search this blog

Monday, November 13, 2017

Who's your (proto) daddy Western Europeans?


Considering the increasingly large amounts of paleogenomic data being released online nowadays, it's no longer practical for me to try to highlight most archaeological cultures and even genetic clusters in my Principal Component Analyses (PCA) of the ancient world. Thus, from now on, I'll be focusing attention in such PCA on the main population shifts that have led to the formation of the modern-day West Eurasian gene pool and genetic substructures, like on the PCA plot below, which includes the new Lipson et al. 2017 data (available at the Reich Lab here).


The relevant PCA datasheet can be gotten here. By grouping several hundred ancient samples into just nine clusters, I'm attempting to highlight four key processes and resulting genetic shifts in Europe, the Near East and Central Asia:

- European forger populations mixing with genetically much more southern early farmers of Near Eastern origin, mostly during the Neolithic, bringing about the total disintegration of the Europe to Siberia Hunter-Gatherer cline

- "Old Europeans" getting overrun and largely absorbed by Y-haplogroup R1-rich Kurgan pastoralists from the Pontic-Caspian steppe during the Eneolithic and Bronze Age, leading to the formation of at least one major new cline from the Bronze Age steppe into post-Kurgan expansion Europe

- the ancient Near East "imploding" or becoming significantly more compact in terms of genetic structure, likely due to a variety of major population expansions from the chalcolithic onwards from the eastern and western parts of the Fertile Crescent, as well as probably the Caucasus and Europe (note how the post-Neolithic western Asian cluster stretches out towards Europe)

- fully nomadic and very wide ranging pastoral and warrior cultures dominating the entire Eurasian steppe during the Iron Age, leading to the emergence of progressively more East Asian-admixed populations from west to east across the Eurasian steppe

An interesting outcome of the denser sampling from space and time in West Eurasia is that Y-haplogroup R1b, once so elusive in the ancient DNA record, is now popping up all over the place. The new Lipson et al. dataset, for instance, includes two R1b "Old Europeans" from Blatterhole in Germany dated to the Middle Neolithic. Below is the same PCA as above except with all of the ancients belonging to R1b marked with an X. The two Blatterhole samples are sitting in the largely empty space between the European/Siberian Hunter-Gatherer cline and most of the "Old Europe" cluster. The relevant PCA datasheet is available here.


So it may seem that we're back to square one in the long running effort to pinpoint the origin of Y-haplogroup R1b-L51, which encompasses almost 100% of modern-day Western European R1b lineages, and thus probably ranks as Europe's most common Y-haplogroup. But at this stage I'd say no, because R1b-L51 is a subclade of R1b-M269, of which the oldest sample comes from the Bronze Age steppe. In fact, as can be seen in the above PCA, this sample is sitting in exactly the right spot to be one of those pastoralists who overran "Old Europe", or at least a very close relative thereof.

Or am I wrong? Feel free to let me know in the comments.

I didn't bother creating a similar plot of ancient samples belonging to Y-haplogroup R1a, because, unlike R1b, this marker is still non-existent in samples from outside of Eastern Europe and Siberia dating to before the late Neolithic. And I doubt that this is simply due to a lack of the right ancient material. Moreover, the recent discovery of Y-haplogroup R1a-M417, which encompasses almost 100% of all modern-day R1a lineages on the planet, in a North Pontic steppe sample belonging to the Eneolithic Sredny Stog culture means that it's game over for the naysayers as far as the steppe origin of most modern-day R1a lineages is concerned (see here and here).

In other words, if you're still hoping to see R1a, and especially R1a-M417, pop up in non-steppe derived ancient individuals in, say, such far away places as South Asia, then you'll probably be waiting forever.

For the linguistic implications of all of this, see...

Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...

Update 15/11/2017: After a couple of days of messing around with the Lipson et al. dataset, I'm certain that Late Copper Age sample Protoboleraz_LCA I2788 shows significant steppe-related admixture. This is the only sample from Lipson et al. with such an obvious signal of steppe-related input that had enough data to be analyzed individually by me with PCA and D-stats.


For the time being, amongst the best proxies for this signal appear to be Yamnaya_Samara and Samara_Eneolithic. But it's likely that the real source of the admixture is yet to enter the ancient DNA record, or at least my dataset. When it does, it'll probably be an Eneolithic pastoralist population from the North Pontic steppe.


Yamnaya_Samara also gives the best statistical fit as the single source population in qpAdm (see here). It's an important result, because it suggests that steppe peoples very similar to Yamnaya were already expanding on and out of the steppe as far back as ~3500 BCE, and perhaps a few hundred years earlier.

234 comments:

1 – 200 of 234   Newer›   Newest»
Davidski said...

I'll update these plots with the new samples from Mathieson et al. 2017 and/or Olalde et al. 2017 if they come out soon.

Rob said...

Oldest sample of M269 is from Iberia
Ha !

Rob said...

Nice summary though

Davidski said...

Oldest sample of M269 is from Iberia.

Which one is that?

Rob said...

ATP3. I think it really could be M269 (TMRCA 13000 y), but not *the* M269.

Davidski said...

Nah, that's nowhere near to being confirmed.

Davidski said...

It's just another heavily degraded sample with weird calls. No way does it qualify as the earliest M269 without being re-sequenced to a decent level.

Rob said...

So you've analysed the BAM ?

Davidski said...

Nope. It can't be analyzed properly considering the coverage. No one can do it.

Rob said...

Nobody is a strong claim. I suspect these little outliers willl keep popping up.
But to reiterate, I dont think it's going to change the overall picture

Davidski said...

Unless the M269 call is based on a new, higher quality sequence then it'll always be in doubt. That's why the Y-hg result for this sample hasn't been published in the string of related papers that have come out since it was sequenced.

If there is a new sequence of this sample and it does obviously show M269, then that's fine, but I haven't heard anything about that.

Samuel Andrews said...

Rob, the chance full blown R1b M269 or relatives of it were widely dispersed before 3000 BC is super small. Assuming R1b M269 originated somewhere in eastern Europe the chances it also existed in Iberia are really low.

Samuel Andrews said...

Everybody needs to understand the "new" data from Mathieson 2017 really strongly suggests R1b1a (TMRCA, 19ky, yfull) and all its descendants originated in Mesolithic Europe. This is a big revelation and was a surprise to most here including myself. In no way could it have been decerned using modern DNA.

Mathieson 2017 sequenced DNA from sizable number of individuals from three Mesolithic European communities/small regions within a country: Latvia, Serbia/Romania border, and Ukraine.

All of them were rich in R1b1a-L754. Plus, the oldest example of a WHG individual, Mr. Villabruna who died in Italy 14,000 years ago, belonged to R1b1a-L754.

R1b1a2-V88, which today is mostly found in Africa and West Asia, has popped up in several Mesolithic Europeans. In addition, R1b1a1-P297 and R1b1a1a1-M73 have popped up in Mesolithic Europe. The oldest P297 comes from Latvia and dates 10,000 years old!

With this data, the chances R1b1a1a2 somehow has an origin the CHGs of West Asia unlike its relatives R1b1a1a1 and R1b1a2 are basically 0. It's just a matter of time before R1b1a1a2 pops up in Mesolithic Europe. Maybe Geneticker will find it in one of the Ukraine HGs once their DNA is made public.

Rob said...

@ Sam

You don.t understand what you're saying.
For a start, the Age of formation of M269, and its split from P297 is ~13 KYA, which is a different entity to TMRCA of modern lineages.
Secondly, it is unlikely R1b originated in Mesolithic Europe. For a start, its Age, as you noted, is c. 22 kya, which is not the Mesolithic, nor is it likely the place to have been in Europe.

Rob said...

@ Sam

I disagree. For a start, the 'Age of formation' of M269, and its split from P297 is ~13 KYA, which is a different entity to TMRCA of modern lineages.
Secondly, it is unlikely R1b originated in Mesolithic Europe. For a start, the age of R1b-M343 Age, as you noted, is c. 22 kya, which is not the Mesolithic, nor is it likely the place to have been in Europe.

Samuel Andrews said...

@Rob,
"You don.t understand what you're saying."

You're the one who doesn't understand what I'm saying you retarted looking, lingustically imparied, motherfucker.

First of all, of course I know M269 is estimated to have split from P297 13ky and therefore relatives of modern M269 existed before modern M269's most recent common ancestor. That makes no difference. M269 related lineages probably didn't reside far outside of eastern Europe.

Second of all, I didn't say R1b probably originated in Europe I said R1b1a and its descendants probably originated in Europe.

Rob said...

@ Sam

FYI

* retarted-> retarded
* imparied -> impaired.

:)

rozenfag said...

Paper about Globular Amphorae will be published next week: https://figshare.com/articles/Electronic_Supplementary_Information_from_Genome_diversity_in_the_Neolithic_Globular_Amphorae_culture_and_the_spread_of_Indo-European_languages/5594287

Samuel Andrews said...

Sweet. I just hope we can have fun making theories using modern DNA before every single ancient human specimen has its genome sequenced.

Olympus Mons said...

Yes.... All from the steppe, I guess.


It is unclear whether Indo-European languages in Europe spread from the Pontic steppes in the late Neolithic, or from Anatolia in the early Neolithic. Under the former hypothesis, people of the Globular Amphorae Culture (GAC) would be descended from Eastern ancestors, likely representing the Yamnaya Culture. However, nuclear (six individuals typed for 597 573 SNPs) and mitochondrial (11 complete sequences) DNAs from the GAC appear closer to those of earlier Neolithic groups than to the DNAs of all other populations related with the Pontic steppe migration. Explicit comparisons of alternative demographic models via Approximate Bayesian Computation confirmed this pattern. These results are not in contrast with Late Neolithic gene flow from the Pontic steppes into Central Europe. However, they add nuance to this model, showing that the eastern affinities of the GAC in the archaeological record reflect cultural influences from other groups from the East, rather than the movement of people.


Olympus Mons said...


"...It is unclear whether Indo-European languages in Europe spread from the Pontic steppes in the late Neolithic, or from Anatolia in the early Neolithic. Under the former hypothesis, people of the Globular Amphorae Culture (GAC) would be descended from Eastern ancestors, likely representing the Yamnaya Culture. However, nuclear (six individuals typed for 597 573 SNPs) and mitochondrial (11 complete sequences) DNAs from the GAC appear closer to those of earlier Neolithic groups than to the DNAs of all other populations related with the Pontic steppe migration. Explicit comparisons of alternative demographic models via Approximate Bayesian Computation confirmed this pattern. These results are not in contrast with Late Neolithic gene flow from the Pontic steppes into Central Europe. However, they add nuance to this model, showing that the eastern affinities of the GAC in the archaeological record reflect cultural influences from other groups from the East, rather than the movement of people.

Mike the Jedi said...

I think your simplified labeling method going forward is very sensible, Dave. Most of us have seen enough of these PCAs to memorize where certain milestone samples are anyway (WHG, EHG, Natufian, etc.).

One thing I'm curious about:
East Asian-mixed Volga-Ural people and ASI-mixed South Asians are usually included on your West Eurasian PCAs so I assume the reason you never include North Africans is due to their significant SSA or SSA-like ancestry screwing up the plot somehow? North Africans are predominantly West Eurasian so I can only assume you have a good reason to omit them. I know Lazaridis doesn't put them in his West Eurasian PCAs either...

Basil S said...

