be just a small step. Was their paternalistic battle cry actually their whining for their mothers teats? Lol.

Simon_W said...

Lol @ Rokus' comment

@ Colin's 3 arguments

I agree that the presence of L23 in far eastern European Yamnaya is somewhat tricky to reconcile with a secondary expansion of West European R1b from Southwestern Europe. It may be considered a hint slightly in favour of the steppe theory. But it's not completely impossible to reconcile it with the alternative explanation: if L23 first arose in a HG in eastern Europe, went east into Yamnaya, and somehow managed to go west as well. I'll soon elaborate a little how this might have happened.

Also, I'm not convinced of the frequency argument. Practically all west European R1b is descended from the first person who had L51, so you don't need large amounts of people with L51. The large amounts are superfluous, because they all go back to one. You're of course right that it's unlikely that R1b ever played much of a role in western Europe if it once was a very rare haplogroup. But we don't know exactly how rare it was in the Chalcolithic, because the sample size is still low. At least it was detected in this very small sample from El Portalon, so probably at least in this time and place it wasn't that rare.

Gioiello said...

@ Alberto

I apologize. I spoke of Colin but were you. Don’t forget that the oldest R1a-M420* so far belongs to Jones, from the Isles. We have discussed a lot on eng.molgen, both Mikal and me, about R1a and its origin. Perhaps, as also the person we cannot name here asked, to test Mesolithic Great Britain with Mesolithic Italy wouldn’t be bad.

Mike Thomas said...

Gio

"I(I2, I1), J2, G2a, T1a, H2, E1b-M78, C1a2, R1b1c, R1b1a2, R1(xR1b1a, R1a1a) are the lineages found in Neolithic farmers dating 5500-3000 BC. All came from italy"

Wow that's a big call!
I guess you'll be proven correct soon ..

Mike Thomas said...

@ Simon
My earlier comment about L23, etc, was replying Chad, but we're speaking the same thing. :)

Simon_W said...

Since ATP3 only appears to be positive for M269, whereas L23 was present in Yamnaya, I'm not quite sure at the moment which theory about the origin of West European R1b to prefer. More evidence is needed. But isn't it odd, there always tends to pop up some R1b in Iberia. First the early Cardium farmer who appeared to have the same R1b1 as the Samara HG. Now we know he was R1b1c and the Samara HG R1b1a. But now again, as if the theory of an Iberian origin doesn't want to die out, there pops up an R1b-M269 in Iberia... Will it again turn out to have been a dead end? Possibly!

But there might be a way how R12b-L23 could have spread to Yamnaya and to Iberia from a third source: Some people believe that there were westward migrations of steppe people or EHG already in the 5th millennium BC. They suggest that, among others, the Gatersleben group in central Germany (starting c. 4400 BC) and the Tisza-Polgar culture in Hungary (4400 – 4000 BC) show evidence for such influence, but it's controversial. But from what I've read about the cranial evidence in Hungary there appears to be some truth in these suggestions. The timing would be right to enable the possibility that R1b-L23 was involved in these migrations. Culturally and physically however, they didn't leave a lasting impact, apparently if there were eastern migrants they were not numerous and their influence quickly got diluted. That's also obvious from the ancient autosomal DNA evidence from central Europe. But the yDNA doesn't get mixed up, so from that time on some people with R1b may have been around. The R1 that has been found in Baalberge may be related with this. And maybe, when the migration of the Globular Amphora people started, which possibly began rather early, c. 3400 BC, some people with R1b-M269 and R1b-L23 were dislocated westwards. But that's of course speculation.

Mike Thomas said...

Simon W

If this M269 is indeed true, then people just need to get real and stop with the EHG BS. Sorry but how are EHG people colonizing Africa and Iberia ? Flying horses ?

People are engaging in so much special pleading. The inescapable conclusion is that M297 arose in western Asia. But that was already obvious in 2009 with modern DNA.

The intricate details of Z2103 and L51 need further resolution.

Davidski said...

We have three similar and unadmixed forager samples from North Eurasia with R*, R1a and R1b.

