The spread of 'Steppe' DNA and autosomal best-fit analysis

[Ygorcs]

Unless it's my tiredness kicking in, do people not realise how important this is?! This is R1b-L151, in Southern France no less (and by the Rhône no less too), with absolutely no sign of Steppe.

Instead, it has a lot of WHG, and a lot of ENF too. Rather Basque-like.

Doesn't it have way too much WHG to be comparable to Basques or any modern population in fact? I'm not sure, but considering WHG is not differentiated from EHG, and EHG itself already carried a minor percentage of ANE (I think ~10%) amidst a whole lot of WHG-like admixture, it is quite likely that this calculator just does not get the EHG signal that could exist in the sample especially if it's mixed with additional WHG. I think other models would be better, this one seems more prone to confuse people more than clarify things, the labels are just so broad, and the proxy populations so vague...

In any case, I do think this sample is very interesting and merited deeper investigation (what paper published about it? Maybe we could try to find their interpretation of the data).

 

[ToBeOrNotToBe]

Doesn't it have way too much WHG to be comparable to Basques or any modern population in fact? I'm not sure, but considering WHG is not differentiated from EHG, and EHG itself already carried a minor percentage of ANE (I think ~10%) amidst a whole lot of WHG-like admixture, it is quite likely that this calculator just does not get the EHG signal that could exist in the sample especially if it's mixed with additional WHG. I think other models would be better, this one seems more prone to confuse people more than clarify things, the labels are just so broad, and the proxy populations so vague...

In any case, I do think this sample is very interesting and merited deeper investigation (what paper published about it? Maybe we could try to find their interpretation of the data).

This was from one of the Olalde Beaker papers; it didn't mention anything about it I don't believe. How could this be from a once Steppe-heavy R1b L51? It would take an impossible level of dilution in such a short time period to reach this low level of Steppe.

That being said, even in my scenario (not going to explain it for the millionth time, you know it already, though I've modified it slightly after reading Los Millares, despite being the oldest copper smelting site in Western Europe, had surprisingly sparse Beaker pottery finds compared to somewhat later copper smelting sites in Southern Portugal), I'd expect a teensy little bit more Steppe. Perhaps this sample has more Steppe than it seems, but it clearly is still too low for the likes of Yamnaya or Corded Ware. That extra level of dilution from my hypothesis is feasible, but certainly not so from being very Steppe-like to Basque/Sardinian-like in a matter of generations.

It is a clear outlier, though, and probably not enough to convince against the mountain of Steppe R1bs genotyped. But it's a start. I'd love to compare it directly against a "pure" Megalithic Spanish I2 type, but I can't find the GEDmatch ID of any.

 

[Ygorcs]

This was from one of the Olalde Beaker papers; it didn't mention anything about it I don't believe. How could this be from a once Steppe-heavy R1b L51? It would take an impossible level of dilution in such a short time period to reach this low level of Steppe.

That being said, even in my scenario (not going to explain it for the millionth time, you know it already, though I've modified it slightly after reading Los Millares, despite being the oldest copper smelting site in Western Europe, had surprisingly sparse Beaker pottery finds compared to somewhat later copper smelting sites in Southern Portugal), I'd expect a teensy little bit more Steppe. Perhaps this sample has more Steppe than it seems, but it clearly is still too low for the likes of Yamnaya or Corded Ware. That extra level of dilution from my hypothesis is feasible, but certainly not so from being very Steppe-like to Basque/Sardinian-like in a matter of generations.

It may be so, but I'm not so sure such a dilution is that improbable for a sample whose average dating is ~2250 B.C., that is, well into the BB period and centuries after the beginning of its expansion. If, as I personally believe, the L51-rich steppe-derived population that spread to Central Europe already arrived there with a higher EEF input, then it doesn't strike me as an "impossible level of dilution" in a not really "short time".

Theoretically, if this population was actually EEF with some occasional exposure to the BB males (a past invasion? Mercenaries? Enslaved prisoner of war?), but not destroyed by the new waves nor heavily transformed by them yet, and it remained mostly EEF, it's not that implausible that someone might have L151 and also a very low steppe admixture, especially if his Early BB ancestor was already ~50% steppe as many other BBs. If all subsequent generations involved fully EEF partners, it'd take only 5 generations (~100-125 years) for the steppe component to be reduced to a tiny 1.5% (the amoutn of ANE would then become almost invisible).

