Ancient Genomes suggest Basque are descended of Late Neolithic Iberians

I'm not totally comfortable drawing a lot of definite conclusions from this Admixture analysis done by Genetiker. I'm not saying it's not an accurate representation of what the calculator shows, mind you. Does anyone know, by the way, who created this calculator, or how accurate it's been in analyzing high quality ancient genomes?

Which brings me to the point that some of these genomes, in particular ATP 3, are not high quality at all.

Also, I find the "African", not "North African" percentages in some of these samples quite startling:

ATP3 3500 BC, 2.05 WEST Negroid

ATP2 2900 BC, 12.47 W. Negroid, 1.42 E Negroid, 14.11 Bushman

ATP20 2300 to 2000 3.5 W Negroid, 4.46 E Negroid

ATP 9 Mid Bronze 1700 2.42 W Negroid, 1.49 E Negroid

I mean, this 2900 BC sample looks at least a quarter Sub Saharan. Do we think that's possible? Ancient EEF sometimes throw up a few percent SSA, like Otzi, for example, which might be noise given the age of the samples, but this much? It's true that the authors themselves say that they found evidence of African in these samples, but they sort of waffle on how much. If this calculator is accurate would it mean there was that much SSA in parts of Iberia from the time of the late Neolithic? Might it represent a movement up the west coast of Iberia or did some migration from the east stop off in North Africa and incorporate some SSA heavy people? I don't know. That's way too much speculation for me based on an amateur (?) calculator using less than optimum quality genomes.

Anyway, if we are going to try to get some clues from the calculator results, I think it may be more helpful to group them by time period and to show at least three of the most important components, Northern Middle East, Northern Euro, and "veddoid".

The results are in that order: Northern Middle East, Northern Euro, and Veddoid

ATP3 3500-3300: 31.97/14.04/3.82

ATP7 3300-2900: 0/18.73/3.82
ATP16 3200-2900: 8.17/12.65/0
ATP17 3000-2800: 11.14/7.33/0
Matojo 3000-2900: .01/15.97/.11

ATP2 2900-2600: 2.60/12.60/0

ATP20 2300-2000: 0/18.92/28.14

ATP9 Mid-Bronze 1750-1600: 3.21/26.32/0

Looking at the "North Euro" score, the only really clear pattern I see is that it is consistently higher in the period from 2300 BC, when it's 18.92, to 1700-1600 BC, when it reaches 26.32%. That makes sense to me. I think this component was always present because it's related to "Euro" hunter-gatherers, but it increased when there were movements into Iberia starting around 2000 with the beginning of the Bronze Age. ( I'm not sure what to make of the 14.04 in ATP3, but given it's one of the worst samples, I think caution is advised. You can also see that this score fluctuates wildly in the big chunk of Chalcolithic people from 3300 to 3,000, with one sample at 18.73, and another at 7.33.)

The "Veddoid" component is absent in the vast majority of these samples, including in the Bronze Age sample which we would think would best represent the new "Indo-European" element. Then we have the 3.82 in the ATP 3 sample, which is the oldest, and an anomalous 28.14 score in ATP 20 from 2300 to 2000 BC, the same sample that has absolutely 0 Northern Middle East. Either this component is very difficult to pin down, or the bad quality of the genomes makes it impossible to pin it down, or this calculator isn't very good at capturing it.

That leaves the Northern Middle East component, which really only makes up a substantial portion of the ATP 3 sample. (low quality as we said) It's at an extraordinary 31.97 level. The Chalcolithic block from 2200 to 3000 BC varies from 0 to 11.14. By 3000 BC it's way down in the single digits.

If I were forced to make a prediction, I would say that if ATP 3 is really R1b M269 then it represents a movement of at least some very Northern Middle Eastern like R1b people from the east into Iberia by 3500 BC (or earlier), whose signal was diluted by intermarriage with locals and then by mixing with "Indo-Europeans" entering Iberia from the north in the Bronze Age.

I'd also say that the Bronze Age invaders of Iberia had very little "Indo-European" left in them by that time if you define "Indo-European" as the Yamnaya.