As a result it's now exceedingly difficult to argue that R, R1a and R1b were common at the same time among foragers in the Middle East and/or Central Asia unless we find direct evidence of it. That's because these areas always had starkly different environments from north Eurasia.

Different forager meta populations didn't just run around all over the place. They were specifically adapted for where they lived.

Obviously, it's much more sensible to assume that some, now probably extinct, R1b lineages trickled into the Near East and/or southern Europe from Siberia and surrounds during the Neolithic, and got caught up in some of the Neolithic expansions.

Mike Thomas said...

True that's possible. But the R*, R1a and R1b aren't really similar or unadmixed.

R* in Mal'ta shos its own peculiarities, not the least, a semi-recent south asian origin. The R1a* and R1b-pre M73 in Samara look indeed like "pure EHG", which is in fact some kind of yet undefined palaeo-Siberian descendent of mal'ta drifted populated with WHG. Then the Yamnaya and CWC look west asian admixed.

So they all show rather distinct histories.

Its a bit artificial to think that the western Russia was the well- spring of Eurasia.

Davidski said...

I wouldn't say Mal'ta boy shows recent origins in South Asia. His relationship to South Asia is far from resolved.

R and R1 still look like markers of North Eurasian foragers. The fact that they seem to have trickled south and west, probably even as far as the Canary Islands, during the Neolithic means very little in the scheme of things.

What really matters is where most of the R1b-M269 that is now found in the world expanded from during the tail end of the Neolithic. The data we have currently doesn't resolve this issue, although the steppe is the front runner.

Karl_K said...

@Simon

"Quite an unusual explanation. I've never had the idea that lactose tolerance could have started with adults feeding on their mother's breasts!"

Very funny. But obviously I was talking about prepubescent children. Especially the 4-8 year olds who may have taken a little extra in times of need. Older children with lactase persistance would have an advantage over children from a different father.

Karl_K said...

@Davidski

"What really matters is where most of the R1b-M269 that is now found in the world expanded from during the tail end of the Neolithic. The data we have currently doesn't resolve this issue, although the steppe is the front runner."

It probably originated in the steppe, but the expansion must have occurred in Western Europe. How could all the R1b people just walk around the R1a people. Excuse me. Pardon me. Coming through...

Alberto said...

David, MA1 belonged to a population that probably went extinct. Siberia got depopulated shortly after. But not forever. A few thousands of years later it got repopulated again, by 2 different populations that mixed and even migrated to America.

From where do you think that Siberia could be repopulated?

One place is East Asia (China, Korea,...). From there Siberians (and subsequently Native Americans) got East Asian admixture.

What could the other place be, from where an ANE population entered Siberia? We do have one hint: S-C Asians have more ANE than all other Eurasians. And another hint: HGs from Eastern Europe also got admixed with this population. So somewhere between Siberia, Eastern Europe and South Asia looks like a possible area of origin, I would say.

Maju said...

L23 is a widely extended haplogroup which is common in West Asia and, for whatever lesser frequencies R1b has in these regions, also the Balcans and Eastern Europe, in all these cases almost always in its L23(xM412) version.

Downstream of it, the M412* and P311* (L11?) stages are scattered in Central Europe, Italy, Iberia and even in parts of Britain. But they are rare and were probably always that way.

And then, downstream of P311, is where the interesting stuff begins: S116/P312 and U106. In the case of S116 a South French origin is almost impossible to question, U106 is less well studied AFAIK but frequency scores suggest the Netherlands as a possible origin.

So L23(xM412) is as pointless as M269(xL23) to infer from it anything re. Western R1b of the S116 or U106 clades. Focus on what really matters please.

Aram said...

Maju

I agree with You that M269(xL23) is pointless for understanding the Western R1b because there is a 7000 years separating him from L23 at minimum. And some 80 new SNPs. The 7000 years for a herder/hunter tribe is sufficient to cross all Eurasia if needed.

But from L23 to M412 there are only few new SNPs. So the the M412 has very little time to move away from the L23's 'homeland'. That's why I would not say that L23 is pointless for Western R1b.

Rokus said...