Got these stats for Blatterhohle-I ran 3 of them,excluding the Blatterhohle HG:

Khomani Blatterhohle_MN Samara_HG LBK_EN 0.0163 3.456
Khomani Samara_HG Blatterhohle_MN LBK_EN -0.0350 -8.181
Khomani Blatterhohle_MN Karelia_HG LBK_EN 0.0205 5.251
Khomani Karelia_HG Blatterhohle_MN LBK_EN -0.0338 -9.837
Khomani Blatterhohle_MN Samara_Eneolithic LBK_EN 0.0242 6.908
Khomani Samara_Eneolithic Blatterhohle_MN LBK_EN -0.0284 -8.999
Khomani Blatterhohle_MN Samara_Yamnaya LBK_EN 0.0290 10.976
Khomani Samara_Yamnaya Blatterhohle_MN LBK_EN -0.0164 -6.541

Could there be better pops to run against Blatterhohle?

Davidski said...

@Olympus Mons

Blathering idiot, here's what I have to say about Globular Amphora. Enjoy...

Globular Amphora people starkly different from Yamnaya people

@Mike

10-20% Sub-Saharan admixture in too many samples messes up the West Eurasian plot, while up to 30% East and South Asian admixture doesn't, but rather helps to flesh out some more detail in the ancients. I suppose this might have something to do with the bigger difference between Eurasians and Sub-Saharan Africans, then between Eurasians. But I haven't tested it in any detailed way.

@Basil S

I wouldn't use Khomani as an outgroup. I'm pretty sure that they have some West Eurasian admixture.

Grey said...

Olympus Mons

"DNAs from the GAC appear closer to those of earlier Neolithic groups than to the DNAs of all other populations related with the Pontic steppe migration."

If there was pressure from the steppe on the closest farmers to them then one of the first effects might be seen on the opposite borders from the pressure as some of the farmer groups moved away from it towards the west or south.

Helgenes50 said...

nMonte2 with these new samples

[1] "distance%=0.2195 / distance=0.002195"

Bell_Beaker_Germany:average

Blatterhole_MN:I1563 28.75
Latvia_LN1:ZVEJ28 26.85
Boncuklu_N:Bon001 12.60
Iberia_Chalco 9.40
Yamnaya 8.85
LBK_EN:I0048 5.80
Mentese_N:I0726 5.20
Starcevo_EN:I1878 2.55


[1] "distance%=0.5167 / distance=0.005167"

Bell_Beaker_Germany:I0108

Blatterhole_MN:I1563 34.60
Latvia_LN1:ZVEJ28 15.45
Boncuklu_N:Bon001 13.00
Iberia_Chalco 9.40
Yamnaya 8.70
LBK_EN:I0048 7.60
LBK_EN:I2037 5.20
Blatterhole_MN:I1594 4.15
Boncuklu_N:Bon005 1.90

[1] "distance%=0.6425 / distance=0.006425"

Ireland_EBA:Rathlin1

Blatterhole_MN:I1563 31.00
Latvia_LN1:ZVEJ28 30.15
Iberia_Chalco 12.55
Yamnaya 11.70
Boncuklu_N:Bon004 6.80
Mentese_N:I0726 4.95
Boncuklu_N:Bon005 2.85

Davidski said...

Those models look overfitted.

Helgenes50 said...

Those models look overfitted.

By removing the Anatolians, the results would probably be better and more logical

Tesmos said...

The individuals from Blatterhole are probably R1b-(pre)V88. We need more people to confirm these results. However, I think it's clear now that M269 and L51 are from the Steppe.

Chad Rohlfsen said...

Blatterhohle is V88. That ATP3 probably will be too.

Anthro Survey said...

@David
I don't mind using this 9D WestEurasia-specific datasheet for the time being, but how soon will the newer samples therein(Blatterhohle, etc.) get their Global10 treatment?

@Helgenes
nMonte version 2? What's changed? If you don't mind, can you please post download link here?

Chad Rohlfsen said...

Dump Boncuklu. They're long dead and gone.

Rob said...

@ Chad, and rest

Why are you guys stubborn, when none of you have even looked at the actual data ? :)

Matt said...

Very cool how many ancient samples there are on these plots now, and how full they look even without moderns.

Since I haven't done it for a while since you've updated the sheet, few attempts to try and visualize how much group structure there is in this PCA data:

Linear Discriminant Analysis on these PCA dimensions: https://imgur.com/a/ofxgv

PCA on PCA for averages: https://imgur.com/a/qQbeg

Some distances and Dendrograms for averages: https://imgur.com/a/hI4RL

So looks like there's enough structure to discriminate

1) Euro HG from Near East ancient,
2) Western ancients (Levant, Anatolia, WHG) from Eastern ancients (CHG, Iran N, EHG),
3) CHG and Anatolia_N (Central+Ancestral to Europe) related from Scythians, Levant_N, Iran_N related
4) Iberian Farmers, from Hungarian Neolithic farmers and from Levant Farmers.

Some nMonte fits on these data, for Bronze Age Europeans: https://imgur.com/a/ORxRN

Not to be taken too literally, and perhaps there is overfitting, but there are some suggestive patterns in the fits for Irish, German, Hungarian, Iberia BA populations...

The BA Balkans, Great Britain, Baltic and the Welzin_BA samples and others will all add another layer to this.

(Btw, David, is it possible to the Tollense Valley samples on this datasheet? Or not possible and too many issues?)

Olympus Mons said...

@Davidski,
You extreme case of NPD... why would you think I care about you wrote?

Iñigo said...

@Rob

That's not true. The data shows that ATP3 is positive for R so he is very likely R1b, but M269 is not the only option. Given that he doesn't have any steppe ancestry in the autosomes I would bet he is not M269.

Anthro Survey said...

A couple of very crude nMonte runs to get a rough idea of how WHG-resurgent EEFs mingling with steppic newcomers in Bronze Age Central Europe really were. So nice to finally have Blatterhohle samples at our disposal.
They and others like them will prove crucial in elucidating this.

From 2D PCAs, Iberia_Chalco-like WHG-EEFs always seemed inadequate for MLBA-like populations.

[1] "distance%=0.5127 / distance=0.005127"

Beaker Average

Yamnaya Kalmykia average 49.9
Iberia Chalco average 31.1
Blatterhole_MN:I1593 18.9

[1] "distance%=0.5852 / distance=0.005852"

Unetice Average

Yamnaya Kalmykia average 55.5
Iberia Chalco average 23.1
Blatterhole_MN:I1593 21.4

Rob said...

@ Inigo

"Given that he doesn't have any steppe ancestry in the autosomes I would bet he is not M269."

How is that relevant ? M269 is 14 ky old, so of course most M269s which ever existed *won't* have steppe ancestry. R1b didn;t originate on the steppe, i'd bet M269 itself didn't even originate on the steppe, but one branch of it found it's way there by the late Neolithic, whilst others could have been far & wide, then become extinct as their steppe cousins expanded with Yamnaya - BB.

Chad Rohlfsen said...

There is no m269 west of the steppes before 2600 BCE, right? Anyway, ATP had an M269 equivalent, but not M269, IIRC. Could be damage issues too. It's a horrible sample to base an argument on. Go where 99.9% of the data is. Blatterhohle supposedly had M269 equivalents, but now we know that's not the case.

Chad Rohlfsen said...

I just wouldn't base anything on a single call for a sample with 0.01% coverage. Shaky is an understatement.

Chad Rohlfsen said...

Can't forget modern contamination too.

Iñigo said...

@Rob

It is very relevant. M269 is absent in Europe before 2800 BC. After 2800 BC it goes very rapidly to high frequency, always associated with a very clear signal of steppe ancestry in the autosomes. I don't know where R1b or R1b-M269 originated (as you say, they might have originated outside the steppe), but it is now very clear that the M269 appearance in Europe is associated with the steppe migration in the 3rd millennium BC. That's why ATP3, who lacks steppe ancestry, is very likely not M269.

Ryan said...

David - any chance you could highlight which of the samples on the PCA are Bell Beakers? Since the span two categories. Would be interesting see if they form a cline or not.

Davidski said...

@All

The following populations/individuals have been added or updated in the Global 10 datasheets.

ALPc_MN
Baden_LCA
Balaton_Lasinja_CA
Blatterhole_HG
Blatterhole_MN
Germany_MN
Iberia_ChL
Koros_EN
Koros_HG
LBK_EN
LBKT_MN
Protoboleraz_LCA
Starcevo_EN
TDLN
Tisza_LN
Tiszapolgar_ECA
Vinca_MN

https://drive.google.com/file/d/0B9o3EYTdM8lQd1hLRFl4OUdiME0/view?usp=sharing

https://drive.google.com/file/d/0B9o3EYTdM8lQb0ZjM0pLVmtvTFE/view?usp=sharing

Davidski said...

@Ryan

I can do a plot with the German and Iberian Bell Beakers highlighted, but which of the new Iberia_ChL samples are actually Bell Beakers? I need their IDs.

Samuel Andrews said...

@Rob,
"you're just making superfluous comments, and you don't understand the evidence - neither archaeology nor genetics "

Why are you always such a jerk? Btw, to everyone I called Rob a "retarted looking, lingusitically impaired, motherfucker" because he's a jerk.

Dis agreement, especially on a subject so frivilous as the origins of R1b, is no reason for name calling.

Samuel Andrews said...

@Rob,

Here's something else to consider. We get the impression it is possible R1b1a in its early days could have resided almost anywhere in the world because of recent expansions which made it widespread.

yHG I2a was the other major yHG in Mesolithic (eastern) Europe. Why couldn't R1b1a be just as native to Europe, but unlike I2a, later made big expansions into Asia making it appear possible it has an old history there?

I don't like to make big conclusions without a whole lot of data. I'm not arguing it has been proven without any shadow of a doubt that R1b1a strictly stayed within the broders of Europe in its early days. All I'm saying is the data suggests R1b1a, along with I2a, is more or less a Paleo/Mesol Euro HG lineage.

Davidski said...

@All

Interesting new paper on winemaking in the Shulaveri-Shomutepe Culture of the south Caucasus. Perhaps the first winemaking anywhere.

Early Neolithic wine of Georgia in the South Caucasus

http://www.pnas.org/content/early/2017/11/07/1714728114.full

Samuel Andrews said...

@Rob,
"The difference is I respond to their comments, and lack of susbtance in them ."

If you haven't noticed you're getting your butt kicked right now. "Duh duh R1b M269 in Chalcolithic Iberia kablonga."

"On the other hand, you just meltdown because you are projecting what you yourself are. Your psychopahtology (mental disorder) is very transparent."

More name calling. Even if I did have a mental disorder, which I don't, I wouldn't deserve name calling. Rob, it seems you value people on the basis of intelligence not the simple fact they're human beings and therefore deserve a basic level of respect. Disgusting. I don't measure my self worth on my intelligence or any ability. Your insults have absoultly no affect on me.

I don't believe I or the many other people you have called mentally ill or whatever are. I just think you're a cold, cruel person. Not suprised, the academic world has a reputation for creating people like you. You think education, money is end all in life, well it isn't.

Rob said...

@ Dave

I think the source of CHG/ Near Eastern admixture in pre-Yamnaya will be difficult to fully elucidate. We're looking at individuals moving from east Anatolia as early as 6500 BC, which left no genome-wide impact , and will be otherwise undetectable apart from future lucky uniparental clues. This early wave was limited to the adoption of ceramics, and spread via a networks between the Don & Volga rivers. Some later movements from the Aral region are detectable in the middle Volga.
The genome-wide impact only manifests after c. 4500 BC as the animal -productive component of the Neolithic package was adopted along the southern part of the western steppe, facilitating greater population growth.

Ric Hern said...

Where did all the R1bs go during the Neolithic Farmer migration into the Balkans ? The only way clear seems to be North of the Carpathians. If the oldest R1b was found in Latvia and the youngest in the Ukraine then surely R1b M269 originated somewhere inbetween. Maybe in Belarus ? Or maybe just north of GAC and West of the Dnieper..?