All of this is only if I were FORCED to make a prediction, mind you. :) I'm not comfortable making predictions based on this kind of evidence.
 
a simple question to put things in perspective : does ATP3 has ofspring today?
my rough guess : any random individual > 5000 yo has max 10 % chance to have ofspring today
 
Angela said:
I'm not totally comfortable drawing a lot of definite conclusions from this Admixture analysis done by Genetiker. I'm not saying it's not an accurate representation of what the calculator shows, mind you. Does anyone know, by the way, who created this calculator, or how accurate it's been in analyzing high quality ancient genomes?

Which brings me to the point that some of these genomes, in particular ATP 3, are not high quality at all.

Also, I find the "African", not "North African" percentages in some of these samples quite startling:

ATP3 3500 BC, 2.05 WEST Negroid

ATP2 2900 BC, 12.47 W. Negroid, 1.42 E Negroid, 14.11 Bushman

ATP20 2300 to 2000 3.5 W Negroid, 4.46 E Negroid

ATP 9 Mid Bronze 1700 2.42 W Negroid, 1.49 E Negroid

I mean, this 2900 BC sample looks at least a quarter Sub Saharan. Do we think that's possible? Ancient EEF sometimes throw up a few percent SSA, like Otzi, for example, which might be noise given the age of the samples, but this much? It's true that the authors themselves say that they found evidence of African in these samples, but they sort of waffle on how much. If this calculator is accurate would it mean there was that much SSA in parts of Iberia from the time of the late Neolithic? Might it represent a movement up the west coast of Iberia or did some migration from the east stop off in North Africa and incorporate some SSA heavy people? I don't know. That's way too much speculation for me based on an amateur (?) calculator using less than optimum quality genomes.
Click to expand...


I think some of it represents very archaic West Eurasian(farmer?) genes in SSA populations which we didn't knew off.
Logically the more ancient the genes the closer to the roots also.

If I were forced to make a prediction, I would say that if ATP 3 is really R1b M269 then it represents a movement of at least some very Northern Middle Eastern like R1b people from the east into Iberia by 3500 BC (or earlier), whose signal was diluted by intermarriage with locals and then by mixing with "Indo-Europeans" entering Iberia from the north in the Bronze Age.
Click to expand...

What if those R1b with high teal represent a Indo European population not yet deluted by EHG ancestry in the Steppes? This would fit one of my theories that Yamna was a secondary homeland for the Indo European coming with the teal admixture from somewhere else.
 
Alan said:
I think some of it represents very archaic West Eurasian(farmer?) genes in SSA populations which we didn't knew off.
Logically the more ancient the genes the closer to the roots also.



What if those R1b with high teal represent a Indo European population not yet deluted by EHG ancestry in the Steppes? This would fit one of my theories that Yamna was a secondary homeland for the Indo European coming with the teal admixture from somewhere else.
Click to expand...

They might represent a population that fed into the steppe people, but I don't think they can be called the "Indo-Europeans" based on the data we have so far.

I still think that the "Indo-Europeans" were the Yamnaya people, although I'm ready to be persuaded otherwise by new data.
 
Good "paper" Angela
I agree for the most; let's not forget today Basques is not the same as ancient geographically "Basques";
that said, I find very boring all these abstracts where we always find unprecise geographic terms as western or eastern and so one. Always I've the impression people are trying to sale washing stuff. And I cannot believe in pure EEF in Iberia nor in any other place. Other paper said Iberians Neolithic people were close to Center-Eastern Europe and they shared just a taste of (W)HG, but the same paper shows the Middle Neolithic people of Iberia (Spain, Cataluna and other places) are more "westwards" shifted and closer to old WHG than Early and Middle Neolithic men of Central and Eastern Europe, even more shifted than today Sardinians, at least in some plotting (I know, plottings ...) - so the HG admixture increase by time, or Iberian Neolithic people were not exactly the same as other Farmers, even if they shared a lot of common ancestry. concerning this (W?)(S?)HG we have to wait to be sure (question of DNA coverage?)...
for Y-R1b I have no religion even if I favour the Steppes orgin. But a Megalitihic origin (4000 BC) with subsequent Atlantic Bronze is not to be discarded. From where through where??? a 3000 BC alley of Neolithic megalithers and first I-Eans Y-R1b (doped by BBs?) could explain the demographic increase of this period and the later Atlantic Bronze
The today Basques show litlle of it but they have ANE, they have 'gedrosia', they share some mt-H and mt-U with North-Eastern European people, they have more HG in them: saying after that "Basques descend almost only from Neolithic people from Near-East" or something like that is going very far; or we are not giving the same meaning to names? We lack aDNA from Portugal (see the surveys showing already differences between Cantabria and Basque country ancient mt-DNA, mt-H stronger in West from Late Paleolithic but well present (about 50%) in Basque Mesolithic: the "neolithical' mtDNA came rather during Chalcolithic than Neolithic and more marked in the Ebro river region than in North stayed "archic"; mt H augmented again in Middle Ages at the depends of mt-U but also at the depends of "neolithical" mtDNA!
 