What may have been the case is that Iberia was in early contact with an external source of M269. The prevalence in Neolithic Iberia of haplogroups that nowadays are quite rare, such as H2, E, I2a2, C, almost exclude Iberia as a primary source of R1b. However, since the autosomal components of ATP3 doesn't seem to point at ANE admixture at all, it is equally likely this external source was somewhere in the vast genetic terra incognito of Middle-Neolithic NW Europe. According to Mallory and others there still exists an influential lobby of proponents of a Rhineland origin of Bell Beaker, whose direct precursors would then be protruding foot beaker, western TRB and Swifterbant. It is inconceivable that both pre-beaker Iberia and Rhineland could be mentioned as the origin of Bell Beaker without some kind of prehistoric Atlantic contact.

Rokus said...

And why not, the Lower Rhineland is still a treasure of R1+ variability, hosting most L51+'s rare and common subclades. An old layer of remarkable R1a variability should be mentioned as well.

Davidski said...

Very interesting. ATP3 clusters close to north Spaniards (Spanish Pais Vasco). Look for Spain_CA here.

https://drive.google.com/file/d/0B9o3EYTdM8lQY2R2XzktRXZhUWs/view?usp=sharing

There's not much data to work with, but ATP3 doesn't seem to be a typical Middle Neolithic/Copper Age farmer. He does show a hint of eastern influence, like Basques and Pais Vasco Spaniards, or maybe even a little more.

I'll try and run some formal stats later today. Gotta run now.

Maju said...

@Aram: part of the problem is that your judgment is based on chronological guesstimates that are at best dubious and at worst totally wrong. MCH-derived guesstimates cannot be taken as "evidence" of anything... unless they are extremely solidly demonstrated. Just looking at what Y-Full claims without any documentation or backing is not "evidence" of any sort.

Maybe someone could make a good case for a properly calibrated chronology of R1b that looks relatively solid... but I have yet to see it. By the moment just disregarding MCH guesses is the best attitude.

Alberto said...

@Davidski

If that turns out to be correct it's really big news! R1b-M269, in Iberia ca. 3400 BCE with some 7% ANE?

If it has enough markers you could also try to run it through your latest K12 that could detect well small amounts of ANE.

It would be good to see ATP2, that is the one with highest coverage and from ca. 2700 BC. If ANE was in the area by the time of ATP3, it should be in ATP2 too, I guess.

CroMagnon said...

David

The PCA on the publication looks interesting and is portrayed differently to others- as if Mesolitiic hunters are themselves " eastern".

Krefter said...

@Davidski, Alberto.

In the West Asia+Europe PCA from Günther et al. 2015 the El Portalon individuals are in-between Sardinians and Basque and some cluster by Spanish. They all looked like they had a little ANE. This could be because of WHG not ANE. WHG-genomes were clustering by SHG, not at the NorthWestern edge in the PCA like they usually do.

Anyways, it'd be shocking if the ANE-affinity in Basque and other Iberians has been there since the Copper Age.

Davidski said...

I'll check this out in more detail later, including with formal stats.

The top 10 IBS pops for ATP3 are these...

Basque_Spanish 0.830285
Sardinian 0.829974
Basque_French 0.829737
French_South 0.829656
Spanish_Murcia 0.829611
Spanish_Cantabria 0.829379
Spanish_Cataluna 0.829140
English_Cornwall 0.829034
Croatian 0.828996
Spanish_Valencia 0.828771

batman said...

L-11 - origin?

http://justpaste.it/m11s

Myers et al, 2011:
http://www.nature.com/ejhg/journal/v19/n1/full/ejhg2010146a.html

Maju said...

That's pretty nice, Batman, and judging only on that and the complementary U106 and S116, we should think of something like North France as the ultimate L11 origin. As I've said several times it's not too important because the big expansions correspond to S116 and U106 and the upstream phylogenetic stages are just blurry, relatively pointless precursors.

But for whatever knowing the origin of L11 and M412 is worth, it must be noticed that M412(xL11) has a more southerly distribution and that when I'm thinking about the origin of L11, I'm usually thinking about both stages, because the northern L11* might well be a yet undescribed single subhaplogroup in spite of the scatter (in fact ISOGG right now has no clear stand on the position of the L11 marker, which is said to be under investigation). By contrast to L11*, M412* is found only (Myres 2010) in Slovakia, Italy and Spain.