Davidski said...

Some of the Balkan forager R1b lineages, usually those belonging to R1b-V88, were incorporated into the Neolithic farmer populations that moved into Central Europe and even Iberia, and became the later WHG-rich Middle Neolithic farmers and Chalcolithic/Copper Age groups there. It looks like those that stayed in the Balkans went extinct.

And I doubt that R1b was in Latvia before it was in Ukraine and southern Russia.

Simon_W said...

Something basic to be considered when talking about yDNA haplogroups - what exactly do we refer to when talking e.g. about R1b-M269? Do we mean people who have precisely and positively this marker? Or do we mean people who have this marker and at the same time all 106 phylogenetically equivalent markers? That's not a trivial question. And related to this is the correct understanding of the older date when it "was formed" according to yfull, in contrast to the TMRCA. The TMRCA is obviously the TMRCA of the tested living people. These share all the phylogenetically equivalent markers. The older date on the other hand, when the haplogroup "was formed", is identical to the TMRCA of the haplogroup that is immediately upstream. It's logical that it cannot have formed earlier than its parental clade. But why is it supposed that it "formed" immediately after the TMRCA of the parental clade? Technically this is impossible. It's impossible that 107 mutations occured within a few years. So to my understanding that date when the haplogroup "formed" means just the beginning of its evolution, the start of its divergence, when maybe the first of the 107 mutations occured, nothing more. So in a way it's not correct to assume that R1b-M269 people were running around 13300 years ago. There must have been people who had one or more of the phylogenetically equivalent markers. Maybe even M269, though it's unlikely given the sheer number of equivalent markers. R1b-M269 in the strict sense, with all the equivalent markers, must be much younger, and hardly much older than the TMRCA.

Simon_W said...

When exactly the M269 mutation arose is impossible to say, it may have been at any time between 13300 and 6400 years ago. And actually it's not even important because all equivalent markers are equally important.

Davidski said...

Yeah, in practice, focusing on one SNP like M269 or L51 is just shorthand for referring to a whole chunk of the R1b phylogeny and everything downstream.

But it does work, because it certainly makes sense to say that L51 is the most common R1b subclade in Western Europe, and Z2103 is the most common R1b subclade in the Balkans and West Asia.

Davidski said...

@Matt

The same datasheet with Welzin_BA...

https://drive.google.com/file/d/1L1DTCpD_aA7G5CYO82iNnMQaFSTCr3a4/view?usp=sharing

Aram said...

Davidski thanks for that paper about wine.

Now this paper makes some disputable assumptions. Areni cave is assumed to be part of Shulaveri horizon. But in most likelihood it is not the case!
Our archaeologists published their summary when they came to the conclusion that Areni seems to be influenced from Iran. And it was in __sharp__ discontinuity ( Areni is a late Eneolithic ) from the preceding Early Eneolithic and Neolithic (Shulaveri) culture.

Thus Areni is probably not Shulaverians but those who pushed Shulaverians to North.

And the presence of similar jars in Areni and Haji Firuz Tepe (Iran) and their absence from true Shulaveri sites simply proves what I said.


--------
One disparity between the analyses of Hajji Firuz and Georgian jars is that the latter showed no signs of a tree resin or any other additive, according to the GC-MS analyses. Pine and terebinth saps were commonly added to wine throughout antiquity. They acted as antioxidants to keep the wine from going to vinegar, or barring that, to cover up offensive aromas and tastes. The tradition continues today only in Greece as retsina.

The Hajji Firuz jars were found partly buried in an earthen floor. No evidence has yet been found of how the Shulaveri and Gadachrili jars were positioned or whether they were partly or fully buried underground, as is the common practice for making so-called qvevri (“large jar”) wine

....


The breakthrough came when numerous underground jars were found inside caves at Areni in a mountainous region of Armenia (37). Desiccated (uncarbonized) grapevine wood, dating to ca. 4,000 BC, together with pips and chemical evidence by LC-MS-MS of tartaric acid/tartrate and the red pigment malvidin, left no doubt that we now had partial evidence for the previously “empty” transitional period. The technology was ingenious: humans had laid out plaster floors for pressing the grapes and running the unfiltered juice into underground jars. Whether similar evidence will eventually be found in Georgia and Azerbaijan, elsewhere in the SSC area, or in the extended mountainous region remains to be seen.

Davidski said...

@Rob and Samuel

Please calm down, or better yet, take things to e-mail and stop communicating on this blog.

Matt said...

@Davidski, thanks for that.

Btw, here's the same Linear Discriminant Analysis of these PCA Dimensions, with the Welzin BA samples: https://imgur.com/a/iPon3

(In the Linear Discriminant Analysis Axis:

1 and 2: Welzin_BA slightly less steppe shifted than other BA populations, slightly more high WHG+EEF shifted.

3: Welzin_BA Diverse (not clear how meaningful this Axis is for Euro ancient samples)

4: Diverse again, more so than other BA populations.

a said...

@Davidski said..
"But it does work, because it certainly makes sense to say that L51 is the most common R1b subclade in Western Europe, and Z2103 is the most common R1b subclade in the Balkans and West Asia."

Comedy central, comedy relief ,entertainment. Like Harvard handing out an award, giving a speech of recognition in the Germanic-Anglo-Saxon/Romance dialect- and then rescinding it. Who knew ?
With Villabruna-Italy R1b-L754 and Blatterhohle Cave-Germany R1b-V88 samples clustering/plotting so far on the PCA into Western Europe, the push to prove R1b-V88 being derived in Africa and Levant-centrists and by the Afro-Nufians + LGBQUE fanatics, has gone the way of the Dodo bird.



Variance in R1b snps L754+ M269+ L23+Z2103 are centered around Volga.
R1b-Z2103 is also very common to the East of the above cluster in an area also known as Arkaim, aka the bad-ass warrior region. More comedy relief.
Now we have selective PCA discriminate analysis. Lets parse that into a contemporary definition of Europe, and European Neaderthal components.

Chris Davies said...

Apparently R1b that is neither R-V88 nor R-M269 exists in North Africa.

"R1b1a (R-L754)
R1b1a (R-L754) was carried by an individual known as Villabruna 1, who lived circa 14,000 BP in north-east Italy, and belonged to the Epigravettian culture.
Living individuals positive for L761, an equivalent to L754, have been found at high frequencies among the Toubou population inhabiting Chad (34%)."*

*Haber, Marc; et al. (2016). "Chad Genetic Diversity Reveals an African History Marked by Multiple Holocene Eurasian Migrations". American Journal of Human Genetics. 99 (6): 1316–1324. Retrieved 27 June 2017. - Y-chromosomal haplogroup frequencies on Table S.4

ISOGG 2017:-
R1b1a L754/PF6269/YSC0000022, A702, CTS3063/V2515, CTS3794/PF6256/S4229, CTS4244/PF6257/YSC0001279, CTS4764/PF6041, CTS7585, CTS8436/PF6259, CTS8612/PF6260, CTS9972/PF6261, FGC35, FGC36/Y409, FGC41/M12190/Y108, L761/PF6258/YSC0000266, L820/PF6262/YSC0000211, L1068/PF6264/YSC0000223, L1345/PF6266/YSC0000224, PF6249, PF6263, PF6271

Roy King said...

The discovery of viniculture in Neolithic Georgia circa 6000-5800 BCE is an important and somewhat amazing finding. Could this explain J2a in the Levant, Anatolia, Eastern Mediterranean, Aegean and perhaps even Lengyel horizon in Hungary? We need an explanation for J2a in these areas post Early Neolithic Era. Wandering sommeliers could be the answer!

Iñigo said...

“M269 will be present in parts of Europe- often distant- at low levels prior to the Bronze Age expansion of one lucky M269 group”

Maybe. We will see if that holds true when more individuals are analyzed, but we already have a fair amount of samples before 2800 BC and none has M269, so there are reasons to believe that it was absent in Europe until the Bronze Age expansion.

Rob said...

@ Inigo.
We don;t know if it will or won't. Perhaps 'wide' is stretching it, but according to branching & development as Simon explained it certainly is possible.
Btw, when you say " there are reasons to believe that it was absent in Europe until the Bronze Age expansion" do you mean that M269 came from the Near East or Asia ?

Anthro Survey said...

@Aram

Do you think Shulaveri people at the time(circa 5800BC) were essentially CHG-like or hybrids of CHG and Tepecik-to-Barcin_N?

If the latter,then it rules the area out as a source for steppe's CHG.

If it turns out that Anatolia_N-like ancestry was also established in the Circassian foothills a millenium later(by 4800BC or so), it can be ruled out as well----at least after that date. It's a reasonable assumption,imo, because it's hard to envision Tepecik-like ancestry making its way there after Neolithic times. Intuitively, it's also tempting to speculate that Maykop had Tepecik/Barcin ANF because of that lineage's association with "civilization-building".
Now, perhaps colonies of CHGs established themselves to the north(e.g.Azov) PRIOR to ANFs expansion into the Caucasus to eventually hybridize with EHGs and produce Srendy Stog.

The limited samples/data we have from Mathieson suggest Sredni Stog area had varied amts of EEF but it was surely from Cucuteni-Tripolye. Yamnaya_Samara and Kalkymia(3000-2700BC) practically lack it when I put them through nMonte, though, and prefer a bit of WHG, not Barcin. If formal stats are giving them Barcin, it may be from the WHG-like component of it.

Grey said...

Chris Davies
"Living individuals positive for L761, an equivalent to L754, have been found at high frequencies among the Toubou population inhabiting Chad (34%)"

they have an interesting look

https://www.google.co.uk/search?q=toubou+tribe&safe=off&dcr=0&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwiZ5Ke06L7XAhXpB8AKHZdwCAQQsAQIMw&biw=1280&bih=614#imgrc=_

i may be imagining it but is there a hint of eyefolds?

(there are/were apparently some little groups tucked away in the remotest bits of the Sahara who look/looked more like Khoisan)

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

it says they (or at least the ydna part) arrived in the region from around 600BC so where they came from before that might be a clue to something.

Grey said...

Roy King said...
"Wandering sommeliers could be the answer!"

that's one of the most interesting aspects of this to me - could some of the various low frequency but widespread haplogroups be a legacy of a particular innovation being spread by a family group over time i.e. eldest son inherits family business and younger sons sent off to set up a franchise in an adjacent node of the local trade network.

so maybe pots vs people is almost always people but in one of two forms: 1) mass migration and 2) artisanal family groups.

capra internetensis said...

@Chris Davies

The Chadian L761 is L761(xP297), apparently not tested for V88 (!?). Anyway 3 of them are fully sequenced and all those are V88.

Lot of people on Wikipedia don't understand the concept of a paragroup, genetics articles are full of newbie errors.

Samuel Andrews said...

Blätterhöhle BLA 16
Y DNA R1b1a2-V88
mtDNA U5b2a2*

Me
Y DNA R1b1a1a2a1a2a-Df27*
mtDNA U5b2a2b1

Which uniparental marker in me and BLA 16 is more related, mtDNA or Y DNA? For a lot of you, I think the answer would be mTDNA. But actually, our mtDNA and Y DNA are equally related.

U5b2a2 and R1b1a are both about the same age, both emerged 15,000-20,000 years ago. Many factors cause us to miss interpret the level of relation between different mHGs and yHGs.

Because founder effects and regional variety is more common in yHGs than mHGs we tend to exaggerate the distance between two subclades of the same yHG and exaggerate the closeness between two subclades of the same mHG.

Here's another example. U5b2a, which reaches peak frequencies at 2% in Baltics and Britan Poland and SCandinavia, is about as old as mHG H which ranges from 40-50% across Europe and 20% throughout most of the Middle East. Because mHG H is so widespread and has so many regional specific subclades people will likely see U5b2a1 and U5b2a2 as more related than H5 and H6.

