search this blog

Wednesday, July 9, 2014

More ancient genomes from Sweden: Pitted Ware forager Ajvide58 and TRB farm girl Gokhem2


Both of these genomes were published earlier this year by Skoglund et al. 2014 (see here).

My analysis shows that Ajvide58 is very similar to Mesolithic Swedish forager StoraFörvar11 (see here), and also in part Ancient North Eurasian (ANE). This can be seen in the Principal Component Analysis (PCA) of West Eurasian populations, in which Ajvide58 is shifted east relative to La Brana-1 (which lacks ANE), much like StoraFörvar11.

However, the Eurogenes K15 results suggest to me that the level of ANE in Ajvide58 is lower than in StoraFörvar11. That's because Ajvide58 shows less of the Eastern European component (17.54% vs. 23.23%), and none of the South Asian component. These two components, along with the Amerindian component, dominate MA-1's K15 results (see here).

On the other hand, Gokhem2 appears not to harbor any ANE ancestry; note it's extreme western shift on the PCA of West Eurasia and complete lack of the Eastern Euro, Amerindian and South Asian components in the Eurogenes K15. This is in line with all scientific literature to date, which indicates that ANE was basically missing from Western and Central Europe during the Mesolithic and Neolithic. Indeed, this sample's best matching population in the Oracle are the Sardinians, one of the few present-day European groups with only a trace amount of ANE.

The absence of ANE in Gokhem2 and all other ancient European genomes from a farming context, like Stuttgart and Oetzi, is a very important point. That's because Neolithic farmers largely replaced indigenous hunter-gatherers across most of Europe, including in Scandinavia. As a result, it's probably safe to assume that this process reduced the amount of ANE in Scandinavia to much less than what was carried there by the indigenous foragers (15-19%). However, present-day Scandinavians carry around 17% of ANE, which must mean that there was another migration wave into Northern Europe after the Neolithic, coming from an area rich in ANE. This was probably the Indo-European expansion from the middle Volga region (see here).

Nevertheless, Gokhem2 does carry forager admixture, which can be seen in its non-trivial levels of Eurogenes K15 components strongly associated with indigenous European forager ancestry: North Sea at 16.74% and Baltic at 3.78%. What this suggests is that the admixture event between the Near Eastern and European ancestors of the TRB farmers didn't take place in Scandinavia, but rather somewhere on the European mainland where ANE wasn't present at the time. Interestingly, the Oracle results are in agreement, because, for instance, they feature La Brana-1 but not Ajvide58.

Eurogenes K15 - Ajvide58

North_Sea 35.8
Atlantic 18.76
Baltic 24.56
Eastern_Euro 17.54
West_Med 0.02
West_Asian 0
East_Med 0
Red_Sea 0
South_Asian 0
Southeast_Asian 0
Siberian 0
Amerindian 2.09
Oceanian 0
Northeast_African 0
Sub-Saharan 1.22

4 Ancestors Oracle results

Principal Component Analyses (PCA) featuring West Eurasian, Eurasian and global reference sets, respectively, show that Ajvide58 is outside the range of modern West Eurasian genetic variation, which is in line with the results of all other ancient European foragers sequenced to date. The cross marks the spot (click on the images to download high resolution PDFs of the plots):




Eurogenes K15 - Gokhem2

North_Sea 16.74
Atlantic 27.71
Baltic 3.78
Eastern_Euro 0
West_Med 45.34
West_Asian 0
East_Med 4.66
Red_Sea 0.78
South_Asian 0
Southeast_Asian 0
Siberian 0
Amerindian 0
Oceanian 0.98
Northeast_African 0.01
Sub-Saharan 0

4 Ancestors Oracle results




The Eurogenes K15 and Alexandr Burnashev's 4 Ancestors Oracle are available for use free of charge at GEDmatch for anyone with genotype data from 23andMe and similar personal genomics companies. Look for the Ad-mix option and then the Eurogenes tab.


110 comments:

Chad Rohlfsen said...

The ANE really makes things kind of messed up. It tries to fit her more as Sardinian, rather than her Northern European, as her percentages of Basal Eurasian is basically the same as Brits and Northern French people.

Chad Rohlfsen said...

Rather than Northern European. Excuse me.

Davidski said...

Yeah, the Gokhem2 West Eurasian PCA suggests to me that modern Northern Europeans are basically TRB farmers shifted east due to extra WHG admixture and a big chunk of ANE that the ancient farmers didn't have.

We can add to that a bit of ENA ancestry from Siberia for some of the Northeast and far Eastern Europeans. However, I now have a feeling after testing Ajvide58 that ADMIXTURE tends to overestimate the level of modern Siberian ancestry among Finns, North Russians and the like. That's because Ajvide58 appears more East Eurasian-like than modern Finns, even though this is in all likelihood not the case. Also, modern Siberians aren't purely East Eurasian; they do carry a lot of ANE and in many cases probably even WHG.

barakobama said...

"Yeah, the Gokhem2 West Eurasian PCA suggests to me that modern Northern Europeans are basically TRB farmers shifted east due to extra WHG admixture and a big chunk of ANE that the ancient farmers didn't have."

I am sure it's more complicated than that. It is hard to put all northern Europeans in one category. The only reason they have been is because they have more WHG and ANE ancestry than southern Europeans(who are put in one category because they have more EEF ancestry).

The TRB culture was never strong in northeast Europe and its eastern border was Poland.

We can't forget that much of northeast Europe and Scandinavia was still ANE-WHG hunter gatherer country thousands of years after the Neolithic age ended. Which could be an explanation as to why northeast Baltic people are probably about 70% WHG+ANE(over 50% being WHG).

While studying Norwegian mtDNA at FTDNA I found a perfect U4a1 HV1+2 match with two 3500YBP hunter gatherer U4a1 samples from Bolshoy Oleni Ostrov, Karelia Russia. This haplotype has a bunch of extra mutations passed U4a1, so it is probably some type of Baltic hunter gatherer-specific lineage.

If we do the research I am sure we can find plenty of Baltic hunter gatherer-specific maternal lineages that take up decent percentages of modern Baltic U4, U5, and U2e, proving that their ancestry is more complicated.

Davidski said...

There was definitely EEF influence in all of Northeastern Europe by the end of the Neolithic. There were no pure WHG/ANE people alive by that time anywhere in Europe.

Chad Rohlfsen said...

TRB has more WHG than Northern Europeans. It has to be the ane that is keeping her from sitting with Brits and myself.

Matt said...

Gokhem2 is an interesting lady.

She doesn't really fit an Oetzi-SF11 cline (or hunter-farmer in Laz parlance) very well, compared to present day Europeans.

Very Sardinian like in the intra-European dimension and overlapping with Scandinavians / Ukrainians (/ perhaps British) on PCA dimensions with inter-continental elements on them.

Northwest Europeans wouldn't really fit a mix of Gokhem2 and SF11 very well - that would leave them more extreme outliers from Europe on the world dimensions than they are.

However NW Europeans might fit a Gokhem2-Lithuanian mix, which wouldn't shift their position on world dimensions very much and give them the right position on intra-European graphs.

Not sure this is a better fit than Sardinian-SF11 though just from looking at the PC (seems slightly better on the intra-European PCA, but slightly worse on the East + West Eurasian plot).

However if we do adna in Scandinavia and the Isles and only find evidence of Gok types before the ANE bearing wave, might become an unavoidable conclusion (if there just weren't any Oetzi and Stuttgart style Sardinian farmer types around to mix with). And the same if we look at the early Corded Ware and Samara samples and they bear EEF affinities (which would indicate against there being a "hunter" population as in Laz actually involved in European admixture).

Chad Rohlfsen said...