So tricky. I say: let's focus on S116 and U106 because those are the clearly expanding haplogroups, not M412 or L11, which may have traveled with a much larger pool before founder effects.

Maju said...

PS- When I wrote: "By contrast to L11*, M412* is found only (Myres 2010) in Slovakia, Italy and Spain", I should have added: "and in all those places it is found along L11*", what I believe is important. L11* is in fact much more widely distributed than the map suggests, the map only enhancing the three areas where it is exceptionally frequent. Some of those were already apparent in Myres 2010, on which I made this map back in the day.

Maju said...

PS2- Actually looking again at that map, one can imagine M412 migrating via Crete, Italy and Iberia, with a branch heading North to Slovakia. As for L11 it is instead much harder to pinpoint because L11* is everywhere at low frequencies, even in some cases in Eastern Europe too.

CroMagnon said...

Maju
Interesting comments. But the critiqued Id imagjne being aimed at them is that some rare , modern basal clades in Italy and Greece are just relics of an early expansion from EHG territory

Kristiina said...

I go back to the Yamnaya question. I compared this R1b tree (https://drive.google.com/file/d/0B9o3EYTdM8lQSnowY29UU09HSmM/view) and R1b tree here (http://genome.cshlp.org/content/suppl/2015/02/18/gr.186684.114.DC1/Supplemental_Figures.pdf) and added some info from yfull.

On the basis of that comparison Yamnaya does not look very IE. R1b-269(xL23) is found in Turkey, Belarus and Sardinia and in RISE524, Mezhovskaya, and RISE 546 Yamnaya, Temrta IV. R1b-L23(xL51) has four branches: 1. Turkish/Moroccan Jews; 2. Armenian, Assyrian, Sardinian, Sicilian and RISE397, Late Bronze Age Armenia, Kapan; 3. Avar branch that is found in RISE550, Yamnaya, Peshany 4. Tabassaran branch, Pathans, Indian, Chinese, Burzyan Bashkirs, Kurd +European branch that could originate in Bulgaria area, and RISE547, Yamnaya, Temrta IV and RISE555, Stalingrad Quarry.

Modern groups speak many different languages: Armenian, Northeast Caucasian (Avar, Tabasaran), Turkish, Bashkir Turkish, Arab, Chinese, Urdu/Hindi, Kurdish, Slavic and Romanic languages. Georgian should probably also be in the list. I do not know if the case is very strong for IE identity of Yamnaya. I would put my money on a NortIheast Caucasian identity.

Aram said...

Lol Kristiina

Are You serious about modern Turkish?
On what haplogroup You will put Your money for Anatolian IE speakers ?


Karl_K said...

@Kristiina

"On the basis of that comparison Yamnaya does not look very IE."

Perhaps you mean that R1b does not look IE? Yamnaya clearly look IE based on the fact that a huge migration of people closely related to them into Europe is strongly associated with IE languages. Are you disputing this based solely on R1b distributions?

Mike Thomas said...
This comment has been removed by the author.
Maju said...

@Cromagnon: IMO what you say is a relic, via wishful thinking, of a hypothesis that does not work no matter how you look at it. I mean: how do you link M412 to Eastern Europe? You can't even force-feed it, let alone make it look parsimonious. So...

Karl_K said...

@Mike

I didn't say Yamnaya invaded Europe. I said a group closely related to them migrated to Europe. The autosomal DNA speaks for itself. How is that in any way circular? People largely descended from a group defined by autosomal DNA largely speak a language group with a similar time of divergence.

What exactly is so speculative?

Mike Thomas said...

Yes a people -*similar to Yamnaya * colonized Europe. Later these similar people admixed in Europe, on the one hand, and on be other they appear to have moved out back out into the stepe lands. So clearly, this wasn't the "final chapter" of europe that some peole believe.

Today , they're most similar to non-IE Ugric and Siberian groups. And R1b is a broad marker, and cannot be fathomed to be "IE".