Samuel Andrews said...

As we get more and more ancient DNA, I bet we'll learn mHG H decends from the very specific Paleolithic population that contributed greatly to modern West Eurasians like mHG U5 decends from WHG. But because mHG H is so widespread and is popular in such distinct populations people have a hard time imagining at one time it was specific to one population/genetic cluster.

supernord said...

"The limited samples/data we have from Mathieson suggest Sredni Stog area had varied amts of EEF but it was surely from Cucuteni-Tripolye"

There is no sample belonging to the classical the Sredniy Stog culture, and even the classical the Dereivka culture. There is only one sample from Alexandria related either to late or after the Dereivka culture. The Sredniy Stog people was not derive from the preceding Neolithic population of the Ukraine.

Seinundzeit said...

David,

"10-20% Sub-Saharan admixture in too many samples messes up the West Eurasian plot, while up to 30% East and South Asian admixture doesn't, but rather helps to flesh out some more detail in the ancients. I suppose this might have something to do with the bigger difference between Eurasians and Sub-Saharan Africans, then between Eurasians. But I haven't tested it in any detailed way."

That's pretty interesting.

An additional factor could be that most ancient West Eurasians had some amount of ENA admixture (based on current modelling, ENA admixture is baked into the ancestry of nearly all West Eurasians), while none sampled yet has displayed any African admixture.

Ryan said...

@David - "I can do a plot with the German and Iberian Bell Beakers highlighted, but which of the new Iberia_ChL samples are actually Bell Beakers? I need their IDs."

I'll try dig into this though I'm pretty screwed for time for the next bit. Maybe just highlighting all Iberian CHL and Bell Beaker samples (and labeling them as such)? I guess without Olade's data there's not much chance of a cline though is there as we don't have any of those intermediate samples.

If you want to "settle" the R1b issue (at least as best as it can be settled until many more high coverage samples become available), you may want to look at the sex bias estimates that Mathieson did. He did qpAdmn first for autosomal DNA and then for X-chromosome only and compared the two results to estimate the proportion of their male ancestry that came from hunter gatherers and the proportion of their female ancestry that came from hunter gatherers. I'd think you could easily do something similar to see if the CWC ancestry in Bell Beakers is primarily from CWC men or from CWC women. If the CWC ancestry is overwhelming from CWC men then there's little doubt L51 comes from them (or a closely related group). If the CWC ancestry is overwhelmingly from women then it's probably not from CWC/steppe men, and instead Z2103 would have been the one that introgressed into the steppe from WHG-like populations.

Obviously if the mixing was relatively balanced that wouldn't solve anything, but for the 15 (unpooled) neolithic/chalcolithic populations Mathieson did this for, 5 came back as exclusively male in terms of HG ancestry (Iberia Chalcolithic, Central_MN, Balkans Chalcolithic, Varna) and two came back as exclusively or nearly exclusively female in terms of HG ancestry (Tyrpillia and LBK_Austria).

One thing I just noted from Mathieson's figure 3 that may be relevant and is at least interesting is that Iberian MN was more "eastern" (closer to EHG vs WHG) than Central European MN and Globular Amphora, but between the Middle Neolithic and Chalcolithic this shifts back towards WHG. They show Iberia MN as modeled as descended from Malak Preslavets too - anyone know what the hell is up with all of that? It seems there's a story there but I don't know what it is. Danubian Neolithic to Iberia rather than Cardial?

Here's a link for anyone who is interested so you don't need to google it like I did.

https://www.biorxiv.org/content/biorxiv/early/2017/05/30/135616.full.pdf
https://www.biorxiv.org/highwire/filestream/57903/field_highwire_adjunct_files/1/135616-2.xlsx

Davidski said...

@Ryan

Central European Beakers = black dots; Iberia_ChL = black circles; Iberia/Portugal_BA = black stars

https://drive.google.com/file/d/1iBIDvgG2ofWcB5mr0sNN1yMjCufBGknB/view?usp=sharing

I have attempted to test the X chromosomes of both Corded Ware and Beaker samples. There's very little data to work with, so the error margins are huge, and honestly overall I can't see a clear difference between the X and the autosomes in terms of steppe ancestry.

Olympus Mons said...

@Davidsky
Which should be read as it did not yield the results you wanted.

Why dont you share it anyway here in a post? Things are quiet enough.

Davidski said...

Yeah, maybe you're right and both R1a-M417 and R1b-M269 are in fact from Anatolia.

Bahaha...

Ryan said...

Mathieson did do autosome vs X chromosome qpAdm models for CWC and German Bell Beakers though by the way. They breakdown is between WHG, AN and Yamnaya, so maybe of limited usefulness since the WHG and AN could have come from multiple sources in each group, and it may be missing any excess EHG (there's an early bronze age sample from Hungary with excess EHG but 0 excess CHG, so wild EHG may have been present in the area), but it's something.

They have CWC as 10.6% WHG 20.2% AN and 69.1% Yamnaya on the autosome and 25.3% WHG 25.5% AN 49.2% Yamnaya on the X chromosome.

Bell Beaker Germany they have as 15% WHG 37% AN and 48% Yamnaya on the autosome and 14.9% WHG 51.3% AN 33.8% Yamnaya on the X chromosome.

If we wanted to model Bell Beakers as a linear combination of CWC + population Z at proportion X, we would have the following system of equations for the autosome:

10.6(1-x)+wx=15, 20.2(1-x)+ax=37, 69.1(1-x)+yx=48, a+w+y=100

Now let me make an assumption here - which is that y (ie the proportion of pre-CWC Yamnaya ancestry in Bell Beakers) is 0. If there's wild EHG inflating Yamnaya ancestry and depressing AN ancestry that got in there this assumption could be wrong, but let's hope it's not very wrong at least. That simplifies things to this:

10.6(1-x)+wx=15, 20.2(1-x)+ax=37, 69.1(1-x)=48, a+w=100

That lets us calculate x directly as 0.305.

So now we have Z-WHG=w=25.0 and Z-AN=a=75.2. That's pretty close to adding up to 100 so I guess those numbers make sense?

Now if I do the same for the X chromosome. I'll have 25.3(1-x)+wx=14.9, 25.5(1-x)+ax=37, 49.2(1-x)+yx=33.8 which for y=0 gives me x=0.313. That gives me a=-7.9 and w=62.2%.

So that makes 0 sense. What the hell did I do wrong or if I didn't do anything wrong what the hell does that mean? That the "Z Beakers" had some excess Yamnaya even before CWC? That CWC isn't a good proxy for the Yamnaya side of German Bell Beakers' ancestry?

Ryan said...

@David - "Central European Beakers = black dots; Iberia_ChL = black circles; Iberia/Portugal_BA = black stars"

Thanks! I guess it does form a a bit of cline doesn't it. Those two beakers to the right of the others - is there anything special about them as far as you know?

"I have attempted to test the X chromosomes of both Corded Ware and Beaker samples. There's very little data to work with, so the error margins are huge, and honestly overall I can't see a clear difference between the X and the autosomes in terms of steppe ancestry."

As you can see above my back of the envelope attempt failed as well lol. I think based on Mathieson's numbers it's pretty clear that there is at least a bit of a male-sex bias for Yamnaya ancestry in both CWC and German Bell Beakers, but I'm not sure what any difference between the bias in each group means if anything.

Maybe we should all be paying more attention to the Makó/Kosihy-Caka culture.

Davidski said...

@Ryan

These are the two outlier Beaker samples.

Bell_Beaker_Germany.SG RISE559.SG F H46 .. 0.127 coverage 140816 SNPs
Bell_Beaker_Germany.SG RISE564.SG M H-T16311C R1b1a1a2a1 0.079 coverage 91035 SNPs

There are two interesting, or at least note worthy things about them: 1) they both appear to belong to farmer-derived mtDNA lineages and 2) both are very low coverage sequences.

Davidski said...

I'm seeing a clear signal of steppe admixture in Protoboleraz_LCA I2788. This individual is a clear outlier on my plot relative to the other Protoboleraz samples, and indeed all "Old Europeans".

https://drive.google.com/file/d/1CH0BeEEIUzJIWFJodAgFi1s13-W2JZhG/view?usp=sharing

So it looks like some groups from the steppe made it onto the Hungarian Plain by ~3700 BCE.

Chad Rohlfsen said...

Koros EN outlier is very interesting too.

Rob said...

I'lll go through them too. Because it gets complex to know what all the cultures & chronologies are

Chad Rohlfsen said...

Other interesting samples

German MN I0802
ALPc MN I2357
TDLN I1902
TDLN I4183
TDLN I2352
LBK I2022
Starcevo I1876

Chad Rohlfsen said...

This Admixture run I have has Protoboleraz less than those above. That sample shows 3% Steppe and 6% Iran.

Chad Rohlfsen said...

I'll graph and qpAdm what I can in the next couple days.

Chad Rohlfsen said...

German MN I0800 and I0802 at 26 and 24% Steppe. Might be coverage though.

Rob said...

@ Chad which ones are I0800 & 802 ?

Davidski said...

@Chad

Almost all of the samples you pointed out don't have enough markers to be plotted accurately in my PCA.

Davidski said...

If you're using qpAdm, I'd be very wary of qpAdm results based on less than 25K SNPs, or even 50K SNPs.

Alberto said...

On the new samples added to Global 10 I didn't find anything too outstanding after a quick run. The biggest outliers:

ALPc_MN:I2378 with 5.7% EHG
ALPc_MN:I1499 with 5.6% EHG
Protoboleraz_LCA:I2788 with 5.4% EHG + 2.3% CHG + 2.4% Iran_N

Nothing that can't be explained by the samples from the Balkans, I'd guess.

Rob said...

ALPc is Alfold linear Pottery, the variant of LBK east of the Danube, tucked within the mountains. Surprising to see EHG this early, no matter how minor. But given all the I2a in it, perhaps not too surprising. The eastern Carpathian range was prObably a refuge for HGs of I2a2 and R1b-V88 type

Alberto said...

Yes, these samples also have some 10-17% WHG, so given what we've seen from the Balkans and Baltic, it's probably some SHG-like admixture from the area rather than any steppe EHG proper.

Davidski said...

@Alberto

Protoboleraz_LCA I2788 has steppe admixture.

It's not SHG, nothing from the Balkans, but actual steppe stuff from the Eneolithic steppe.

supernord said...

"Protoboleraz_LCA I2788 has steppe admixture."

This is normal. In Hungary the time there is the burial of the Sredniy Stog culture.

Rob said...

Blatterhohle don;t make much of a diff. for BB

Bell_Beaker_Germany:I0108
Germany_MN 43.1 %
Yamnaya_Samara 22.9 %
Latvia_HG:ZVEJ32 22.45 %
Armenia_EBA 10.5 %
Dai 0.5 %

Bell_Beaker_Germany:I0112
Yamnaya_Samara 29.45 %
Hungary_CA:I1497 27.95 %
Latvia_HG:ZVEJ32 24.35 %
Iberia_ChL 8.05 %



let's wait for GAC or Romanian Eneolithics

Rob said...

The ProboBeloraz don;t have any steppe admixture.
They're basically
ANF : 78-92 %
WHG: 8-12*
One (#12788) has ~ 6% ANE.


@ SN
"This is normal. In Hungary the time there is the burial of the Sredniy Stog culture."
No there was not. The Suvorovo burials in the Carpathian basin date to. c 44/4100 BC
Proto-Boleraz is 2 - 4 hundreds of years later. There were no 'steppe burials' (which disappeared like Varna) at this time, only Cernavoda & Usatavo culture.

Rob said...