This fits nicely with a paper we discussed not too long ago that stated the mixing of EEF and WHG took place 5000-3500 ybp, I believe it was. I believe David stated correctly that the excess WHG picked up by farmers was in Central Europe. The aboriginal tribes of Northern Europe may have been mostly wiped out, with a possible 25% replacement by IE-speakers.

I still stand by my statement that she has more WHG than Northern Europeans. Beings that she is 23% Basal, she has to be in the 50-55% EEF range. The remainder has to be WHG, since she has no ANE. That would make her more WHG than almost all modern Northern Europeans. You would have to go to the NE Baltic to find a range like this, for WHG.

About Time said...

I have a hard time imagining Volga ANE getting to Britain / Scandi without picking up lots of middle pops.

So that might mean British ANE was really a package of (Volga ANE + Pontic EEF + C Eur EEF + maybe N Eur WHG).

So no wonder modem Brits aren't a simple with of Gokhem + Gotland HG.

TRB was in Pommerania (Baltic Poland) Btw. Complete with megaliths. Whether that was acculturation in part, we need lots of DNA. Same deal for Lusatian.

I posted stuff on last thread about Comb Ceramics there. Did they adopt TRB stuff with less mixing? Like Kashubians in Middle Ages.

Chad Rohlfsen said...

The Bronze Age Dane didn't appear to pick up middle populations. He's plotting on the edge of Finns, towards native Americans and east Eurasians.

truth said...

David,

Could you make a plot with BOTH the foragers and Gokheim ? I want to see if the plot resembles that of Skoglund :

http://3.bp.blogspot.com/-hoBLZ0dZXto/U1luxCudqwI/AAAAAAAAJk4/J3X428YdPzo/s1600/PCA.png

Grey said...

"The aboriginal tribes of Northern Europe may have been mostly wiped out, with a possible 25% replacement by IE-speakers."

The presence of so much y DNA I in and around Scandinavia and the Balkans says different.

If I came with an incoming wave why is it most prevalent in the most defensible terrain?

.

"I have a hard time imagining Volga ANE getting to Britain / Scandi without picking up lots of middle pops."

Two routes.

1) northern land route mixing with the native (I/U) foragers.

2) southern maritime route from an origin near the Black Sea bringing ANE with them.

If correct there could be two strands from the same origin which meet up again somewhere around Denmark/Holland.

Chad Rohlfsen said...

Grey, i2 is basically non existent in the north. I1 is from later migrants. The i's entered Neolithic farmers via the mix I spoke of.

Grey said...

The idea of a maritime route from near the Black Sea makes more sense if the model involves three waves not two:

1) neolithic farmer wave
2) chalcolithic farmer/miner wave from the cucuteni/vinca regions west of the Black Sea displaced by early I-E expansions
3) bronze age I-E conquest wave.

In this model a lot of prime maritime sites are already settled by the first wave (Capsium) so people retreating from the I-E taking the maritime route wouldn't necessarily be conquering all those existing settlements but bypassing existing settlements by hopping along the coast.

Grey said...

"Grey, i2 is basically non existent in the north. I1 is from later migrants. The i's entered Neolithic farmers via the mix I spoke of."

If I came with later arrivals why is it concentrated in the two regions most likely to have been refuges of native foragers: Scandinavia and the Balkans.

Grey said...

distribution of y DNA I

http://2.bp.blogspot.com/-lhXgXOSp_a0/UKljuK_d9AI/AAAAAAAAABM/gB_lmZj0Jqc/s1600/792px-Distribution_Haplogroup_I_Y-DNA.svg.png

i.e. the places least effected by LBK

Chad Rohlfsen said...

I2 is big in the Balkans, as they were absorbed by farmers. I1 has not been found in a Mesolithic Scandinavian. Nor is there any ANE in a Scandinavian farmer to this point. These things make it look like there weren't many people absorbed in the North. At least to this point.

Chad Rohlfsen said...

I2 does not really exist in the North anymore, Grey! That is the point I'm making. All Scandinavian Mesolithic remains are a single I*, and the rest are I2.

barakobama said...

That's a good point. There are multiple west European hunter gatherer I clades(I1, I2a1b, and I2a2) which went through founder effects and take up big chunks of Y DNA in some modern Europeans. PWC hunter gatherer Ajv52 most likely belonged to I2a2a1 which is pretty popular in northwest Europe today.

barakobama said...

"distribution of y DNA I

http://2.bp.blogspot.com/-lhXgXOSp_a0/UKljuK_d9AI/AAAAAAAAABM/gB_lmZj0Jqc/s1600/792px-Distribution_Haplogroup_I_Y-DNA.svg.png

i.e. the places least effected by LBK"

We must always remember that Y DNA haplogroups are just paternal lineages, which are prone to founder effects and other things in human societies that screw up percentages in a way that ancestry can't be accurately predicted with Y DNA.

I2a1b originated in post-Mesolithic(Mesolithic samples are pre-I2a1b) and originated in western and or northwestern Europe. Nearly all of east European I2a1b is under the very young clade I2a1b1a1-L147.2.

Most Y DNA I today is under clades that expanded after the Neolithic. The only ones which obviously been where they are today since the Mesolithic are the various I2* clades and I2a1(I2a1b-, I2a1a1-) clades, which are very rare.

Grey said...

@chad

The distribution says different.

If your theory is correct why is I so concentrated in certain regions?

That there are specific clades of I which survived to make up such a large proportion of surviving European DNA and these are relatively recent makes perfect sense if you consider a plausible mechanism for forager survival in the Balkans and Scandinavia.

If you have an advancing farmer wave the foragers who survive will be those in favorable terrain who *adapt* to farming (or more likely herding) and thus increase their population density enough to resist the farmers.

This is hinted at by the y DNA bottlenecks around the correct times for this to occur.

So the farmers didn't absorb the surviving native foragers on the periphery who adapted to farming.

Those surviving foragers (in alliance with the I-E) absorbed the farmers instead.

Grey said...

@barak

"We must always remember that Y DNA haplogroups are just paternal lineages, which are prone to founder effects...Most Y DNA I today is under clades that expanded after the Neolithic."

Yes that's my point. The only way low density forager populations are likely to survive an advancing wave of higher density farmer populations is if they adapt to farming early enough to resist.

If correct that process would create bottlenecks *during* the neolithic from those who first adapted.

Chad Rohlfsen said...

Grey, I2 has been the most dominant mesolithic signature in Europe. Not a single I1 has been found. Today I2 is only 0-5% of the lineages of Northern Europe. The most popular male line has been virtually wiped out. Until we have an i1 mesolithic or Neolithic mixes with ANE, it's not looking good for much of a survival in the north.

Davidski said...

Truth,

I can't run them together because they tend to stick to each other, probably due to high levels of homozygosity.

But the reason my plots don't resemble those from Skoglund et al. is because I didn't project the ancient samples, so they don't pull into the middle of the plots because their PCA space isn't shrinked.

Chad Rohlfsen said...

I2b barely tops 10% in central Germany and NE Sweden. Just to expand on the convo. I2 is not popular in the North. I1 could've hopped on the back of a Neolithic group, for all we know.

jackson_montgomery_devoni said...

Does this mean then that the North_Sea component from the Eurogenes K15 calculator at GEDmatch is of purely European Mesolithic hunter-gatherer origin?

Davidski said...

Probably not. I'd say it's overwhelmingly of hunter-gatherer origin.

But I might be able to check this by running the K15 components in TreeMix along with the ancient genomes.

Exciting times ahead.

Grey said...

@chad

"it's not looking good for much of a survival in the north."

The distribution clearly implies there was significant survival in a couple of distinct regions one of which was Scandinavia and which (mixed with I-E) expanded over a much wider area later as a component of Germanic.

.

@jackson

"Does this mean then that the North_Sea component from the Eurogenes K15 calculator at GEDmatch is of purely European Mesolithic hunter-gatherer origin?"