I would say that European levels for 42A5 have risen everywhere, even in the far northeast, which would go to Mathiesen et al's point about selection within the last 5,000 years.[/QUOTE]



Interesting Angela, but what about the other: SLC 24A5? These so short time of selection seems to me a bit curious; the selection concept is confusing sometimes: if a mutation on a locus is enough to strongly lighten the skin, why an other mutation with the same effect would be so imperiously needed and selected? Selection is complicated in fact: but we can also figure out an other natural selective pressure acting upon a gene or genes, with different "task" but very close on the same chromosome? Linkage?
 
bicicleur said:
afaik the walls of Los Millares collapsed and were rebuild 3025 BC
dates for the 1st construction of Los Millares are unknown

probably explorers where at first attracted by alluvial copper which maby natives had allready found
after they would have been looking for copper ores

don't know whether in Atapuerca it was the same situation
Click to expand...

one of the things I was thinking about early copper working is say for the sake of argument there were five levels of difficulty in extracting copper with level one sites being panning for it in streams or simply chiseling it out of exposed rock and levels two to five being progressively more difficult to reach then the first level one dudes might have traveled far and wide looking for level one sites before they were able to extract it from level two difficulty sites - and then the same again between difficulty level two and three.

A process like that might have made the spread very erratic.
 
MOESAN said:
"Angela: I would say that European levels for 42A5 have risen everywhere, even in the far northeast, which would go to Mathiesen et al's point about selection within the last 5,000 years."



Interesting Angela, but what about the other: SLC 24A5? These so short time of selection seems to me a bit curious; the selection concept is confusing sometimes: if a mutation on a locus is enough to strongly lighten the skin, why an other mutation with the same effect would be so imperiously needed and selected? Selection is complicated in fact: but we can also figure out an other natural selective pressure acting upon a gene or genes, with different "task" but very close on the same chromosome? Linkage?
Click to expand...

One possibility might be a mashup of the various theories.

Say lighter (but not light) skin had an advantage in northern Eurasia (or interior northern Eurasia) and so multiple depigmentation genes developed among separate populations in that region.

Say for the sake of argument there were five separate ones in total and they'd spread around a bit for the same reason they developed or a different reason until say most of the people who had them had 3 out of the 5.

Then farming either magnified the (or one of the) original selection pressures or added a second (or third) selection pressure on top.

Just an idea.
 
Last edited:
Interesting Angela, but what about the other: SLC 24A5? These so short time of selection seems to me a bit curious; the selection concept is confusing sometimes: if a mutation on a locus is enough to strongly lighten the skin, why an other mutation with the same effect would be so imperiously needed and selected? Selection is complicated in fact: but we can also figure out an other natural selective pressure acting upon a gene or genes, with different "task" but very close on the same chromosome? Linkage?[/QUOTE]

I don't recall any paper saying that one depigmentation mutation was enough to lighten the skin to, say, "European fair" levels. Rather, the way I understood it was that each one acted on a different part of the melanin making or transportation process, or rather with the disruption of the melanin making process, with each contributing a certain amount of the final depigmentation result.

In other words, the effect was cumulative.

These are two links that I just happened to have in my files. I've posted lots more on various threads.

https://en.wikipedia.org/wiki/SLC24A5

http://www.pubfacts.com/detail/2605...Melanosomal-pH-and-Influences-Tyrosinase-Acti

As for the Mathiesen et al paper, the time period discussed was actually the last 8,000 years. I don't see it as so terribly surprising that in 8,000 years a certain mutation could rise to fixation. Look at the lactase persistence gene. We find it in one Bell Beaker person in central Europe and in a few people in Iberia, and it's now at huge frequency levels in central, northern and northwestern Europe.