It'd be very nice if they analyzed other samples from the same area, especially since most other Beakers had much higher steppe ancestry, so this one is a much more interesting outlier.

 

[ToBeOrNotToBe]

It may be so, but I'm not so sure such a dilution is that improbable for a sample whose average dating is ~2250 B.C., that is, well into the BB period and centuries after the beginning of its expansion. If, as I personally believe, the L51-rich steppe-derived population that spread to Central Europe already arrived there with a higher EEF input, then it doesn't strike me as an "impossible level of dilution" in a not really "short time".

Theoretically, if this population was actually EEF with some occasional exposure to the BB males (a past invasion? Mercenaries? Enslaved prisoner of war?), but not destroyed by the new waves nor heavily transformed by them yet, and it remained mostly EEF, it's not that implausible that someone might have L151 and also a very low steppe admixture, especially if his Early BB ancestor was already ~50% steppe as many other BBs. If all subsequent generations involved fully EEF partners, it'd take only 5 generations (~100-125 years) for the steppe component to be reduced to a tiny 1.5% (the amoutn of ANE would then become almost invisible).

It'd be very nice if they analyzed other samples from the same area, especially since most other Beakers had much higher steppe ancestry, so this one is a much more interesting outlier.

That's possible, I guess they just need to sample more. But even in that case, how can his WHG be so high? Perhaps the calculator has misinterpreted EHG?

 

[ToBeOrNotToBe]

Turns out I am REALLY good at Googling - I've found some data that nobody else has found (it seems). Not related to Beakers or anything to do with my crazy (though I think they're well-founded) theories.

 

[ToBeOrNotToBe]

Oh boy, I'm pretty certain I wasn't meant to see this, one of them at least doesn't seem to be released

And I'm not t-rolling. Just give me a day to look through this. I'll post it in a new thread of course.

EDIT: So many wrong SNP calls, what a killjoy. Still a few things though

 

[ToBeOrNotToBe]

I've found one mistake in it, but almost all of it is accurate (EDIT: Past me, there are loads more, you look like an idiot now). And as mentioned, some of this data is unpublished. I'm not some hacker or anything, I just stumbled upon it on Google (and I can guarantee it's legitimate based on the website it was downloaded from).

 

[ToBeOrNotToBe]

Interesting paper on Los Millares' links to the Middle East:

https://www.researchgate.net/public...ssible_relations_to_the_Western_Mediterranean

"Symbolkeramik was long regarded as diagnostic for Los Millares I and Vila Nova de Sao Pedro I (Early Chalcolithic). The find contexts in the graves, however, suggest that the Symbolkeramik belongs mainly to an advanced phase of the Early Chalcolithic and the subsequent Bell Beaker period."

 

[Pip]

But today, the distribution of U106, and even ancient DNA, seems checking an East-West move in North Europe or at least North the Carpathians towards maybe Unetice, and Saale post-CWC region.

My analysis of U106, based on intra-SNP STR diversity over thousands of samples on an FTDNA database, suggests an initial most likely West to East move (Belgium to N Germany/Sweden/Baltic) with some subsequent drifts backwards. Perhaps early extant development associated with Battle Axe incursions?

 

[Pip]

Ok. Really interesting.
Sole-Morata et al 2017 shows something that is awkward for a east to west entry of R1b to the region of Bell beakers, Portugal.

Like in the Map of L51 you have shown, Portuguese people today have a level of M269 XP312 that is not at all "Iberian" but similar to levels only, only, found in Norwest France (Alsace, Nord-Pas-de-Calais) which is the French region that differ from the rest o France, where levels of M269 not P312 is very low indeed. That low level of L51 (?) is also true in TSI (Tuscany 1000 genomes). UK had the highest.

The point is ...

1. Portugal with 12% of M269-XP312 (L51?) is much higher than all other Spanish places, TSI (Tuscany area) and France.
2. Then, the level of Portuguese that are P312 xDF27 is the lowest. So, when P312 entered Portugal was already pretty much DF27.
   1. Note- All labs better start looking for Df27 on Europe ancient samples…
3. Spain is DF27 +Z196. Portugal is not. Portugal has lowest Z196 than even the rest of Europe were we find DF27! And we know that the TMRCA of the son Z196 is very similar to father DF27. They knew each other!