I think labeling may be important here as i think the forager component that survived to the present day to be a significant percentage of the European population descend from the small (possibly very small) percentage who made the forager -> herder transition via a very tight bottleneck before they were displaced.

So if corect they'd be the founders who made the transition from mesolithic hunter-gather to neolithic farmer (herder) rather than mesolithic hunter-gather per se.

Chad Rohlfsen said...

The I1 in Northern Europe doesn't imply anything at this time. We don't know how it even got there. You are making assumptions. I am basing my points here, on what can be proven. I2 is virtually lacking in Northern Europe. There is no I1 Mesolithic sample. Northern Europeans are almost identical in EEF/WHG/ANE percentages. The cline is North to South, way more than it is East to West. To this point, WHG ancestry looks to be largely descended from Central Europe. Until we have a Neolithic sample with ANE, it is impossible to say that they survived in any sizable amount. Again, until an I1 is found in Northern Europe, we don't know where it came from. It could have gone to Scandinavia with the TRB farmers.

jackson_montgomery_devoni said...

David,

Indeed I think it would be very informative if you checked these components with the ancient genomes with TreeMix. We may then be able to untangle them and find out what they are truly made up of. That will be very exciting for sure. :)

Grey said...

@chad

"You are making assumptions"

We both are. Your assumption is that the I DNA came from outside. Fair enough but then you need to explain the very odd distribution.

My assumption is it is native forager survival in two of the hardest to reach spots filtered through extreme bottlenecks of those few foragers - possibly single families - who made the forager to farmer transition hence the neolithic dating of the bottlenecks.

In both cases farmers replaced the foragers but in my version it was ex-forager farmers who replaced other foragers who hadn't made the transition.

Chad Rohlfsen said...

I1 is not found in Mesolithic Scandinavia. I1 may only be 5000 years old. That means that it may not have arisen in a hunter society. y-dna really doesn't tell the story. The more important thing is the equal levels of WHG and ANE across Britain to Scandinavia. Neither one is really more Mesolithic than the other. Neither one is really a hold-out, more than the other. Just because one has more I, means nothing. The fact that it's I1, could mean a lot. I1 could very well have developed in Central Europe, in a Neolithic mixed farmer that is I*, 5000 years ago.

Chad Rohlfsen said...

The Balkans have a lot of I2, but way lower levels WHG than Northern Europeans. Hunters mixed into farming societies. Uni-parental markers and aDNA don't fit each other well.

Fanty said...

Well,.... Farmer Y-DNA is also almost "nonexistant in the north".

It almost stops short after the alps (the typical landmark for the northern/southern Europe split today.

And not only is Farmer Y-DNA nonexistant in the north, its also very weak in the south.

So, whats going on here?

Fanty said...

There is one thing that however always comes to my mind when thinking about mesolithic forrager genes.

These people had been really big (1,70m-1,90m) and massive. "Beary".
And Farmers had been small (1,50m-160m) and fragile.

I dont know what Indo-Europeans are supposed to be through.

Romans famed Germanics as beeing super big people. Up to totaly ridiculous and unbelievable numbers (One Warchief with suposed 2,50m height, warriors at 2,10m). No skeletons come close to these numbers (average Germanic skeleton is 1,72m)

The largest people in modern Europe are Scandinavians, followed by Yugoslawians and Dutch.

The smallest people are Portuguese, Spanish, Italians and Greeks. (Must be the Farmer blood) All others are average.

Ok, what I wanted to say is....
Is it coincidence, that the two body size hotspots are identical to the two Y-DNA "I" hotspots in Europe or does it have anything to do with mesolithic ogre blood?

The of course, Yugoslawians dont have much WHG autosomal DNA and stuff like "possible size range" should be part of the autosomal DNA, right?

But then again, why are people like Finn and Balts who possess the most WHG autosomal DNA are only of average size? Maybe its too cold? (cold adaption shrinks bodies)

Well yeah and then there is diet. Animal proteins (Meat and Fish) is what makes you tall. Farming products make you small.

blabla...

Can someone explain to me why its Scandinavians and Yugoslavians who are so big? What do they have in common?

Davidski said...

Fanty,

Tallest are Serbs and Croats from the Dinaric Alps, followed by the Dutch and then Scandinavians. But I have a feeling that young people from the East Baltic, especially Lithuania, aren't any smaller than Scandinavians these days. They might end up being even taller as the socio-economic situation there improves further.

Also, Greeks aren't one of the shortest people in Europe from the recent stats that I've seen. They actually seem to be the tallest southern Europeans and as tall as Central Europeans.

Maju said...

I wonder why your PCA results, David, give a more Sardinian-like placement for Gok2 than what she gets with the original studies, which place her near Basques, Spaniards and Italians. Is that because you did not use "projection" or is for some other reason (or both)?

Besides, I kind of like particularly this comment by Matt:

Northwest Europeans wouldn't really fit a mix of Gokhem2 and SF11 very well - that would leave them more extreme outliers from Europe on the world dimensions than they are.

However NW Europeans might fit a Gokhem2-Lithuanian mix, which wouldn't shift their position on world dimensions very much and give them the right position on intra-European graphs.


This, although surely oversimplified, makes good sense based on the archaeological record. I.e. a Megalithic farmer wave from presumably SW Europe (Iberia or France) was partly superseded by the Western Indoeuropean intrusions (Corded Ware and such, for which Lithuanians are almost certainly the best proxy alive).

However it probably falls short in order to explain the ANE element, even if we assume that the TRBK population had Basque-like ANE affinities of c. 10% (Lazaridis' figures). So I would think that a third locally rooted HG element (with important ANE) is also present - what also makes good sense to my mind.

barakobama said...

Fanty,


You're going off fibs and feelings, and giving hunter gatherers and farmers simplistic characteristics. Almost nothing is black and white, especially when you're talking about 1,000's of differnt sub cultures, 100's of generations, 1,000,000's of people, over 1,000's of years, and in many differnt parts of Europe.

Someone has to be an expert in archaeology and on how the human mind and human society works to get an accurate idea of how the various European hunter gatherer and farmer societies thought about their world and of each other.

Loschbour was only 5'3. It would be very interesting if European hunter gatherers really were around 6'0 on average because that would make them the tallest non-20th century humans in history. Can you show me some sources?

"Romans famed Germanics as beeing super big people. Up to totaly ridiculous and unbelievable numbers (One Warchief with suposed 2,50m height, warriors at 2,10m). No skeletons come close to these numbers (average Germanic skeleton is 1,72m)"

The overall average height for a man in most populations is about 5'6-5'8, and in the last 200 years or so western men have grown to 5'10-6'0. Prove is old bones and records from Europe, and west Africans are 5'6-5'8 while African Americans are 5'10-6'0.

The idea that norther euros are abnormally tall I think is mostly a myth. The stero type could have started with Romans and Greeks, because they mentioned it. Ancient bones tell us that the average pre-Roman Gaul was 5'7, the average German during the prime of the Roman empire was 5'7-5'8, and that the average man from Italy and Roman centurion was 5'6-5'7. There was a slight 1 inch difference.

Today, pretty much all Europeans and Americans(including African Americans) are around the same height. I am sure if west Asians were as healthy as Europeans they'd be just as tall.




barakobama said...

"Ok, what I wanted to say is....
Is it coincidence, that the two body size hotspots are identical to the two Y-DNA "I" hotspots in Europe or does it have anything to do with mesolithic ogre blood?"

Probably no connection with Y DNA I. I doubt hunter vs farmer ancestry has much to do with height. Yugoslavians are strangely tall. All the ones I know are above 6'2. But i bet they were only 5'6-5'8 just a few hundred years ago.

Davidski said...

Maju,

If I was to project Gokhem2 onto that plot she would indeed land among the Basques, French and Spaniards.