I'm not saying that it was all selection, by the way. I'm just saying I don't see why selection couldn't have had a part in it after the random mutation and perhaps some initial admixture with people bearing it.

I also think that the selection in terms of an 8,000 year time span looks more like an increase in SLC42A5 to me, as you say. SLC24A5 derived was at very high levels even in the earliest Neolithic. Given that the derived snps are fixed in the Near East as well I think there's some chance that it arose or at least was being selected for quite a bit before 8,000 years ago, just not in all human groups apparently.

It will be very interesting to see the phenotype snps in the ancient Anatolian farmers. If they weren't derived for SLC24A5 then the farmers must have picked the mutation up on the way into Europe. Then selection would take over.

Wait, didn't the abstract for an upcoming paper on the Greek Neolithic say that they were "dark"? Of course, we don't know what that means precisely, but maybe the farmers did pick up some depigmentation snps (SLC 42A5? SLC24A5? Both?) somewhere in Europe.

There's no percentage in spending too much time speculating. We should know in two weeks.
 
@Angela,

The "dark skin" report in Greek Neolithic is probably a miss interpretation from the media. They've miss interpreted authors before, because it's not an interest/hobby of theirs. By Dark skin my guess is Greek Neolithic had little 374F.
 
Angela said:
Does anyone know, by the way, who created this calculator, or how accurate it's been in analyzing high quality ancient genomes?
Click to expand...

I suspect Genetiker probably made it. That's just my opinion based on reading his blogs, though. And I think it was just recently made, and will have to stand the test of time as to how well it will hold up. I'd much like to see these new genomes' results on Dienekes' and Davidski's calculators.

Anyway, awesome find! It sheds light on how accurate or inaccurate the history we've been told can be. Maybe the Bell Beakers spread R1b-L11 and much of our Western European languages/language components from Iberia.

Also perhaps of note is the Southern Middle Eastern component in ATP3 of 3.55%. I guess it may or may not have been brought there by R1b. Part of Gedrosia?
 
Davidski is doing analysis of ATP3 and suspects he has some ANE. In Davidski's PCA of West Eurasia ATP3 clusters by Northern Spanish not Neolithic farmers. He'll probably post more on analysis of all El Portalons in the net few weeks.
 
Fire Haired14 said:
Davidski is doing analysis of ATP3 and suspects he has some ANE. In Davidski's PCA of West Eurasia ATP3 clusters by Northern Spanish not Neolithic farmers. He'll probably post more on analysis of all El Portalons in the net few weeks.
Click to expand...

If Atp3 has modern Northern Middle Eastern/Caucaso-Gedrosia genes, of course he will have ANE, Since it's the decisive difference between western and eastern/teal farmrs. No need for suspections there.
 
Last edited:
@Angela: OK, maybe I rised too quickly my voice!
but It's pity I've not the today distribution of mutated genes at SCL45A2
I've only maps about CL24A5 111*A, OCA2 355*, ASIP 8818*A, MATP 374*C and TYR 192*A
concerning skin colour, the only making total sense among these 5 is the SCL24A5, typically "europoid" and checking very well perceived skin pigmentation distribution -
none of the others 4 show a strong correlation to skin pigmentation among Europoids so they could be of little importance in depigmentation among them but ASIP 8818*A seems having a strong enough effect for light skin among 'mongolid' East Asians which lacked SCL24A5 mutated for the most- (the higher scores among Amerindians: North Siberian in origin?)
ASIP 8818 mutated is found too at low enough levels in some SSA groups as OCA2 mutated - this OCA2 mutated is strong enough among East Asians (not too much in America) and it seems that the cumulative effect of ASIP and OCA2 mutated genes was necessary to lighten the East Asians skin, even it this lightening is not completley as strong as among european 'europoids' (confirmed by the fact SSA people are not very fair skinned spite they have a bit!): so I suppose these 2 last mutations had a very less strong imput on skin pigmentation, compared to SCL24A5 mutation.
all that doesn't tell us anything about the SCL45A2 mutation penetration, helas for our question;
selection? YES! but principally for what mutated gene? ( Orcadians (fair skinned I suppose) and North Russians with lower %s of mutated genes at ASIP, OCA2, and TYR than some southern Europeans and others darker skinned groups...

to come back to the thread: all that SSA DNA in Iberia seems a problem of calculator at first sight even if we cannot discard other but surprising explanations (very mobile elite) - an archaic DNA seems amazing at these dates at a so high level!)
sorry it's the my aperitive time! Good lick!
 