To me, it tells me that the story of L51 departing from the estuary TAGUS river is pretty plausible. We know that there were bell beaker originals and the ones returning centuries later. The ones with steppe admixture (DF27).

Do we know the precise SNP readings of the M269XP312 in Portugal? This would help answer the Portugal question.

 

[Pip]

Doesn't it have way too much WHG to be comparable to Basques or any modern population in fact? I'm not sure, but considering WHG is not differentiated from EHG, and EHG itself already carried a minor percentage of ANE (I think ~10%) amidst a whole lot of WHG-like admixture, it is quite likely that this calculator just does not get the EHG signal that could exist in the sample especially if it's mixed with additional WHG. I think other models would be better, this one seems more prone to confuse people more than clarify things, the labels are just so broad, and the proxy populations so vague...

In any case, I do think this sample is very interesting and merited deeper investigation (what paper published about it? Maybe we could try to find their interpretation of the data).

Good point. The WHG-UHG ("unknown" hunter gatherer) seems to include EHG (?), and what is called WHG ("Western" Hunter Gatherer) was in any case predominant across much of the Steppe.

A more interesting feature to me is the inherent variability of the CHG component within otherwise standard-looking Bell Beaker and Corded Ware samples. Some such samples have no CHG element whatsoever. It is difficult to explain, and suggests either that much of CHG was a very late addition to the mix or was initially present only in selective castes within proto-BB/proto-CW.

 

[Pip]

Oh boy, I'm pretty certain I wasn't meant to see this, one of them at least doesn't seem to be released

And I'm not t-rolling. Just give me a day to look through this. I'll post it in a new thread of course.

EDIT: So many wrong SNP calls, what a killjoy. Still a few things though

There have been plenty of mistakes on SNP calls in various studies - that's why I generally prefer using mass, modern, fully-published databases.

I also wonder whether there is a lot of data that we are not meant to see, so it would be interesting to see what you have found.

I'm not saying the L52* French sample fits into this category, but it is certainly the most striking R1b-L51 sample that has been published and it is a little curious that most of its data was removed from the internet once people started noticing it.

I will not give the full details, as I gather there might be a preference for anonymity. However, what I will say is that:
1. I am not aware of any other sample with SNP readings like it.
2. Its STR readings are extraordinarily diverse for L51, and (if not ignored as freak results) set back the estimated origin point for L51 by a millennium or so.
3. My assessment of its probable earliest-known paternal ancestry point is Autun in East Central France.
4. Its unique yDNA position has been cross-checked by two reputable sources (FTDNA and yfull).

Its location in France matches my coalescence point estimates for most of the other branches of L51.

Its diverse readings yield early predicted dates for western branches of R1b-M269, preceding the Bell Beaker expansions and the appearance of Yamnaya west of the Steppe.

 

[berun]

Before I get excited, why hasn't anybody mentioned this?

Sample ID / genetic sex (M/F): I1388 / M
Find location: Marlens, Sur les Barmes, Haute-Savoie
Country: France
Associated label in publication: Beaker Southern France
Date: 2455–2134 calBCE (3805±40 BP, Ly-3099)
MtDNA haplogroup (mother): H
Y-DNA haplogroup (father): R1b1a1a2a1a (L151)
Reference: Olalde et al. 2018
Colour group: Anatolian farmer (autosomal)
Comments: Clusters closer to Anatolian farmers in f-statistics (sup info p. 155)
Other references: null

​

I1388, Bell Beaker FR CA (2455–2134 BC), Kit T122731
1 Basque_Spanish + Basque_Spanish + Sardinian + Sardinian @ 3.817203

he was lumped by Reich & co. with other French BB as to provide enough steppe admixture into the group.

 

[markod]

Potentially very important to note about this sample, I didn't even expect this:

Population ANE - ASE - WHG-UHG 72.71 East_Eurasian - West_African - East_African - ENF 27.29

No ANE! Wow...

f4(Mbuti, X; Yamnaya_Samara, Anatolia_Neolithic) indicates that this sample is closer to Yamnaya than early Iberian beakers however.