Fanty said...

I dont know a source at the moment.

I read an article that said
Paleolithic Europeans would be the tallest non 20th people.

MEsolithic Europeans would suposedly be smaller than Paleolithic ones but taller than neolithic ones.

It claims that bronce age europeans are as small as neolithic ones, but iron age europeans are almost as tall as mesolithic ones.

Then, medieval Europeans are as small as neolithic Europeans and late 20th century (1970 upwards) Europeans get as large as paleolithic ones.

Thats roughly what it said.

Fanty said...

It also claimed paleolithic Europeans would not only be taller than modern Europeans but also have larger brains than modern Europeans.

Fanty said...

Oh and thats a modern brain size map:

http://radishmag.files.wordpress.com/2013/01/world-map-of-brain-volume.gif

Matt said...

Another comment, the combination of PCA for Gok2 also show that the alleles which place a population West and South on an intra-European (or intra West Eurasian) PCA are at least somewhat different from the alleles which place a population West on an African-Eurasian weighted world PCA. They're different dimensions which are apparently just correlated in the present-day.

Seems interesting, as I had thought it would be Basal Eurasian-like vs ENA-like on both. Instead it seems like it could specific alleles which are more important on the intra European level are only loosely connected to Basal Eurasian and ENA.

Davidski I can't run them together because they tend to stick to each other, probably due to high levels of homozygosity.

If I'm understanding this right, Pitted Ware girl and TRB girl cluster *together*? I guess that must be an artefact but it's surprising.

...

Also, the Oracle results are worth checking out. We might have doubted the Skoglund treemix finding of around 25% increased WHG composition to Gok2 relative to Oetzi, yet these Orcale results, which, when allowed to vary to above 2 populations essentially show 25% La Brana, 75% Sardinian. Or, put another way, exactly what Treemix shows, 75% WHG, 25% BE.

About Time said...

@Fanty, I wonder how much those size changes are due to:

1. Diet (obviously)
2. Relative degrees of inbreeding depression in each period (did people marry in the same tiny village or meet mates through geographically extended peer networks?)

And lastly 3. Migrations of small/big bodied populations and their numerical dominance each period.

Which also relates to class issues and Darwinian forces operating in the social structure of each period.

Or 4. Combine all 3 factors

About Time said...

Oh, one more factor: ascertainment bias due to burial practices and where you are digging.

The kurgan males were often huge, well over 6 foot in many cases (6'5" etc). But these were kings or Kshatriya or whatever. Eating well and probably selected in the social environment, maybe on a multi generational basis.

Now, were the guys building the chariots huge like that? What about the drivers (think of a jockey or a navy pilot -- what is efficient for the job?)

Grey said...

@Fanty

"Is it coincidence, that the two body size hotspots are identical to the two Y-DNA "I" hotspots in Europe or does it have anything to do with mesolithic ogre blood?"

It seems plausible enough to test.

There are other smaller I dna hotspots around Europe mostly(?) in mountainous areas also so they could provide further samples.

.

@barak

"I doubt hunter vs farmer ancestry has much to do with height. Yugoslavians are strangely tall. All the ones I know are above 6'2. But i bet they were only 5'6-5'8 just a few hundred years ago."

Genes for potential height and genes for varying height with diet may exist separately.

So population A might have a potential average height of 6' and a second population a potential average height of 5' 8" and with a bad diet they might both be 5' 6" but with a good diet both might rapidly increase to their different potentials.

Something like this might explain the height explosion of the Dutch.

Fanty said...

Yeah I think it must be all of this too.

Differences between actual Farmers and Hunters can be explained by food anyways.

I recall an article that claimed Farmer bones show signs of very bad, unhealthy diet or famine and that they carried all kinds of sickness based on their unhealthy lifestyle. While hunter bones show they had been perfectly fed and their overall health must have been better.

Fanty said...

there are food scientists who dare to claim that the largest mistake humanity ever did was to invent bread. Its poison to human health to have carb bombs.

Grey said...

"There are other smaller I dna hotspots around Europe mostly(?) in mountainous areas also so they could provide further samples."

Fehnman island is mentioned as a tall/robust hotspot but i don't know about the dna.

Davidski said...

Here's what a composite of Ajv58 and Gok2 gets on the K15. Note the almost complete lack of exotic admixtures. The best fit in the Oracle is French Basque + Lithuanian.

North_Sea 29.12
Atlantic 22.44
Baltic 23.3
Eastern_Euro 4.64
West_Med 20.27
West_Asian 0
East_Med 0
Red_Sea 0
South_Asian 0
Southeast_Asian 0
Siberian 0
Amerindian 0.12
Oceanian 0.11
Northeast_African 0
Sub-Saharan 0

jackson_montgomery_devoni said...

Interesting composite results there David. So how would you describe the origins of the North_Sea component from the Eurogenes K15 calculator? By this I mean when seen in modern day Europeans what kind of ancestry would you say it represents?

Helgenes50 said...

What is the best fit for this composite in modern Europeans ?

barakobama said...

You should do a Irish+Gok2 combination, and see how close it is to Iberians.

barakobama said...

I mean't southwest French north Iberians.

Davidski said...

The best fit for the composite is 50% French Basque and 50% Lithuanian, with a distance of just over 6, which is pretty good. The best single fit is East German, but that's with a distance of over 11, which isn't good. So no single modern population is a good fit for this composite. What this suggests is that no present-day Europeans are a simple mixture of TRB farmers and Scandinavian foragers.

I'm still learning how to use TreeMix. Once I figure it out, I'll calculate the WHG and ANE ratios of the North Sea, Baltic, East Euro and Atlantic components.

Maju said...

@David: this "ancient Swedish alchemy" of the last comments looks a most interesting finding with some implications. Ara you planning to write a more detailed entry on it?

Particularly striking is that the composite looks, as you describe it, almost as the child of a modern Basque-Lithuanian couple and that the "exotic" affinities are resolved into nothingness with the mix, what makes the composite a "modern European" in the basics although not quite any specific one.

It is possible, I presume, that the lack of good fit with any modern European is because the two samples do not belong to the actual ancestral populations (Scandinavia was in general destination rather than origin of migrant flows, with very punctual exceptions but even these relate more to Denmark than to the Scandinavia Peninsula).

In any case it would seem that your "alchemy" makes the ANE problem disappear, right? If so, I guess that the solution to the "ANE problem" is, in essence that most modern Europeans (excepting the SE probably) are a variable blend of ancient Atlantic farmers and Eastern European (~IE) herder-conquerors.

Deducing from your East German approximation (to be fine-tuned, I guess), the line of 50-50 mix line goes approximately through Berlin.

Does this make sense to you or am I speculating wildly on too little data?

In any case your "alchemy" seems very interesting and I hope to read more on it soon.

barakobama said...

"If so, I guess that the solution to the "ANE problem" is, in essence that most modern Europeans (excepting the SE probably) are a variable blend of ancient Atlantic farmers and Eastern European (~IE) herder-conquerors. "

Haven't people been suggesting this idea for decades? There is truth to it, but it sounds to simple.

Chad Rohlfsen said...

Here's mine, for comparison


1 North_Sea 36.88
2 Atlantic 26.25
3 Baltic 10.3
4 West_Med 10.14
5 Eastern_Euro 8.84
6 East_Med 5.8
7 Siberian 0.8
8 Oceanian 0.66
9 West_Asian 0.3
10 Red_Sea 0.03

Chad Rohlfsen said...

David,
What would be your explanation of her discrepancies of Basal ancestry? If they have her at 22.8% Basal, relative to Stuttgart, should she not be closer to Northern Europeans, than Mediterraneans? Can you run a plot that includes Ajv58, Malta, and Gok2 with modern Europeans? I find it curious that she is plotting near people that are apparently 10-15% more Basal Eurasian than herself. Thanks!

barakobama said...