None of the new Hungarian MNE - Copper Age specimens have anything from Samara steppe. Their HG admixture is mostly barn-standard WHG and that kind of latvia_HG which probably represents a HG region region from the outer-Carpathians through Poland to East Baltic. Probably reciprocal exchange between Polish Neolithic cultures and Tisza Copper bearers.

Germany MNE: standard ANF & WHG, with the Blatterhohle gnoting slight eastern affinities probably from Baltic HGs (? assoc. with R*)

Davidski said...

Protoboleraz_LCA I2788 does have steppe admixture. I've checked this out in detail now.

supernord said...

@Rob, don't not write what I do not know. Not rip bits of information, not owning it.

This burial is called Csongrad (in Romania is other). It was naturally a earlier than Protoboleraz culture. So we see that the influence of the steppe is extremely small, that is, it's been a long time.

Rob said...

Not really seeing any 'steppe' admixture in it, tbh
On this occasions, it's pseudo-steppe stuff , due to Asian CHG and North Euro baltHG

Protoboleraz_LCA:I2788
LBK_EN:I0056 65.65 %
Tepecik_Ciftlik_N:Tep006 10.15 %
Latvia_HG:ZVEJ32 9.35 %
Villabruna:I9030 8.6 %
Iran_Neolithic:I1290 4.25 %


Rob said...

@ SuperNord

Not really sure what you're saying about rips of bits, but the CsongradKettashalom burial you're referring to "recently obtained RC date of C-k grave is 4370 BC-4240 cal BC, in good corrrelation with Ochre Grave Data, parrallel with the MCA Bodrogkreeztur culture. .. In E.E. this is the period of ...Khvalynsk and Skelya cultures'
Horvath, Dani et al. MultiDsci. Contributionto Study of Pit Grave Culture .."

This is 600 years before Boleraz. It was an isolated phenom., after which steppe-type burials only re-appear in the pre-Yamnaya/ early Yamnaya period c. 3300 BC

supernord said...

This effect is not calculated with qpAdm. qpAdm is modelling. It is calculated using D-stats. Need to calculate D-stats like template
Mbuti Yamnaya_Samara I2788 WHG
Mbuti Yamnaya_Samara I2788 EHG
Mbuti Yamnaya_Samara I2788 EEF
Mbuti I2788 Yamnaya_Samara WHG
Mbuti I2788 Yamnaya_Samara EHG
Mbuti I2788 Yamnaya_Samara EEF
etc...

@Rob, you in qpAdm results don't write P-value! Then you do not understand. It's MAIN value!

supernord said...

@Rob, "This is 600 years before Boleraz. It was an isolated phenom., after which steppe-type burials only re-appear in the pre-Yamnaya/ early Yamnaya period c. 3300 BC"

1. Stepp influence is not equal Yamnaya influence.
2. I2788 has date 3909-3651 calBCE (4960±40 BP, VERA-5402). Between they may be only 300 years.
3. This is not isolated phenom.. There was another Dacia Muresului in the north of Romania.

Rob said...

Yes i agree it has EHG, but that's not 'steppe' (CHG/EHG), is it ?
The I2c haplogroups of these individuaöls supports my models

Alberto said...

@Davidski

I wouldn't argue strongly either way just based on autosomal genetic data. It's really difficult to say if a small amount of EHG and CHG came from a Yamnaya-related population or not. Unless it shows a typical steppe uniparental marker.

I think this will be better known when we get a more precise chronology of Ukraine. It's still unclear when exactly a Yamnaya-like population arrived there. We have a sample from the eastern edge of Ukraine that might be from around 3500 BC. We'd need more samples from across Ukraine from 4000-3000 BC to establish a chronology and then we can see if it's plausible to find steppe admixture in Hungary around 3700 BC or not. For now I'd be sceptic to call it steppe.

Rob said...

@ <SuperNord

''@Rob, "This is 600 years before Boleraz. It was an isolated phenom., after which steppe-type burials only re-appear in the pre-Yamnaya/ early Yamnaya period c. 3300 BC"

1. Stepp influence is not equal Yamnaya influence.
2. I2788 has date 3909-3651 calBCE (4960±40 BP, VERA-5402). Between they may be only 300 years.
3. This is not isolated phenom.. There was another Dacia Muresului in the north of Romani''

1. Agree ! it could be as early as 4200 BC

2. it may only be 300 years what ?

3. Yes i know that very well. Dont worry

Regardless, the S-N/ West was a precocius event. We still don't ye tknow what individuals of that would 'look like' autosomally. They could be 'steppe'', or they could be more on the WHG/ANF spectrum
Whatever the case, the Boleraz culutre dates #after 4000 BC, and is syncrhonous with Cernavoda. IMO neither the autosomal output nor his-Hg argues for real 'steppe' admixture, but am happy to be convinced otherwise. It would certainly be possible, with the caveat I made about how S-N might look

supernord said...

@Rob.

Your model is overly complicated.
It is anachronistic, Iran_Neolithic obviously flied on the plane.
It is obvious that the program gives you this incredible model (i.e. wrong), so you do not cite the P-values without which the model simply does not exist.
Therefore the models go in the trash.



"Yes i agree it has EHG, but that's not 'steppe' (CHG/EHG), is it ?"

Protoboleraz_LCA:I2788 has got CHG-liked! See even your model. (Iran_Neolithic belongs to CHG/Iran_Chl cluster).

Davidski said...

@supernord

result: Mbuti Yamnaya_Samara I2788 WHG 0.0210 5.668 800553
result: Mbuti Yamnaya_Samara I2788 EHG 0.0481 12.209 750959
result: Mbuti Yamnaya_Samara I2788 LBK_EN -0.0081 -2.751 801079

result: Mbuti I2788 Yamnaya_Samara WHG 0.0257 6.692 800553
result: Mbuti I2788 Yamnaya_Samara EHG -0.0063 -1.651 750959
result: Mbuti I2788 Yamnaya_Samara LBK_EN 0.0436 16.538 801079

Rob said...

@ Supernord, the fit is decent. I only used nMonte.

Protoboleraz_LCA:I2788
Barcin_Neolithic:I1099 64.4 %
Latvia_HG:ZVEJ32 13.1 %
Levant_N 10.1 %
Ukraine_HG1:StPet2 7.9 %
Loschbour:Loschbour 2.8 %
Iran_Neolithic:I1290 1.7 %

(other source groups included, but were Zero: Bichon:Bichon 0 % Kotias:KK1 0 % Samara_HG:I0124 0 % Koros_HG:I1507 0 % Koros_HG:I4971 0 %). Dist : 0.001

nMonte usually agrees with formal/ published results, and what we're seeing here is similar to what we've already seen with Hungary samples - elevated WHG/SHG with minimal CHG. This is not 'Yamnaya' admixture.

Rob said...

If you want we can bet on it, and see what shows up when S-N & Cernavoda are sequenced.

Salden said...

https://www.nature.com/articles/s41598-017-15480-9

Paper on the Ancient Iberian Peninsula is out.

Davidski said...

My PCA shows that compared to the rest of Protoboleraz_LCA, I2788 has admixture from Steppe Eneolithic/EBA, rather than EHG. Note the trajectory of the shift from Protoboleraz_LCA to I2788.

https://drive.google.com/file/d/171GoWxIJwv-dRVltvxHvZh7yaVcEQAIx/view?usp=sharing

And that's what my D-stats show as well...

result: Baden_LCA CHG Protoboleraz_LCA I2788 0.0145 3.902 775967
result: Baden_LCA EHG Protoboleraz_LCA I2788 0.0160 4.025 729048
result: Baden_LCA Samara_Eneolithic Protoboleraz_LCA I2788 0.0196 4.977 580740
result: Baden_LCA Tepecik_Ciftlik_N Protoboleraz_LCA I2788 0.0002 0.051 571261
result: Baden_LCA Yamnaya_Samara Protoboleraz_LCA I2788 0.0175 6.299 773810

So there's nothing really to discuss here. And there's no need for any skepticism (or rather denial), considering that the Hungarian Plain is basically an extension of the Pontic-Caspian steppe.

Davidski said...

@Rob

This is not 'Yamnaya' admixture.

It's actually very, very similar.

Rob said...

Dave,
Do you consider it problematic to model Boleraz on something much younger than it ?

Davidski said...

@Rob

I consider it problematic to overfit models with several very similar reference samples, some of which don't seem even remotely chronologically or geographically relevant (Latvia_HG?!).

And here's another set of the same D-stats using the older TDLN as a reference. There's nothing to discuss here.

result: TDLN CHG Protoboleraz_LCA I2788 0.0180 4.753 761613
result: TDLN EHG Protoboleraz_LCA I2788 0.0199 4.758 716225
result: TDLN Samara_Eneolithic Protoboleraz_LCA I2788 0.0214 5.245 573159
result: TDLN Tepecik_Ciftlik_N Protoboleraz_LCA I2788 0.0039 1.130 561791
result: TDLN Yamnaya_Samara Protoboleraz_LCA I2788 0.0208 6.878 759864

Chris Davies said...

@ capra internetensis -

"The Chadian L761 is L761(xP297), apparently not tested for V88 (!?). Anyway 3 of them are fully sequenced and all those are V88.

Lot of people on Wikipedia don't understand the concept of a paragroup, genetics articles are full of newbie errors."

I emailed Marc Haber, author of the paper. Turns out that the Chad populations were additionally tested for V88 using TaqMan assays. Table S4 showed only SNPs found on the Illumina array.

Rob said...

@ Davidski

Obviously the Latvia HG is a proxy. Similar groups probably existed in Poland and the Carpathian range, which is where the they received most of their 'eastern' shift from.
If you wish to model Boleraz (3800 BC) on Yamnaya samples from 2700 BC, then good for you.

Davidski said...

@Rob

I can't think why Latvia_HG would be a useful proxy in this case, but I can certainly see why Yamnaya_Samara is producing that strong signal in the D-stats, considering the existence of Yamnaya-like people on the steppe during the Eneolithic at Khvalynsk and Alexandria.

Rob said...

Sure that's possible, aso why don;t you use Khvalynsk in your proof, then ? - oh wait you did
You can't make a convincing arguement from something which didn't exist.

supernord said...

@Rob
"Yamnaya samples from 2700 BC"

Yamnaya_Samara is proxy for the Steppe component, because we have not other samples from eneolithic Steppe (any Sredniy Stog/Dereivka/other Eastern European cultures after 4500 BC). But it is known that the Steppe component already existed at this time.

Davidski said...

@Rob

One of the Khvalynsk samples is very similar to Yamnaya.

It's the other two that aren't that are messing up the stats here, because they have too much EHG.

And I'm willing to bet that the Eneolithic R1a-M417 guy from Alexandria will do very well in these stats. He might even beat Yamnaya_Samara.

Ryan said...

@David - "My PCA shows that compared to the rest of Protoboleraz_LCA, I2788 has admixture from Steppe Eneolithic/EBA, rather than EHG. Note the trajectory of the shift from Protoboleraz_LCA to I2788."

The trajectory could just as easily be pointing towards the peak of the Europe to Siberia HG cline. I don't think you can distinguish between the two with just those two dimensions.

Ryan said...

@David - In terms of "steppe admixture" vs "steppe-like admixture" do you have a rough idea of the excess of EHG vs the excess of CHG?

I'd note that Mathieson has excess EHG >1% for the following populations:

Trypillia (14.6% WHG 2.4% EHG 83.1% AN)
Malak_Preslavets (13.4% WHG 4.6% EHG 82% AN)
Varna (7.0% WHG 9.2% EHG 83.8% AN)
Balkans_Chalcolithic (5.4% WHG 4.0% EHG 90.6%)
Minoan (-5.2% WHG, 4.9% EHG, 100.3% AN)

Can you distinguish between genuine steppe ancestry and ancestry from a place like Varna or even the Varna outlier? Sincere question - I'd be curious as to what degree you can separate the two - especially for small amounts of ancestry.