If we're going to follow strictly genetic arguements- folly IMO- then those above facts should alert us against conclusions that ANE/EHG= PIE.

Karl_K said...

@Mike

Your problem with R1b is not from something I said. I agree, which was why I said as much to Kristiina.

As for the autosomal DNA.

"Yes a people -*similar to Yamnaya * colonized Europe. Later these similar people admixed and appear to have moved out."

What data are you looking at? Where do you think the IE languages originated? It doesn't look like it's from the first farmers, based on Sardinia and Basque and the timing of divergence.

I would love to hear your alternative hypothesis.

Mike Thomas said...

Definitely around the Ponto Caspian, although I'm open to which side (N / S). But the fact is PIE was still unitary circa 2500 BC. So whichever language expanded with the movements of 3000 BC,it wasn't Indo-European

That's all I'll say at the moment

Mike Thomas said...

I accept this might be construed as wacky and fringe. But I'll be proven right in five years time

Maju said...

Only some very particular, not really archetypal, Finno-Ugric groups, Mike. Or are you talking of the Nenets or Khanty? No you are not, you are talking of Estonians. Anyhow early IE and early Finno-Ugrics were in intense contact in Eastern Europe and even West Siberia later on.

Mike Thomas said...



And we need to stop assuming that just because cultures or people might appear generically homogeneous they all spoke the same language. Not all Yarmnaya spoke the same language although there must've been some linguistic convergence and basal relatedness. Similarly not all neolithic Europeans were Proto Basques

Krefter said...

@Maju,
"IMO what you say is a relic, via wishful thinking, of a hypothesis that does not work no matter how you look at it. I mean: how do you link M412 to Eastern Europe? You can't even force-feed it, let alone make it look parsimonious. So..."

R1b-M412's ancestor can be from East Europe and therefore R1b-M412 does have something to do with East Europe. No one is disputing it expanded within West Europe.

The R1b-L23 dominance in Central European Bell Beaker and Yamnaya, and the former having 50% Yamnaya-like blood gives plenty of room for R1b-M412 to be originally from East Europe.

I'm currently 50/50 on a local/Eastern origin. And I've commented on this before, but I think you trust modern Y DNA too much.

The whole idea that diversity=origin is flawed. It doesn't always work that way. Sometimes when there's a migration all of the clades move into the new territory and some don't stay in their homeland. So, 1,000 years later the homeland will have less diversity. Also, lots of diversity can be erased overtime.

And when we have a small data set(like I assume we do of high coverage R1b and R1a in all of West Eurasia) we should be especially hesitant of assuming the region that has slightly more diversity is the place of origin. We need Ancient Y DNA. With modern Y DNA we can only guess.

Maju said...

"Similarly not all neolithic Europeans were Proto Basques"...

Let's see: we have a genetically homogeneous people expanding in around 1000 years from a common source over a different aboriginal population (with whom they sometimes mixed but often not even that) and you're telling me that they did not speak the same language family? It's possible but it's not very parsimonious you have to admit. More so when a bunch of linguistic data does suggest a Vasconic pre-IE substrate all around.

Maju said...

"R1b-M412's ancestor can be from East Europe"...

Might, yes. Can is an overstatement here. But it's much more parsimonious that the ancestor is from West Asia or the Balcans, where frequency and diversity fit much better. Morelli 2010 provided some arguments for a Balcanic origin, Myres studies also rather support that one or a West Asian one, etc. What do you have against that? Nothing: just a quite specific derived branch in the Volga, about the only place of Eastern Europe where R1b is more than a mere anecdote today as well.

"We need Ancient Y DNA".

We'd be safer with it but at the rate good quality Y-DNA is produced we can grow old waiting, it seems. Scarce, obliquely related ancient Y-DNA will hardly be enough to replace modern references. Considering modern distribution is not anyhow a mere "guessing" exercise but studying the bulk of the data we are likely to have in many years or even decades. Ancient Y-DNA can complement it but it can hardly become so abundant as to replace it as primary informative source.

«Oldest ‹Older   1 – 200 of 403   Newer› Newest»