Chad, there must be some type of mistake with the basal Eurasian thing. Everything else shows that Gok2 is only a little more WHG than Sardinians.

Table S14, clearly shows northern Europeans are more related to Ajv58 than to Gok2. This would not make sense if Gok2 had more similar EEF-WHG percentages to northern Europeans than Ajv58 does.

barakobama said...

Table S14, of Skogland 2014, my bad I just woke up.

Chad Rohlfsen said...

Aren't their plots projected onto each other, instead of running the samples together? Even if she is around Sardinians, this should put her in their range of 79-83% EEF. Most Northern Europeans should still be closer to her than Ajv58. I would like to see a plot with her, Stuttgart, Ajv58, and modern Europeans. This would be very interesting.

Davidski said...

Chad,

The level of Basal Eurasian admixture in one TreeMix graph can't be the last word on where Gok2 should cluster, since we don't know how much Basal Eurasian ancestry the Sardinians would show in the same TreeMix run. My guess they'd show more than Gok2, but not much more.

By the way, the shared drift stats in that paper indicate that Gok2 is more similar to Sardinians than anyone else in Europe.

Chad Rohlfsen said...

If she ends up 75-80% EEF, so be it. It's nice to finally have a mixed individual. Is it the basic lack of ane that drives the drift?

Davidski said...

ANE appears to have a very important influence on the second (horizontal) dimension in that PCA. I think this is what creates the two parallel west to east clusters (North European and West Asian) on such West Eurasian plots, which is what David Reich was talking about recently. So anyone lacking ANE will be pushed to the far left of the plot, near the Sardinians.

The first dimension seems to be based on Near Eastern versus combined WHG and ANE ancestry, and this is why Gok2 is just above the Sardinians.

Chad Rohlfsen said...

How soon do you plan on doing tree mix runs on her and various European populations? Thanks, in advance!

Maju said...

Your last comment, David, makes good sense.

But we know that, when the PC graph is European-only, the Sardinian/Basque apparent similitude splits into two polarities that pretty much define both axis. We know where Alpine EEFs plot in that axis (near Sardinia or Italy) but where does the Atlantic farmer population represented by Gokheim falls in them? On first thought I would think that near Basques but in this particular analysis, they happen to fall closer to Sardinians, so I do wonder.

barakobama said...

"Alpine EEFs plot in that axis (near Sardinia or Italy) but where does the Atlantic farmer population represented by Gokheim falls in them? On first thought I would think that near Basques but in this particular analysis, they happen to fall closer to Sardinians, so I do wonder."

The established reason is that Basque are something more complicated than Gok2-like Neolithic farmers. Basque are a tiny ethnic group who are and were surrounded by many differnt ethnic groups, so even though they are mostly Neolithic-Mesolithic west European relics, they have shared post-Neolithic ancestry with other west Europeans(eg, R1b).

Maju said...

The Basque People is not a "tiny" ethnicity when compared with a lot of of other European peoples. We're talking of around 3 million people within the conventional borders (notice that at least Gascons and some Spaniards cluster with them strongly, so it's a conservative take). This is more people than any of the following: Lithuania, Estonia, Lithuania, Slovenia, Albania, Cyprus, Sardinia...

It's small but not "tiny". Tiny is Iceland, Malta, Luxemburg. Even more importantly it is an obviously important population in European analysis, describing one of the axis in the PCA (usually vs. Eastern Med populations - the other axis is usually Sardinia/Italy vs. Eastern Europe).

In general somewhat loose terms, (purebred) Basques can be considered, I understand, as a particularly unmixed SW European remnant, and hence are important when considering this region which is becoming more and more obvious that cannot be ignored in European population paleohistory (on the contrary: it has been a major pivotal contributor). By SW Europe I understand basically Iberia & France, although it obviously projects genetically beyond this core area.

The same as Sardinians or Lithuanians are a reference population, so are Basques. This may be less obvious when West Asians are included in the (the axes are defined by +/- ANE and Europe/West Asian polarities) but it becomes much more notorious when West Asians are removed (or minimized).

Matt said...

@ BO,

On table S13 though, note D-statistics show that "Gökhem2 is not a clade with Sardinians and the Iceman, but closer to hunter-gatherers". Need to have some kind of HG mixture to make that happen. It's hard to see how that wouldn't decrease Basal Eurasian relative to Sardinians. But that's not calculating the level.

Perhaps one idea for how to estimate Basal Eurasian ancestry in Gok would be to run sets of D-statistics of along the line of to S14 of the form Yoruba, X, Gok2, (West Eurasian) across East Eurasian populations, for various West Eurasian populations as the reference and East Eurasians as the target.

For populations without ENA mixture (e.g. Lithuanians, not Estonians or Finns, Greeks, not Turks, etc.) as WHG+ANE are supposedly equidistant from ENA, only the absolute ratio of BE, WHG+ANE should affect the outcome (obviously ENA mixture will increase similarity to ENA populations much more than a similar level of WHG+ANE).

If the West Eurasian population is less BE than Gok2, the statistic will be statistically significantly negative (East Eurasians less related than to Gok2), if positive, it should be more BE than Gok2.

This wouldn't tell you the exact amount, but would indicate relative amount, e.g. as much as Orcadians, less than French, etc (for example).

(You could also use chimpanzee in place of Yoruba I guess).

Also be interesting to see what an D statistic with Admixtools would be like for Gok2 for Yoruba, X, Sardinians, La Brana, where for X we could use Basque, Orcadian, Bedouin and Iceman. That would give us an index of where the populations sit on a scale of relatedness Sardinian-La Brana unaffected by Lithuanian or Ajv specific population history.

D statistics are pretty easy to run, so I might be able to have a go at this if prepared Gökhem2 data is publicly available. Not sure if there are any additional challenges I haven't considered.

barakobama said...

"D statistics are pretty easy to run, so I might be able to have a go at this if prepared Gökhem2 data is publicly available. Not sure if there are any additional challenges I haven't considered."

The Skoglund 2014 genomes used to be here.

http://www.ebc.uu.se/Jakobsson/data/SkoglundEtAlData2014/


Grey said...

If y DNA I represents mesolithic survival (and may be it doesn't) and Sardinia has c. 40% I then it's hard to see how Sardinians can represent archetypal EEF unless the populations are geographically split and the sampling for Sardinians was done in the non I areas?

Chad Rohlfsen said...

Sardinians are in the 79-83% EEF range according to some figures. Up to 88% is the highest that I think I've seen. y-DNA and Autosomal don't really match. The British and Irish have almost exactly the same EEF as Scandinavians, but lack any significant I lineages. The same goes for the Balkans large EEF%, while still having lots of I2. Hunters joined the farmers. The paper was clear that hunter genes entered the farmers, not vice versa. Don't get hung up on uni-parental markers.

Maju said...

@Grey:

In my interpretation, the reason why yDNA I2 is associated to both Epipaleolithic and Neolithic populations is because the European Neolithic genesis in Thessaly was more complex than we could imagine. Not just the West Asian input seems less obviously Anatolian than we used to think (more Palestinian-like instead, although I guess that E1b was already saying it loud) but it clearly incorporated a large amount of European aboriginal blood (UHG in Lazaridis) which seems most clearly represented as haplogroup I2 (at the very least the Dinaric-Sardo-Pyrenean variant of it).

Then it was largely a matter of founder effects: Sardinians got the I2 lineage in large frequencies, in the Western Mediterranean (ancient Alps included) G2 was very important, E1b-V13 seems to have remained strong in the SW Balcans, while we are still uncertain about the exact role of J2 and R1b.