And on that front, could we maybe discuss what the hell was going on in Iberia?

In the WHG / EHG / AN model we get the following progression:

Iberia_EN - 8.9% WHG -1.6% EHG 92.7% AN
Iberia_MN - 25.4% WHG -1.1% EHG 75.6% AN
Iberia_Chalcolithic - 28.1% WHG 0.0% EHG 71.0% AN

In the WHG/CHG/AN model we get:

Iberia_EN - 8.0% WHG -3.2% CHG 95.2% AN
Iberia_MN - 23.4% WHG -12.3% CHG 101.0% AN
Iberia_Chalcolithic - 23.5% WHG -31.8% CHG 108.3% AN

So I realize these aren't great fits, andthose negative numbers are coming from something other than WHG/CHG/AN dragging Iberia MN and Iberia Chalcolithic away from CHG, and that something is likely El Miron. For the straight up WHG/AN model they include the following notes for Iberia_Chalcolithic and Iberia_MN because of the poor fit of WHG/AN alone:

* Iberia_Chalcolithic fits a model of 23.7% Vilabruna + 8.7% ElMiron +67.6% Anatolian Neolithic (p=0.29)

** Iberia_MN fits a model of 16.3% Vilabruna + 12.4% ElMiron + 71.4% Anatolian Neolithic (p=0.26)

So what exactly was going on here? Does it go something like this?

Iberia EN - mostly Anatolian Lithic with modest WHG admixture not necessarily from local hunter-gatherers, but also more eastern (Balkans?) hunter-gatherers.

Iberia MN - a metric crap-tonne of El-Miron-derived men come down out of the hills and take over the farming communitites.

Iberia Chalcolithic - a second metric crap tonne hunter-gatherer men come down out of the hills, but this time they weren't El-Miron derived and took over farming communities somewhere else and then migrated to Iberia?

capra internetensis said...

@Chris Davies

Thanks! Good to know.

Rob said...

@ Dave

"One of the Khvalynsk samples is very similar to Yamnaya.

It's the other two that aren't that are messing up the stats here, because they have too much EHG."

Sure, and that's sort of the point, which ive mentioned previously, if the signal is weak , or only formative in the 'homeland' of Yamnaya, how would it have an impact hundreds of miles away ?

"I can't think why Latvia_HG would be a useful proxy in this case,"

Evidently, despite explanation.
Boleraz appears to have begun near the north Carpathian arc- surrounded by areas we know to be home to foragers on that brink between EHG-WHG, which is why 3 years ago Motala was being picked up, and now Latvia_HG.
You seem to be under the false impression that every other region or group was standing still, and Yamnaya groups were the only ones moving about & admixing.
You risk getting unusable models despite being statistically fit and pleasing to your audience.



Davidski said...

@Ryan

The trajectory could just as easily be pointing towards the peak of the Europe to Siberia HG cline. I don't think you can distinguish between the two with just those two dimensions.

No, but that's why I also ran the stats.

I'd rather not comment about the other stuff. Seems too speculative at this stage.

Davidski said...

So Ryan, you reckon an AfontovaGora3-like population moved into the Carpathian Basin during the late Copper Age? :)

Well, anything's possible I suppose. But if we're sensible about things, then the most likely option is that there was a population like Yamnaya_Samara already expanding on and beyond the steppe at this time.

Davidski said...

Yes Rob, I'm pandering to the audience here, and don't care about the facts. It just so happens, tough, that my pandering lines up very nicely with all of the facts. Lucky break I guess.

Btw, post updated with the new finding!

Ryan said...

@David - "No, but that's why I also ran the stats."

Fair enough. How exhaustive were you in which alternative populations you tested if you don't mind me asking? Samara Eneolithic would make some sense. I'm just curious if something like Ukraine Neolithic or Ukraine Mesolithic are viable candidates. Basically did they bring CHG with them or not. Are the populations you posted the only ones tested?

"I'd rather not comment about the other stuff. Seems too speculative at this stage."

Speculation is the fun part! :3

I think it needs to be kept in mind at least when discussing the spread of both R1a and R1b (and I for that matter). This process of hunter-gatherers on the Pontic Steppe hijacking farming communities was happening in parallel across Europe it seems, so ignoring the Europe-wide changes may obscure some of the details of the local picture too.

"So Ryan, you reckon an AfontovaGora3-like population moved into the Carpathian Basin during the late Copper Age? :)"

No, I'm reckoning something more like EHG to be honest. They may not have had to have moved very far either if they were just on the other side of the Carpathian mountains or close to the Danube Delta.

"But if we're sensible about things, then the most likely option is that there was a population like Yamnaya_Samara already expanding on and beyond the steppe at this time."

Agreed, though "like Yamnaya_Samara" doesn't necessarily mean the same as Yamnaya_Samara in terms of location or language. It could be different or it could be the same. We don't know yet.

Ryan said...

@David - Keep in mind too that just before there was a exodus of migrants out of the steppe, there was an influx of immigrants into the steppe. Take a look at the change in ancestry between Ukraine_Mesolithic and Ukraine_Neolithic - it's pretty obvious there was a big influx of WHG ancestry into the Ukraine. Don't discount the possibility that these migrants had a lasting impact on Yamnaya's DNA pool.

I also think you're being a bit too focused on the steppe itself. It's possible that the people living near the steppe were always similar genetically similar to the people on the steppe. I wouldn't assume exclusivity there.

Davidski said...

@Ryan

But you have to admit that your propositions are very, very tenuous and realistically only serve to stir debate and get some replies, no?

For instance, how and why would an EHG-like population be expanding or even moving around during the Late Copper Age so as to influence the Protoboleraz population so unevenly?

This is precisely the time when pastoralism and mobility really begin to be a major feature of steppe populations. So why EHG and not anything like the R1a-M417 Eneolithic guy from Alexandria?

And it seems that you're not yet aware of all the recent outcomes, like for Globular Amphora from the forest steppe in Ukraine, which show that it has only noise levels of steppe input. So which other near to the steppe populations are you referring to? Surely not Protoboleraz, of which 3/4 samples show no steppe input whatsoever.

Matt said...

Off the current topic:

Still been looking at those Welzin samples and so created a datasheet using a K6 PCA Davidski's ran to split up Europe and ancients structure and maintain the split between present day East and West Europe (https://drive.google.com/file/d/0B8XSV9HEoqpFbFBQc2xGRWdleFU/view) and using a combination of Globe10 and West Eurasian PCA to generate 6 more dimensions to cover any differences in ancient components (HG, steppe, AF, CHG) which weren't covered clearly in the above.

Ideally this should result in fits which work both in the general West Eurasian context and in the specific recent European context.

So:
PAST3 Datasheet: https://pastebin.com/3ayrkCze
Populations FIle: https://pastebin.com/DB4eSE98
Ancient Calc File: https://pastebin.com/dF3qTCz5
Example nMonte runs: https://imgur.com/a/kA2YI

Seems to works pretty OK and pick the most geographically logical ancestors, while topping up with appropriate ancients as need be for extra broad scale deep ancestry as necessary (e.g. Polish take most of their ancestry from the Welzin and most HG Hungary BA, due to the Europe K6... but as in deep ancestry, this results in too much WHG and not enough steppe, they get extra steppe ancestry from Poltavka to balance it out).

(However fits work better for regions where there's lots of sampling in this set, mostly around the North Sea to Baltic. Quite a few more southern populations start picking up Mycenaean, Minoan, Hungary_IA likely to deal with southern BA-IA ancestry that's not well represented here...).

Samuel Andrews said...

@Rob,
"Yes i agree it has EHG, but that's not 'steppe' (CHG/EHG), is it ?"

Hungary HGs, Romania HGs, and Latvia HGs were all much more WHG than EHG. Even HGs from eastern Ukraine were about halfway between WHG and EHG. Local HG admixture or even HG admixture from a nearby region can be excluded as explanations.

Maybe the only possible non-Steppe cause would be EHGs moving from Russia into Hungary. But that really just sounds like special pleading.

Ryan said...

@David - "This is precisely the time when pastoralism and mobility really begin to be a major feature of steppe populations. So why EHG and not anything like the R1a-M417 Eneolithic guy from Alexandria?"

I think he's a great candidate. He lacks the CHG that is one of the hallmarks of steppe ancestry though. That's exactly the sort of person I'm talking about. Whatever he spoke it wasn't IE (maybe some sort of para-IE), and its linguistic descendants are probably all dead. It may still have had ah impact in Central Europe before later cultures supplanted it.

Other populations that fit the bill would be Latvia_LN, Ukraine_Mesolithic, Latvia_MN, Latvia_HG, Romania_HG or a more EHG rich relative of Koros_HG and Iron_Gates_HG. Or just an unsampled population - like whoever the hell was living in the Polonynas.

The Bronze Age Makó/Kosihy-Caka culture sample from Olalde has EHG but not CHG in their analysis. Do you dispute this? Because if not, then it's pretty obvious that that EHG had to have come from a population that didn't have CHG with it. And if it happened once, maybe it happened twice, and that is worth checking. Hence why I'm asking if a population with less CHG than Samara would be a better fit.

"But you have to admit that your propositions are very, very tenuous and realistically only serve to stir debate and get some replies, no?"

Which part? I agree that some parts are very speculative but I'd like to think I qualify my statements accordingly? Not sure about what "get some replies" means but I think debate and trying to push the boundaries of what we can infer from existing data is pretty interesting and worthwhile.

And if these questions about the spread of WHG in EHG can't be answered yet, then we don't fully understand the dynamics of Eneolithic and early Bronze Age Europe.

"For instance, how and why would an EHG-like population be expanding or even moving around during the Late Copper Age so as to influence the Protoboleraz population so unevenly?"

I don't know - I'm not saying why, just that maybe it happened. The only sample we have from the Makó/Kosihy-Caka culture is in Olalde 2017, and he has the sample as having a lot of EHG but not the CHG you would expect from Yamnaya. That at the very least isn't tenuous, and is something I would challenge you to explain otherwise.

"And it seems that you're not yet aware of all the recent outcomes, like for Globular Amphora from the forest steppe in Ukraine, which show that it has only noise levels of steppe input."

I haven't read up on the Globular Amphora results but this doesn't surprise me. Look at this map of the language families of North America:

https://upload.wikimedia.org/wikipedia/commons/a/a5/Langs_N.Amer.png

Does that seem nice and ordered to you? Because it looks like a Rorschach blot to me. There are 5 continent-spanning expansions on that map, all more recent than the last 3,000 years, and all but two happening without agriculture.

I don't think Mesolithic/Neolithic Europe was any more orderly. I think it was rather similarly highly structured but in a very patchy and haphazard way. Does that make sense?

If we only had the amount information about Chalcolithic Europe to go by to figure out what pre-contact North America was like, I think we'd probably have the broad strokes right, but a lot of the fine detail we wouldn't have a clue. So I wouldn't expect everything in Europe to fit in nice boxes tied with bows, and I think you shouldn't either. If it was simple what would be the fun in that anyways?

And it just so happens that some of these weird edge cases in Europe happen to be in areas that may have played a big roll in shaping the future of the continent.

Ryan said...

@Samuel - "Hungary HGs, Romania HGs, and Latvia HGs were all much more WHG than EHG. Even HGs from eastern Ukraine were about halfway between WHG and EHG. Local HG admixture or even HG admixture from a nearby region can be excluded as explanations. "

IF we accept Mathieson's numbers, Romania HG was about 50/50 (55.9% WHG vs 44.1% EHG), and I don't see how you can assume that's the peak of EHG ancestry off the steppe.