 

[berun]

maybe by an extra WHG bonus?

 

[markod]

maybe by an extra WHG bonus?

Maybe, but to convince anyone it would require some rigorous testing and a detailed comparison with other Beaker samples.

 

[berun]

Savoy has a climatology that don't allows agriculture, so this BB would be surely a herder, and that implies a bonus in WHG just by simple maths, herders colonize a given area in much less numbers than farmers so they get more local HG admixture, look per example at Gauchos.

 

[halfalp]

Irrespective of what the early L51 population's language was, I think the it's the now more or less unavoidable conclusion regarding late linguistic change doesn't sit well with people. Those early western European Beakers certainly look like what we think of as 'Celtic' genetically, it's just extremely unlikely that they actually spoke Celtic languages. It's possible that the Celtic dialects replaced older Indo-European languages in Western Europe of course.

At the same level, most Beakers are P312 wich is the brother of U106. So if U106 brought IE languages to become the Proto-Germanic, it's likely to assume that P312 spoked an IE languages, like Maciamo hypothetized years ago as Proto-Germano-Italo-Celtic. Following R. Rocca from Anthrogenica, most Eastern Beakers had positive calls for S28 -> L2 wich means it's very likely that British and Dutch Beakers spoked already an IE languages. The differentiation between Celtic, Germanic, Italic and probable languages that were transitional between those like Celtic > Ancient Belgian > Germanic or Italic > Ligurian > Celtic might confuse us in their spatial distribution at the times of the Bell Beakers. What i mean is that we dont know for a certainty when Celtic became Celtic etc. Keep in mind that Archaic Latin was very close to Celtic languages to be intercomprehensible until Classic Latin put a little bit of distance between them. Also there is the question of the role of Unetice into the IE languages contexte, wich had huge cultural exchanges with Wessex wich is related with Bell Beakers no? In conclusion, things seems to be clear that already at 2500 BC most of Western Europe was under the influence of an IE lingua franca who little by little differentiate between different dialects.

 

[markod]

At the same level, most Beakers are P312 wich is the brother of U106. So if U106 brought IE languages to become the Proto-Germanic, it's likely to assume that P312 spoked an IE languages, like Maciamo hypothetized years ago as Proto-Germano-Italo-Celtic. Following R. Rocca from Anthrogenica, most Eastern Beakers had positive calls for S28 -> L2 wich means it's very likely that British and Dutch Beakers spoked already an IE languages. The differentiation between Celtic, Germanic, Italic and probable languages that were transitional between those like Celtic > Ancient Belgian > Germanic or Italic > Ligurian > Celtic might confuse us in their spatial distribution at the times of the Bell Beakers. What i mean is that we dont know for a certainty when Celtic became Celtic etc. Keep in mind that Archaic Latin was very close to Celtic languages to be intercomprehensible until Classic Latin put a little bit of distance between them. Also there is the question of the role of Unetice into the IE languages contexte, wich had huge cultural exchanges with Wessex wich is related with Bell Beakers no? In conclusion, things seems to be clear that already at 2500 BC most of Western Europe was under the influence of an IE lingua franca who little by little differentiate between different dialects.

As I said it's possible that most of Western Europe was already IE speaking by the early Bronze Age considering how overwhelming the demic impact of BB was. But those languages are now lost to us, wiped out by the Celts in the Iron Age and the Roman expansion.

I dimly remember a Roman source that claimed that the Celts were actually a minority in Western Europe, but I can't locate it for the life of me. I think it wasn't Gaius Iulius.

 

[halfalp]

As I said it's possible that most of Western Europe was already IE speaking by the early Bronze Age considering how overwhelming the demic impact of BB was. But those languages are now lost to us, wiped out by the Celts in the Iron Age and the Roman expansion.

I dimly remember a Roman source that claimed that the Celts were actually a minority in Western Europe, but I can't locate it for the life of me. I think it wasn't Gaius Iulius.

I dont know exactly what ancient romans or greeks considered by their times as Celtic or Celts. I know Gallia Celtica and Celtiberians. But where Brittons and Belgians or Helvetii considered as Celts by romans?