On these my bet is that J2 is either primarily Neolithic (as E1b, G2 and some I2) or secondarily so, having then arrived only with the Vinca-Dimini group (invaders from West Asia probably related to Halafian). As for R1b (Western subclades) my bet is that they were incorporated from WHG groups and re-expanded within the Atlantic Neolithic (alternatively it might have arrived with the EEFs but the chronology and geographic patterns I manage are rather consistent with pre-Neolithic presence in Europe, whatever its exact extent).

Many will say: R1b has not been detected (yet) in European populations before the Bronze Age. True. But we only have two samples of WHGs and they are from non-core areas. I'd need some samples from Paleolithic SW France and Portugal to be truly dissuaded from this idea (also more westerly samples from Northern Europe, such as from Denmark or Britain, would help here a lot).

Whatever the case it seems obvious by now that the European Neolithic had a complex genesis in Thessaly (with both transmediterranean immigrant and aboriginal European founders) and that this genesis was followed by founder effects and further admixture firstly, and, secondly, by an even more admixed expansion of the Atlantic Neolithic (~ Megalithic phenomenon) already towards the Chalcolithic period, with critical reshaping effects on European demographics West of the Elbe and North of the Alps. A third factor is of course Eastern European (Indoeuropean) expansion but this one seems to have inflicted a much more limited demic impact, at least West of the Elbe.

barakobama said...

Maju, I mean no offense, but I think your theory that R1b was in Mesolithic Europe is a desperate attempt to prove that the Y DNA R1b and mtDNA H=Basque-Iberian Palaeolithic survivors of west Europe theories that were so popular and believed by so many back in the day is partly true.

When considering that nearly all of R1b in west Europe falls under R1b1a2a1a-L11, L11 has language-specific branchs(eg, U106), it's equal diversity in west Europe, its dominance in all modern west Europeans(with clear founder effects), its complete absence in Mesolithic and Neolithic west Europeans, and L11's recent age estimate of 5000YBP, makes that idea hard to believe.

I am very interested to finally learn one day who exactly the post-Mesolithic fathers were who brought R1b-L11 and made it dominate in west Europe were. I am already pretty certain they traced their lineages originally to west Asia, arrived from east Europe(duh...) after the Mesolithic, were somehow connected to each other ethnically, and were a very powerful force who by the same means made their paternal lineages dominate in many differnt regions of west Europe.

I do think I and others wrongly assumed it had of been Indo Europeans.

A few of you're non-mainstream ideas about Y DNA have been proven incorrect. Just a few weeks before Laz 2013 came out you claimed on this blog that Y DNA I and especially I2a1-P37.2 were descended from early Neolithic farmers in Greece and not west European hunter gatherers. Laz 2013, Skoglund 2014, and modern Y DNA variation, prove Y DNA I and specifically I2a1-P37.2 was the main paternal lineage of Mesolithic west Europeans.

You have also claimed on this blog that R1a in south Asia is mostly pre Indo European, even after Keyser 2009, the Kurgan hypothesis, etc. Now there is confirmed R1a-Z93 in bronze age Mongolia, along with some European-specific maternal lineages, to add to the list of evidence for an Indo European origin.

Maju said...

The persistent problem is that R1b-S116, which is by far the most common R1b subclade, more so in Europe, has an obvious expansion epicenter in Southern France.

Is it possible to rethink it in terms Neolithic? For example a founder effect originating in Occitania? The aDNA evidence is strongly against it and there's much more ancient yDNA known for Neolithic than for Epipaleolithic (excluding Sweden, which is surely not representative).

Then we do know that there must have been a major demographic impact from the West (probably Iberia or France) with Megalithism (and Bell Beaker?), reflected in growing frequencies of mtDNA H, which has some notable correlation with yDNA R1b (European clades) in present populations. So my only plausible conclusion is that R1b (Euro clades) expanded from the West.

It is therefore result of founder effects in Atlantic Europe, which may be Paleolithic or Neolithic.

But I have another issue with the Neolithic hypothesis beyond the stubborn lack of aDNA evidence: using Underhill's 2014 fully Y chromosome age estimates for R1 (plus minor recalibration on my side to fit with the archaeologically confirmed 100 Ka old arrival to India of the Eurasian branch of Humankind), I get that R1b-M412 and R1b-S116 must have expanded in Magdalenian times (or Epipaleolithic using the most conservative calibration). I'm not just pushing for that: it is what the most up-to-date possible chrono-estimate says (even to my own surprise).

Of course they could in pure theory have experienced such expansion anywhere but the lineage is very rare outside Western Europe. So, while awaiting for further evidence, naturally, I do suspect that the lineage was incorporated somehow to the mixed EEF-WHG Atlantic Neolithic peoples, which then expanded over the EEFs.

I have all kind of doubts, notably to what degree these founder effects originate in Neolithic or are more genuinely remnants of the Paleolithic (notice that they can be a combo of both processes, as happens with I2). And also about the origin of the associated languages of which Basque is almost certainly the sole survivor, which right now I strongly suspect not being descendant of Paleolithic languages of Europe but of Thessalian Neolithic. But my speculation on R1b is, I am pretty sure, well founded on all the available evidence.

Re. R1a, notice that Z93 is NOT original from Europe but from South-Central Asia. Cf. again Underhill 2014.

Re. I2-P37, I do think it is a founder effect originating in the Balcans indeed (it is very common in the Western Balcans) but not necessarily Greece, rather Dalmatia I'd say. It's hard to believe otherwise considering the important founder effect it had among Sardinians (but impossible to know with full certainty unless aDNA shows up to help with the riddle).

Do you really think that Sardinian yDNA comes from Luxembourg? Or can you understand in spite of your arrogant ethnocentric naiveté that we only have a most irregular sampling of Paleolithic Europe: 5 samples from a particular locality in Sweden, one from Luxembourg and another from an anomalous site in NW Iberia with an even most anomalous patrilineage? We just do not have enough samples to discern patterns in Paleolithic Europe, at most we could infer (with some serious risk) that NW Europe was dominated by I2 but we can say near to nothing about the rest of Europe, where I or R1b may well have been present in patterns that we cannot realistically infer with the sparseness of data available as of now.

barakobama said...

"Or can you understand in spite of your arrogant ethnocentric naiveté"

This "kid" just dedicated himself to crushing beating you in this sport over the next few years. Laugh all you want, but it'll happen. I'mm tired of the heartless, dark, hateful, and snobby "adults" like yourself that fill these blogs.

Maju said...

Sure, whatever. I'm quite rusty but I'm not heartless, hateful nor snobby nor I would consider myself an "adult" except under the law - in fact I can understand well "hatred of adults", although it's the first time ever I get it on me.

I want "the kids" to take up on this quest of knowledge but first and much more important is being a kid because you're never going to be it again and it's for sure the best of life in spite of certain immaturity (physical apogee is around 30 but intellectual one is around 40).

But whatever you plan to do, you're not going to do anything anyhow bearing an account id that you hate yourself. Get a new one. You are not going to get anywhere either banging the dead horse of racialism and nordicism. You will get somewhere when you get your beliefs and put them under the microscope and dissect them mercilessly, because the most annoying "adult" is inside oneself: the one we have been taught, the Gestapo man inside our heads.

I don't laugh but I'm no afraid of your silly bravados either.

And, by the way, you have not answered my question: only reacted emotionally to it. Fair enough but do it offline if you want to be respected.

barakobama said...

"You are not going to get anywhere either banging the dead horse of racialism and nordicism. You will get somewhere when you get your beliefs and put them under the microscope and dissect them mercilessly, because the most annoying "adult" is inside oneself: the one we have been taught, the Gestapo man inside our heads."

I know this. I am serious about learning.

This is very insulting because it is one of the last things I ever want to be.

The reason I had raciest tendencies is because when I was younger I was excluded and made fun for being white, I had multiple raciest teachers, most of what i learned about in history class was why white people are bad and inferior(my teacher literally said that multiple times), and I felt white stereotypes restricted me in many ways.