And there are samples in the Ukraine that are EHG but without the characteristic steppe ancestry of EHG+CHG that could fit the bill.

Rob said...

@ sam

"Maybe the only possible non-Steppe cause would be EHGs moving from Russia into Hungary. But that really just sounds like special pleading."

No it's not, we know this is the case
And I'm not saying that it was actual hunter-gatherers mixing, but Western " sredny stog" & Cernavoda- actual semi pastoral cultures from same period as Boleraz in & around the steppe. They'd be somewhat different to later Yamnaya on the EHG-> WHG , CHG-> ANF path

Pretty sure thisll be the case; but whatevs its just an interesting side point

Rob said...

"like for Globular Amphora from the forest steppe in Ukraine, which show that it has only noise levels of steppe input"

That's because GAC originated in north central Poland and then moved southeast toward the steppe
So it behaves exactly as expected.

Davidski said...

Right, so that crosses out Poland as a potential source of anything relevant here, doesn't it?

Davidski said...

@Ryan

For whatever reason, it looks like you're focusing on the least plausible possibilities not backed by multiple lines of evidence, or even any sort of strong evidence. Perhaps you work that way, by first striking off the least likely scenarios? Or perhaps you're just attracted to more complex and surprising solutions, I don't know?

But clearly, the best you can say is that maybe it happened that way you'd like it to have happened. Maybe, but chances are slim when multiple lines of evidence are pointing in the other direction.

Also, you sometimes base your theories on a lack of knowledge of new data, or even a lack of a complete understanding of old data. For instance, you're claiming now that Hungary_BA might not have any CHG. But I can assure you that it does, and in fact it has a lot of the classic Yamnaya-like steppe admixture. This has been shown multiple times not only by me, but also in really high level scientific literature with formal stats. So focusing on one shaky ADMIXTURE run in one paper isn't a very useful.

Ryan said...

@David - "For whatever reason, it looks like you're focusing on the least plausible possibilities not backed by multiple lines of evidence, or even any sort of strong evidence. Perhaps you work that way, by first striking off the least likely scenarios? Or perhaps you're just attracted to more complex and surprising solutions, I don't know?"

The simplest solution is usually the correct one. It's not always the correct one though. I'm just saying keep an open mind. It's harder to correct an old assumption than it is to avoid making a new one.

"But clearly, the best you can say is that maybe it happened that way you'd like it to have happened. Maybe, but chances are slim when multiple lines of evidence are pointing in the other direction."

I don't have any particular stake in para-IE groups dispersing before the main kurgan migration. It seems that you and I agree on the situation so... isn't mi casa su casa here?

"Also, you sometimes base your theories on a lack of knowledge of new data, or even a lack of a complete understanding of old data."

Well since you already seem to have the same theory as me as to a plausible source of this ancestry it seems my theory is your theory.

But I'll freely admit that staying by my mother's side during her long and unsuccessful fight with leukemia has got me behind in my academic reading. How inconsiderate of me.

"For instance, you're claiming now that Hungary_BA might not have any CHG. But I can assure you that it does, and in fact it has a lot of the classic Yamnaya-like steppe admixture."

Could you just post a link to that then? I can't even find the damned original admixture run.

Though to be clear I'm not saying 0 CHG, just that the CHG can be explained from the Anatolia Neolithic ancestry.

Rob said...

That paper Sladen linked very interesting - on Iberian mtDNA transect
A lot of H1 & H3 there

Samuel Andrews said...

Rob, almost all data was released last year. I know it inside and out. Also, no N1a1a, which is really significant considering how frequent it was in Barcin and LBK and Hungary farmers. Also, popular K clades across most of Europe today like K1a4a1, K1b1a1, and K1a1b1 have popped up in Iberian farmers but not central European farmers.

Outside of Autria, Croatia, and the Baltic states I see no significant amounts of eastern farmer derived N1a1a in Europe. I think they went the way of the do do bird which is why I get frustrated that they are assumed to be the only farmers Yamnaya and Corded Ware mixed with in northern Europe.

Samuel Andrews said...

Or more accurately it frustrates me that LBK/Hungar farmers are seen as the de facto Neolithic farmers. Their U8b1, N1a1a, T2c1, is rarelly ever seen today.

I don't know how else to explain that away other than to say it's evidence other farmer groups who had beeen seperate maybe before the migration to Europe, contributed bigger amounts of ancestry to modern Europeans. One attractive candidate are the Atlantic (Iberia, Britain) farmers.

Nirjhar007 said...

https://www.sciencenews.org/article/how-asian-nomadic-herders-built-new-bronze-age-cultures

supernord said...

@Ryan "IF we accept Mathieson's numbers, Romania HG was about 50/50 (55.9% WHG vs 44.1% EHG)"

You look at supervised version, but it is incorrectly. There need look unsupervised version where Romania HG is ~70% WHG vs ~30% EHG.

Davidski said...

https://www.sciencenews.org/article/how-asian-nomadic-herders-built-new-bronze-age-cultures

Thanks Nirjhar, I left a comment under this strange article, and so did Capra.

Nirjhar007 said...

Where did practice of Herding emerge first?.

Davidski said...

Where did practice of Herding emerge first?

Probably the northern Levant.

But are you going to claim as a result that Yamnaya was located in the Levant rather than Eastern Europe?

That would be somewhat strange. Sort of like saying that India is part of the British Isles because English is widely spoken there.

Nirjhar007 said...

No. I am just pointing ,that the practice of herding came to Europe from Asia . The continous advent of CHG/Iran ancestry (Perhaps EEF also in some parts of Europe? ) is likely related to it , so technically herders did come from what is now called 'Asia' to what is now called 'Europe'.

Davidski said...

But technically the article is wrong, because it claims that Yamnaya herders expanded from Asia, and technically the Yamnaya horizon, every bit of it, was located in what is now technically considered to be Europe.

It's all very technical, to be sure, but not overly complicated.

Grey said...

(this is just an attempt at squaring the circle of various comments above - it may not fit some data or other but anyway...)

"For instance, how and why would an EHG-like population be expanding or even moving around during the Late Copper Age so as to influence the Protoboleraz population so unevenly?"

The problem with labeling post-mixture EHG+CHG as "steppe" is it creates the impression that the pre-mixture steppe (EHG) was at the pre-mobile (i.e. pre-horse) HG stage but maybe it wasn't - maybe there was an overlap where a post-horse but EHG only "steppe" existed.

And if there was a stage where the yamnaya "steppe" was strong enough to physically force their way onto other people's territory then there was probably an earlier stage when they weren't strong enough: say post-horse but pre-cavalry.

Putting those two together: 1) post-horse steppe population before the CHG mixture and 2) post-horse but pre-cavalry so no dramatic military advantage yet with 3) "influence the Protoboleraz population so unevenly" then one possible model might be traders - and what might a post-horse but pre-cavalry population be trading off the steppe - horses seems like the simplest answer.

So the model is an early expansion phase of wagon-steppe (mostly EHG) horse traders creating pocket settlements on good pasture along trade routes from the steppe before the development of the cavalry-steppe (EHG+CHG) population who expanded later.

Nirjhar007 said...

But technically the article is wrong, because it claims that Yamnaya herders expanded from Asia, and technically the Yamnaya horizon, every bit of it, was located in what is now technically considered to be Europe.

It's all very technical, to be sure, but not overly complicated.


Well the herders came to Europe from Asia with the Iran/CHG related ancestry . So Herders are Asian, who taught the Yamnayans and Proto-Yamnayans and others to herd . Plain and simple it is. This fact will remain unchanged , no matter what.

Grey said...

Samuel Andrews said...
"I don't know how else to explain that away other than to say it's evidence other farmer groups who had been separate maybe before the migration to Europe, contributed bigger amounts of ancestry to modern Europeans. One attractive candidate are the Atlantic (Iberia, Britain) farmers."

If the Atlantic coast was relatively under populated with farmers due to the climate being relatively unsuitable for early neolithic crops (but great for dairying) and then some dairy herders arrived I can imagine the possibility of the original farmer + dairy herder population having a dramatic population boom as they filled the available space (and maybe expanding a little inland to the east as a result). So if the original farmers along the Atlantic coast were a bit different to the LBK farmers and they expanded with the dairy herders would that model fit?

Davidski said...

@Nirjhar

Maybe, but the Yamnaya horizon was located entirely in what is now Eastern Europe. So the article is wrong by calling Yamnaya Asian, since Yamnaya wasn't located in Asia nor did it develop in Asia.

The author Bruce Bower also thinks that the steppe is hilly. LOL

Nirjhar007 said...

In that case of course, it should me considered, that people who brought CHG/Iran did contribute largely in the formation of Proto-Yamnaya and Yamnaya . Contribution was both in case of lifestyle and also in case of culture/language . So the development of Yamnaya had considerable Asian origin. With the contribution, there were certain migrations which can be suggested with robust data of aDNA .


Davidski said...

But Bruce Bower isn't talking about Neolithic farmers from the Caucasus who entered the "hilly" Pontic-Caspian steppe.

He's talking about the Yamnaya people and calling them Asians who migrated to Europe, even though the Yamnaya horizon was located in Eastern Europe and so they originally lived in Eastern Europe.

Nirjhar007 said...

Its a 50/50 thing , Yamnayans were of course also Asians :) .

Davidski said...

Yamnayans were not Asians, just like I'm not an Asian.

I would never describe myself as an Asian for very good, practical reasons, although I suppose I could call myself Eurasian in some circumstances.

Salden said...

Yamnayans and similiar groups cluster nearest to modern Europeans. Not East Asians, Central Asians, etc.

Ryan said...

@Grey - "The problem with labeling post-mixture EHG+CHG as "steppe" is it creates the impression that the pre-mixture steppe (EHG) was at the pre-mobile (i.e. pre-horse) HG stage but maybe it wasn't - maybe there was an overlap where a post-horse but EHG only "steppe" existed."

Thanks for stating it that way. That's more or less what I was trying to get across. Proto-IE has been influenced a great deal by Caucasian languages, so that means by definition any groups from the steppe that pre-date this heavy interaction would not be Indo-European, but rather Pre-IE or Para-IE. I'm just saying we should keep an open mind to the possibility.

@David - I think going with the older more western boundary to Asia of the Don may be appropriate here, as it emphasizes that this isn't just some older wave of western European colonizers going out into Asia, but rather that these people were intrusive to the vast majority of Europe too.

Davidski said...

@Ryan

And what if the steppe peoples who cashed into Europe came from west of the Don? Should we move the border some more for one article?

Ryan said...

@David - I'm just saying it's not the worst bent truth to have.

a said...

>United Nations geoscheme for Europe
>Eastern Europe
>Geographical
>"The Ural Mountains, Ural River, and the Caucasus Mountains are the geographical land border of the eastern edge of Europe."

Mike the Jedi said...

You're being nicer about that "Asian herders" article than I would've been, Dave. It's just flat-out wrong and reeks of sensationalism. "Hey, if I throw 'Asian' in the title, I'll get more readers!"

Yamnaya isn't Afanasievo; it isn't even borderline a la Sintashta. It's clearly well within Europe so that writer is either incompetent or just dishonest. And the article isn't talking about CHG, Nirjhar, it's talking about Yamnaya, which had its ethnogenesis on the Pontic-Caspian steppe, i.e. IN EUROPE. Ashkenazi and Sephardi Jews also had their genesis in Europe; the fact that half their ancestry is Levantine in origin is immaterial to where the groups themselves actually formed. It's not that hard to figure out.