MOESAN said:
@Angela: OK, maybe I rised too quickly my voice!
but It's pity I've not the today distribution of mutated genes at SCL45A2
I've only maps about CL24A5 111*A, OCA2 355*, ASIP 8818*A, MATP 374*C and TYR 192*A
concerning skin colour, the only making total sense among these 5 is the SCL24A5, typically "europoid" and checking very well perceived skin pigmentation distribution -
none of the others 4 show a strong correlation to skin pigmentation among Europoids so they could be of little importance in depigmentation among them but ASIP 8818*A seems having a strong enough effect for light skin among 'mongolid' East Asians which lacked SCL24A5 mutated for the most- (the higher scores among Amerindians: North Siberian in origin?)
ASIP 8818 mutated is found too at low enough levels in some SSA groups as OCA2 mutated - this OCA2 mutated is strong enough among East Asians (not too much in America) and it seems that the cumulative effect of ASIP and OCA2 mutated genes was necessary to lighten the East Asians skin, even it this lightening is not completley as strong as among european 'europoids' (confirmed by the fact SSA people are not very fair skinned spite they have a bit!): so I suppose these 2 last mutations had a very less strong imput on skin pigmentation, compared to SCL24A5 mutation.
all that doesn't tell us anything about the SCL45A2 mutation penetration, helas for our question;
selection? YES! but principally for what mutated gene? ( Orcadians (fair skinned I suppose) and North Russians with lower %s of mutated genes at ASIP, OCA2, and TYR than some southern Europeans and others darker skinned groups...

to come back to the thread: all that SSA DNA in Iberia seems a problem of calculator at first sight even if we cannot discard other but surprising explanations (very mobile elite) - an archaic DNA seems amazing at these dates at a so high level!)
sorry it's the my aperitive time! Good lick!
Click to expand...

Well, it's mine in a few minutes (Although I have to cook as I drink it!) so I'll be brief. :)

I posted the derived and ancestral results for SLC42A5 for the ancient samples upthread.

This is a chart from Lucotte et al of modern percentages of derived SLC42A5 by city. Not great, but better than nothing.
Lucotte et al pigmentation data.jpg
Lucotte et al pigmentation data part 2 374 or 42A5.jpg
.
There is a map of the distribution of derived SLC42A5 in this post by Razib Khan:
http://blogs.discovermagazine.com/gnxp/2013/03/pigmentation-the-simplest-of-complex-traits-not-so-simple/#.VfNEm5dvBT8

Ed. The Lazaridis paper will be read in a few weeks so we'll know whether the farmers came into Europe with derived SLC24A5


 
Alan said:
If Atp3 has modern Northern Middle Eastern/Caucaso-Gedrosia genes, of course he will have ANE, Since it's the decisive difference between western and eastern/teal farmrs. No need for suspections there.
Click to expand...

I really don't think tremendous weight should be put on such a low coverage genome. It might give us some hints, but that's about it, in my opinion. That's probably the reason why the authors didn't even include ATP3 on their PCAs.

As to the source of the ANE, you may indeed be correct. I tend to think that people in the northeastern Near East may have had ANE before the arrival of any people from the steppe. Perhaps some of it, in fact, went onto the steppe from that region. When the Reich Lab gets done with all those samples from the Caucasus and surrounding areas we should have a much clearer picture.

I've also felt for a long time that later migrations of farmers into Europe from the Near East might have carried a bit of it as the ANE admixed Near Easterners moved slowly west. I can't remember now, but didn't Otzi have a smidgeon of West Asian in one of the calculators? Also, if people are so eager to do runs on ancient genomes, why has no one analyzed Barcin?

Therefore, I don't think it's a given that the ANE in this sample necessarily came from an EHG person. I'm also disinclined to give temendous weight to amateur generated analysis given all the confusion about ENF versus EEF. It would have been much better, in my opinion, to have waited for an actual genome from Anatolia.