I hated it and became very bitter. My racism was an inferiority complex and retaliation.

It was a big and very annoying weight that had been on my shoulders for most of my conscious life that only recently with alot of help from my family I have started to finally let go.

Do you finally understand now that I'm interested in this stuff because of genuine interest, not because I want to express extremist views or arrogant theories?

Many of the bloggers here have their own agendas, and most are too afraid to admit it.

"I don't laugh but I'm no afraid of your silly bravados either.

And, by the way, you have not answered my question: only reacted emotionally to it. Fair enough but do it offline if you want to be respected."

I'm not going to fly a plane to the Basque country where I can't speak the main languages, and look for a one in a million blogger named Maju, just to prove my manhood. That would be taken this too seriously.

"But whatever you plan to do, you're not going to do anything anyhow bearing an account id that you hate yourself."

I don't hate Barak Obama. I don't keep up with the news and I know very little about politics, so I don't have much of an opinion on him.

I consider myself conservative because I dis agree with many of the social ideologies liberals push for. Like extreme political correctness which turns into brainwashing, violation of individual rights, discrimination, labeling, and racism.

"Get a new one."

I won't change my ID because this is my ultimate secret. There are some people I know who know I post on these blogs, and they keep it a secret for my own sake because it is very embarrassing. I have tried before bringing this stuff into subject and it only leads to awkward pauses.

There are somethings we should not be embarrassed of and admit them and there are somethings which no one will understand and should be kept a secret.

I know you have your own deep secrets, everyone does.

"I want "the kids" to take up on this quest of knowledge but first and much more important is being a kid because you're never going to be it again and it's for sure the best of life in spite of certain immaturity (physical apogee is around 30 but intellectual one is around 40)."

That's very sad.


Davidski said...

I'd hate to interrupt, but if you guys could steer the discussion back to ancient Swedish DNA that would be awesome.

Grey said...

@Chad
"Sardinians are in the 79-83% EEF range according to some figures. Up to 88% is the highest that I think I've seen. y-DNA and Autosomal don't really match."

Yeah I just wonder which Sardinians i.e. if the sampling was widespread or from a particular area.

It may be nothing but I'm curious.

@Maju

"In my interpretation, the reason why yDNA I2 is associated to both Epipaleolithic and Neolithic populations is because the European Neolithic genesis in Thessaly was more complex than we could imagine."

It seems to me that if you look at the distribution of I and then subtract the part that expanded with the Germanic expansion then the distribution of the various I clades is very clearly little clusters in mountainous and / or very remote regions.

It seems to me there are two most plausible explanations for this.

1) Miners - possibly connected as you say to a much more complicated genesis zone around the Eastern Balkans.

2) Cro-Magnon survival - clusters of foragers who somehow made the forager -> farmer transition in time to avoid being swamped by farmers.

Either of those options makes me wonder about the I in Sardinia and whether it is regionally concentrated.

Map of Sardinia

http://mapsof.net/map/topographic-map-of-sardinia#.U8iUeZRdWSo

If it follows what seems to me to tbe the pattern for I it would be concentrated in the central mountainous zone.

I just wonder if there are two distinct populations in Sardinia - a central mountainous population and a valley population?

Davidski said...

The I2 in the Balkans shows low diversity and age - about 2,000 years or less. The most basal lineages of the Balkan or Dinaric I2 are today found in southeastern Poland and western Belarus.

Grey said...

@Davidski

"The I2 in the Balkans shows low diversity and age - about 2,000 years or less."

Yes but unless I've misunderstood something it seems to me that diversity and age are often assumed to indicate a direction of travel i.e. older diversity in region A than B means people moved from A to B.

Can't low diversity just mean a more recent bottleneck?

That is, A and B might at one point have been two equally aged and diverse clusters and then B goes through a tight bottle neck.

Just a thought.

Grey said...

@Davidski

"The most basal lineages of the Balkan or Dinaric I2 are today found in southeastern Poland and western Belarus."

Relief map of Poland

http://upload.wikimedia.org/wikipedia/commons/d/dc/Poland_topo.jpg

Just curious - would that be the mountainous bit in the SE?

barakobama said...

"The I2 in the Balkans shows low diversity and age - about 2,000 years or less. The most basal lineages of the Balkan or Dinaric I2 are today found in southeastern Poland and western Belarus."

Exactly, people need to understand I2a1-P37 is a west European haplogroup, and the high amount in the Balkans is from a very young founder effect. The fact there is more diversity of stone age west European P37(M26, pre-I2a1b, and I2a1*(a-, b-, d-, e-), than modern east European P37 is very telling. I don't understand why Maju sees a connection with southeast Europe and P37.

The big I experts, are starting to come out saying P37 is probably a marker of the SW-C European refugees during the last ice age, and U5b being its main female partner.

Grey said...

Sorry to be annoying :) but that region - Galicia it looks like? - does anyone know if the people from there have a reputation as being
particularly robust?

barakobama said...

"Yes but unless I've misunderstood something it seems to me that diversity and age are often assumed to indicate a direction of travel i.e. older diversity in region A than B means people moved from A to B.

Can't low diversity just mean a more recent bottleneck?

That is, A and B might at one point have been two equally aged and diverse clusters and then B goes through a tight bottle neck."

The higher diversity=place of origin theory is common sense. We can't say I2a1-P37 originated in SE Europe, if it most basal clades exist in western Europe. That wasn't very understable let me try again; if a 4,000YBP R1b clade exists in Ireland that doesn't mean R1b was in Ireland 7,000YBP, we can only trace it as far back as 4000YBP, saying it was there 7000YBP is pure assumption with no evidence.

If we strictly followed that theory though everyone would be saying Eurasians migrated out of central Asia, because central Asians have very mixed Eurasian ancestry. So, there are other tools we must use, but the diversity one is very constantly true(Humans originated in Africa, mtDNA H originated in west Asia not Europe, etc.).

Grey said...

@barak

"The higher diversity=place of origin theory is common sense...saying it was there 7000YBP is pure assumption with no evidence."

True. I'm making an assumption and then seeing if it leads anywhere.

There does seem to be - if you subtract the later Germanic expansion - a fairly specific mountain distribution of I combined with fairly young age and that is an odd combination.




Maju said...

"The I2 in the Balkans shows low diversity and age - about 2,000 years or less."

What are you basing this claim on?

Grey said...

@Davidski

"I'd hate to interrupt, but if you guys could steer the discussion back to ancient Swedish DNA that would be awesome."

Thing is *if* I clusters do represent concentrated paleo survival (or concentrated something else like miner/traders) then they might have unusual WHG/ANE/EEF proportions which might be generally useful for explaining regional anomalies.

Davidski said...

You'll find some interesting stuff on HG I in this directory.

http://knordtvedt.home.bresnan.net

Like this map...

http://knordtvedt.home.bresnan.net/Tree%20and%20Map%20for%20Hg%20I.pptx

Quote from Lazaridis et al: "We thank Kenneth Nordtvedt for alerting us about the existence of newly discovered Y-chromosome SNPs."

So like I say, Dinaric I2 is a marker of the Slavic expansion from somewhere in eastern Poland.

Grey said...

@Davidski

ty

Maju said...

Thanks, I know the Knordtvedt site and, while he has interesting info (and a lot of speculation!!!) on the phylogeny and distribution of Y-DNA, his age estimates are clearly way off. Just an example: he estimates I-M270 (I-root) to be a mere 24 Ka old, while I estimate, based on full chr. and correct calibration point of ~100 Ka BP age for CF, that haplogroup I-M270 has exactly doubly that age: ~48 Ka (which is also the age of R1, not accidentally).