IF this person is an admixed individual, part new arrival, part "local" Neolithic type, and not a migrant himself, the "father" or actual migrant must have had a heck of a lot of northern Middle East.

As to how he got to Iberia, I have some doubt that this individual, even if it's proved that he has a "Yamnaya" kind of autosomal signature, would have ridden a horse from the steppe across a mostly forest covered European landscape at a time when the horse hadn't even been domesticated yet, or was just being domesticated (see Anthony for a date about 3500 BC), and the later Corded Ware people still often used oxen to pull their primitive carts. Not to mention the fact that they would have had no reason to go due west on the European landmass to Iberia, leaving no trace of their passage in the Copper Age cultures of central Europe.

I think it's probably more likely that any travelers from the east would have sailed along already well established sea routes, used in the copper trade and prospecting at the time, and by interlocking obsidian networks before that, and the neolithic farmers before that. The sea routes were always the same because the winds and sea tides were generally the same.

Given all of that, I would tend to think that this original migrant, if he came around or just prior to this time, would more likely have departed from somewhere around the Balkans, or possibly the Aegean, perhaps with the collapse of the copper age cultures around the Balkans from either environmental collapse and/or pressure from the steppe. That's if he didn't arrive even earlier...

As to the objections that he's too early for the Copper Age in Iberia, I really don't see that. I did a little digging and there are some new texts on the whole topic of metallurgy.
https://books.google.com/books?id=Q...nt ortiz et al 2003 metallurgy/copper&f=false

They don't contradict previous findings that intensive copper ore extraction was going on at Monte Loreto in Liguria from the mid fourth millennium BC (Maggi and Pearce, 2005). Copper objects are found in Arene Candide in the late 5th millennium BC. Copper production, not objects, is apparent in Sardinia at the same time, in the latter part of the 5th millennium.

In Iberia itself, the mine at isn't attested until . However, the author of the above text maintains that copper smelting was going on in Iberia in the late 4th millennium BC going on until the early 3rd millennium BC. There is even fragmentary evidence of copper oxide slag at Cerro Virtud in SE Spain radio carbon dated to the first half of the 5th millennium BC (Montero Ruiz 1999) The author does recommend caution with this result because it is one millennium earlier than many of the others. Unfortunately, it seems to be the case, according to him, that likely sites are not being actively investigated, and/or are not secure.
http://www.qub.ac.uk/sites/EarlyMiningintheLowerSeguraValleySESpain/Image,280457,en.jpg

If any analysis is done despite the fact it's so low coverage I hope it's compared to a lot of ancient genomes to get a better handle on this, including the prior Portalon genome that plotted near modern day Tuscans, especially if that's a different genome than the ones analyzed in this paper. Actually, if it's the same genome or genomes, we need to know why it's plotting differently, if in fact it is....
 
@Everyone,

BTW, my Dad lacks the two "Lactose Tolerant" mutations and he's drank milk his whole life with no problems. No one has questioned the effect of those mutations. There are certainly other factors involved. The same is true for skin color. More factors are involved than the SNPs rs1426654 and rs16891982. I'm sure they make decent affects though.

The easiest way to figure it out is looking at ancestry. All Europeans are mostly EEF/WHG+Steppe, including Greeks. West Asians appear to have little of such ancestry. So, somewhere in the EEF/WHG and Steppe world is where the Light skin comes from. The confusing thing is EEF/WHG and Steppe were as differnt from each other as West Asians and Europeans are today. So, it's hard to imagine it comes from both, but maybe.

It's probably a very complicated origin. Because Basque for example have Light skin and little Steppe-blood and Scythian had Light skin and had little EEF/WHG ancestry.
 
The author does recommend caution with this result because it is one millennium earlier than many of the others. Unfortunately, it seems to be the case, according to him, that likely sites are not being actively investigated, and/or are not secure.
http://www.qub.ac.uk/sites/EarlyMini...,280457,en.jpg
Click to expand...

Interesting on that map that the gold and silver mining sites - which I assume came first - are located away from the copper sites in the Pyrenees implying that miners would have had to move to those sites from elsewhere - maybe from Iberia, maybe not.

copper rush? miner 4049-er (BC)
 
Last edited:
You must log in or register to reply here.
Back
Top