Anyhow, the clade under discussion here is I2a1-P37, which he dates to c. 20 Ka ago (so it'd be 40 Ka but in any case very much pre-Neolithic). As for I2a1a-M26, which is the Sardinian-Pyrenean sublineage, the branch leading to it is dated to c. 19 Ka (i.e. realistically 38 Ka, assuming everything else is correct), however I can't find the "M26 tree" he points us to in order to discern this particular sublineage's details.

So I can't say much more but one thing is clear: the Balcano-Mediterranean branch I2a1-P37, also found among Central-North hunter-gatherers (dotted orange line in Knorvedt's graph), is very diverse.

I guess you could argue that the Sardinian-Pyrenean I2a1a-M26 was not incorporated to the Neolithic migrations from the Balcans but from somewhere else (Italy?) but it is clear that it is not a "recent" lineage nor has anything to do with the quasi-imaginary Slavic migrations (military-political conquests with negligible demographic impact).

Harder would be to argue, as some have done, that it is a Western Paleolithic lineage which arrived to Sardinia prior to the Neolithic. That's because there's nothing alive closer to an early European farmer than a Sardinian, so almost certainly every single important lineage they carry arrived to the island with Neolithic migrations and has remained there ever since.

barakobama said...

"So I can't say much more but one thing is clear: the Balcano-Mediterranean branch I2a1-P37, also found among Central-North hunter-gatherers (dotted orange line in Knorvedt's graph), is very diverse.
"

We can all agree I2-Dinaric is a pretty young founder effect, which like R1b-L11 in west Europe is not supportive of the idea P37 originated in the Balkans. Why do you think P37 originated in the Balkans?

Maju said...

Uhm, I2a1-P37 has two main subclades: I2a1a-M27 in Sardinia and the Pyrenees and I2a1b-M423 in the Balcans, greatest variance in Romania. This last has a low frequency sister branch in the British Islands, knowledge of which may well be product of commercial DNA oversampling and not relevant therefore (I mean: about half of Iberian I for example is unclassified, and so happens in other low commercial value areas of Europe like Romania and Ukraine).

Being hyper-strict in geometric extrapolation (and including the Isles subclade), we could conclude that this lineage originated in Northern Italy* but most likely this is product of several biases and the very direction of flow, so my understanding is that a Balcanic or Ukrainian origin is most likely, especially considering that the Ukraine-Romania area is high in diversity for I in general (I2a2a-M223, I1).

[*method: halfway between Britain and the Danube Delta is Hungary, halfway between Sardinia and the Basque Country is Catalonia or Languedoc; halfway between Catalonia and Hungary is Northern Italy]

So, barring the Epipaleolithic aDNA, my general impression is that I overall seems to have originated in Ukraine with overflow to Romania (and by extension surely other parts of the Balcans) probably in the Epipaleolithic. It is possible that it spread to Central Europe also in that period. Alternatively Ukraine may have been acting as "refuge" but then why is Romania so rich as well, if it has a very different Neolithic history?

In any case, I have no particular reason to think that I in general and I2a particularly did not originate in Ukraine/Romania/South Russia. I2a1 instead seems more Western-shifted, so I feel the need to infer that it originated in the Balcans (where it's most common) and was picked by the Neolithic current, which brought subclades to the Western Mediterranean and the British Islands in respective founder effects, associated to other lineages like E1b-V13 and G2a. Later demographic episodes may have semi-fixated the Balcanic variance into a single sublineage.

As I say, the alternative is Northern Italy, which is not particularly notorious for these lineages, nor seems to have been source of major population movements ever.

Grey said...

@Maju

"That's because there's nothing alive closer to an early European farmer than a Sardinian"

Which Sardinians?

http://www.ata.org.tn/fichier_PDF/Article2.pdf

"Sardinian Population (Italy): a Genetic Review"

"A high degree of internal heterogeneity was also found and it can be
attributed first to strict isolation and consequent high levels of endogamy and
consanguinity, secondly, to selective factor linked to endemic malaria that influenced
the distribution of some gene frequencies."

(malaria)

"Sardinians are characterized by extremely high frequencies of two mutant genes: the gene of thalassemia and the gene ...The frequencies are very high in the two coastal sides and decrease sharply towards the central mountainous area"

"The distribution of haplogroups in Sardinia appears strongly heterogeneous. In particular the frequency of haplogroup G – M 201 is significantly higher in the north and the Haplogroup I – M 26 is more significantly represented in the central-eastern area, characterized by a reduced presence of haplogroup R – M 269."

It's an interesting paper. There is a lot of genetic structure on Sicily. There is the archaic zone in the central east mountainous region which has the most I dna and then around the coastal plains there are distinct populations from different colonization events with lots of linguistic diversity to match.

So "Sardinian" may depend on where the sampling was done e.g. if it was done in the northern zone with all the G dna there may be a lot more EEF than in the mountainous zone.

Separately interesting is the possible effect of malaria in the coastal plains and the very high frequencies of malaria resistance genes in the coastal population.

EEF ~ malaria resistance ?

i.e. are Sardinians in the coastal plains so high in EEF because EEF had some African malaria resistance genes?

Grey said...

To add to the linguistic diversity point

http://en.wikipedia.org/wiki/Ligurian_language_(Romance)

"Besides Liguria (Ligurian Ligùria), the language is traditionally spoken in coastal, northern Tuscany, southern Piedmont (part of the province of Alessandria), western extremes of Emilia-Romagna (some areas in the province of Piacenza), in northern and southern west parts of Sardinia"

So Ligurian is spoken in the parts of Sardinia with a lot of G.

(i.e. maybe tying in with the Capsium wave, Otzi etc)

Grey said...

(Cardium not Capsium) doh

Shaikorth said...

By the way David, can you run the Ajv's, Gok's and StoraFörvar through K13? It'd be interesting to see how consistent the results stay.

barakobama said...

Maju,

We're going in circles here. You must have a lot of data you piled up over the years that points to the Balkans as I2a1-P37's place of origin, because I obviously don't see what you're seeing.

" especially considering that the Ukraine-Romania area is high in diversity for I in general (I2a2a-M223, I1).:"

BTW, most hg I experts are now saying I1 is ~5000YBP and was mainly dispersed in the Iron age by Germanic people, which would explain why it is found in southeast Europe(Britain, Russia, etc.), and the same is probably true for I2a2a-M223.

Chad Rohlfsen said...

An interesting correlation is the age and spread of I1, closely matching the age and spread of TRB Neolithic. Whether or not there's causation, is the question.

Davidski said...

Here are the Eurogenes K13 results for the ancient Swedes.

https://docs.google.com/spreadsheets/d/1SUDZOJx3svqWsXdvRnhdV3b0Fssumg3IZXwluW_c_H0/edit?usp=sharing

Shaikorth said...

Thanks.

Farmers are consistently over 40% West Med and all have some East Med too. Lack of West Asian in everyone makes one wonder whether it's got a separate origin.

Chad Rohlfsen said...

Wow, what is the ANE, of Ajv59?

Shaikorth said...

Ajv59 has easily the lowest coverage of all those Swedish genomes and consequently the least matching SNP's with the calculator. Its scores are probably just noise and not indicative of extra ANE in comparison to other Ajv's.

Chad Rohlfsen said...

Gotcha

Tesmos said...

Who is Ire8? Is he also a Hunter gatherer?

Davidski said...

Yes, Pitted Ware HG from Gotland.

Maju said...

Notice please that Pitted Ware "hunter-gatherers" are not mere residual HGs from the Paleolithic but actually Neolithic foragers with some of the Neolithic package (at the very least pottery and pigs, possibly some cereals too), who had cultural links (origins?) with Eastern Europe. Pitted Ware clearly derivates from Dniepr-Don Neolithic and is not exclusive from some parts of Sweden but is also known at the other side of the Baltic shores and even inland; it's more a subculture of trappers fishers with clear links to Ukraine and the Don basin.