As for the current borders of Europe: hey, they may only date to the 19th century, but they've been the standard for long enough and make the most sense anyway. Why would a cartographer use rivers like the Don and Volga to demarcate the Asian border when expansive mountain ranges like the Urals and Caucasus are so close by? Only the Ural River is needed to mark the southeastern-most border but it fits the "Ural" theme nicely, doesn't it?

The Ural area is a convenient "genetic" boundary, too. You don't find many modern populations west of the region with more than 25% East Asian ancestry (only ones I can think of are Kalmyks and some Nenets in the extreme north). Pretty much every native group just east of the range is well over that. Same for most of Turkestan. Bashkirs largely live west of the mountains, but even then only barely. So, yeah, the Urals make for a very nice boundary separating Eastern European people from Western Siberians. Its usefulness shouldn't need much more defense than that.

You want a real challenge? Try figuring out which Aegean isles should be considered part of Europe and which should be considered part of Asia Minor. Good luck! LOL

Davidski said...

If things are slow next week I'll write a blog entry explaining that the Yamnaya horizon was located in what is now Eastern Europe.

And I'll refer to Bruce as Bruce "hilly steppe" Bower.

Anthro Survey said...

Re/Yamnayans, Asians, Europe-----

It is unwise to refer to Yamnaya-like individuals as either "Asians" or "Europeans" because to do so is to back-project modern-day genetic variation and the geographic boundaries that coincide with it into the world of 4000BC.

First off, the Yamnayans essentially lacked ENA admixture----apart from that archaic "ENA" present in ANEs due to lineage mixing shortly after WestEurasian-ENA split as Qiaomei Fu explored.

The borders of Europe that are relevant today simply weren't back then. It was Dnieper, not Volga, that was more significant in delineating clines: EEFs from early Steppe. There was no European cline as we know it back then and the clines that did exist in Western Eurasia are not so relevant today.

After 2700BC or so, WHG-EEFs and steppe began to hybridize in a process that took several stages and intermediates to come to completion, resulting in a Euro continuum we're familiar with. In short, the territory encompassing modern-day Poland, Germany, Czechia, Hungary as well as parts of Belarus and Ukraine pioneered in this. There were then subsequent expansions into what is today considered Western, southern and easternmost Europe from this "MLBA horizon" over next 3 millenia.

Slavs, presumably, were one of the later MLBA-like waves who expanded, among other places(like Greece), into modern-day Russia. It was after this time that Volga came to have the significance that it has today(all this coupled with important medieval cultural diffusions reinforcing the genetic cline and the concept of Europe).

Davidski said...

Nah, the cline that we see on the PCA from Steppe_EMBA to Europe_MN already existed on the steppe during the Eneolithic, except it covered a fairly short distance, from the Caspian steppe to the western end of the North Pontic steppe.

At the eastern end there was Khvalynsk, looking like Steppe_EMBA, but usually with more EHG, in the middle there were people like the R1a-M417 guy from Alexandria, with a lot of EEF, and at the western end there were typical farmers, but usually with excess WHG compared to EN farmers.

Anthro Survey said...

But, yes, it's looking like the first Yamna-like populations came into being on the territory associated with the present-day European genetic cline around eastern Ukraine or so.

Anthro Survey said...

The middle guy with a lot of EEF hailed from the Dnieper region, a contact zone at the time it seems. (The other one from Eneolithic Ukraine was mainly just steppe-like.)

While guys like that existed, they were rather peculiar for the time---perhaps smth like modern Tuaregs. The population density in that region of the cline that exists today and constitutes much Europe's heartland simply wasn't there. Fast forward a few millenia and the situation has reversed considerably.

Aram said...

Anthro Survey

Imho Shulaveri will have alot off ANF.
The source of CHG for Yamna is from lowland Azerbaijan who entered steppe via Dagestan.
The extra CHG they got probably when they crossed Dagestan.

Rob said...

On the topic of Yamnaya formation, and whilst discussed ad nauseum, things are anyway slow.

With the availability of new data from Armenia Chalcolithic and Iran, the best match for Yamnaya is archaic CHG, IIRC. The problem is it is difficult to demonstrate where such a population would have lived north of the Caucasus before 4500 BC, because there;s little to find.
My current hypothesis therefore is to link the appearance of Meshoko-Zamok horizon (roughly same territory as Majkop, but preceding it by hundreds of years - 4500 vs 3700 BC) with the Darkveti culture in Georgia. The latter is a baroque highland agro-pastoral formation, probably emerging from Neolithicizing Kotias-type natives.
I think that much is a fair assumption.

The problem is with Majkop being a rather different entity - different settlement & material culture, which succeeds but is partly contemporaneous with Meshoko. This would have added Suvorovo-Skelya elements, as well as wide-ranging contacts from Uruk to south central Asia mediated at an individual level.

Yet another problem presents itself with the lower Don Neolithic. Unlike the the rest of Russia/ east Ukraine, here the 'Neolithic' actually is a Neolithic - with domesticates and fixed dwellings which several scholars have drawn analogies to eastern Anatolia, as if by direct migration across the Black Sea, because such sites are lacking in between, in the Caucasus. In fact, O.M.'s Shuvaleri people appear rather late in the South Caucasus and fully formed, c. 6000 BC (i.e. after sites such as R. Yar on the Don). One would imagine these E. Anatolians would be something proto-Tepecik like ANF/ CHG), and given the fair presumption that the lower Don culture in turn catalysed a productive economy further east, in the lower Volga/ Caspian, the lack of ANF there is a bit surprising (but we can simply appeal to down-the-line exchange).

As a final trajectory of 'southern' input, there is the central Asian Kelteminar-like people impacting the north Caspian area (certain flint repertoires & pre-ELshanka pottery).

Rob said...

@ Aram

"The source of CHG for Yamna is from lowland Azerbaijan who entered steppe via Dagestan.
The extra CHG they got probably when they crossed Dagestan."

I see no evidence for that.
As I mentioend, i think the CHG came from western Caucasus (Georgia-Abkhazia).
The Dagestan plain, from what I read, earliest evidence is c. 4300 BC, somewhat later than Meshoko. This region seems to have then become a secondary mediator between north (Majkop) and south (Uruk)

Rob said...

At some point (C. 3200 BC0 there seems to have been another migration or formation at NW Caucasus, Dolmen people, which eventually took over former Majkop territory. The are probably proto-NWC speakers.

Grey said...

Ryan
"any groups from the steppe that pre-date this heavy interaction would not be Indo-European, but rather Pre-IE or Para-IE."

I guess. Is it known whether IE was originally based on an EHG language modified by CHG or a CHG language modified by EHG? If the former then any mainly EHG post-horse but pre-cavalry "steppe" population might have lots of similarities to PIE and if the latter, less so. Are there historical examples of this happening elsewhere which might provide a clue to which way round it was through certain words that don't change?

For example if you have a population speaking language A and a subset of that population moves away and then later the original source population's language changes from A to AB due to mixing with population B do things like personal names change?

Anthro Survey said...

@Aram
At any rate, if Azerbaijan was a source, it would still have to predate 6000BC since ANF would have penetrated into those lowlands at roughly the same time (we think) it impacted the Colchis. I see no impediment for it not to have done so.

@Rob
Could there have been two independent formations of steppe-like ancestry?: One around the Azov(stemming from hypothetical CHG-heavy folks in the Don Delta established early on) and one around the Astrakhan delta or thereabout? The latter might well have had Kelteminaric influences as source of CHG-like if there is little evidence of an influx from Dagestan.

The material evidence from lower Don neolithic is quite intriguing, but yeah, de-facto cultural diffusion was probably the culprit.

How about Crimea? Anything notable taking place there?

Olympus Mons said...

@Anthro Survey
Early Shulaveri will be a mix of Iron Gates HG (+) and some Barcin (-). They picked up Barcin while in places like Fikirtepe (7000-6000bc) but kept moving on the northern part of the dense forest that divided Anatolia (south of Forest heavy Barcin), sosouth shores of black sea. they shared the same story of the stock of people that lived in Ovcarovo Gorata (Bulgaria), Fikirtepe, then the ones that showed at Hagoshrim (north Israel). The Shualveri are the ones that picked up barcin (a little) and later went to South Caucasus known as Shulavari shomu - Their level of Pastoral (cattle, not just sheep or pigs) was always different than the surroundings.

So EARLY Shulaveri will be WHG+EHG (iron gates) and Barcin (fikirtepe). Maybe better represented by the current DNA from SHulaveri we have, which are the one south of lesser Caucasus, hence the Aratashen/Araknashen kind of SSC. Later (1200 years later) will be a bit heavier on Barcin and with a fairly good proportion of CHG/Iran.
I Agree with Rob (I think) that the ones we are talking here, were the ones that lived in Georgia, in places like Arukhlo, Mentesh tepe, etc, (the biggest part of SSC) that when “the Others Came”, whomever they were that made them flee, some flee to nearby mountains (hence Bagvalins) but most run to east shores of Black sea, up to Svobodnoe and Mesokho and later north. These had a much bigger component of CHG than the ones south of the lesser Caucasus.

So,
Early Shulaveri = WHG+EHG+EEF (iron gates + Barcin)
Later (or Georgian) Shulaveri = WHG+EHG+EEF+ CHG/IRAN (a good chunk).

Davidski said...

Useful tip: whatever Olympus Mons claims, it's very likely that the opposite will be true.

Olympus Mons said...

@Davidski.
Classy as usual.

Olympus Mons said...

@@Davidski.,
... and which part of it you disagree?
(ie - Which part contradicts Steppe and proto-Poland/CWC as the center of the universe?)

Davidski said...

You're kidding yourself if you think Shulaveri will show anything from the Balkans and any real European hunter-gatherer admixture, except maybe some CHG-derived faux EHG.

It'll just be a normal southern Caucasus/eastern Anatolian Neolithic population, similar to Tepecik but with more CHG.

Olympus Mons said...

@Davidski. Go play with stats playSation.

Shulaveri will be very close to bulgaria, Romania, serbia Neolithic... Because that is where they obviously came from.

If you don't know that....

Basil S said...

@Samuel Andrews

The new Lipson had LBK as 96% EEF, with 4% WHG, so prob the EEF with least admixture but from then until LN Europe is about 2000 years so plenty of time for EEFs to accumulate extra drift to become the EEF that Yamnaya/Corded/BB mixed into. Obviously each part of Europe will be different but certainly the Olalde Bell Beaker paper had BB outside Iberia taking most of their EEF from Globular Amphora and Sweden TRB and by the time BB got to Britain, they had on average only 3% local EEF input from what I would assume to be 'Iberian' like farmers or at least with some Iberian input but yeah those LBK haps you described aren't exactly common anymore.

On the subject of Sweden TRB and local EEF input into British Beakers, the Cassidy paper with Rathlin had Sweden TRB also sharing a signal with Ballynahatty, but as you'll probably know, that EEF didn't contribute too much to Rathlin.

Chad Rohlfsen said...

You don't have a clue. We already have samples that fall at the end of Shulaveri-Shomu from Armenia, Pre-KA. Nothing like Bulgarians and Serbians. You seriously have never been right about anything. Even when data is available you still manage to get it wrong.

Samuel Andrews said...

@Basil S,
"The new Lipson had LBK as 96% EEF, with 4% WHG, so prob the EEF with least admixture but from then until LN Europe is about 2000 years so plenty of time for EEFs to accumulate extra drift to become the EEF that Yamnaya/Corded/BB mixed into"

What I'm thinking is the biggest EEF contributor to Corded Ware and BBC didn't even descend from LBK. TRB and Globular Amphora may not have been descended completely from LBK and its relatives.

Matt said...

Slightly off topic post, could be of interest to some:

I was thinking again about the polygenic height scores from the Berg et al paper - https://www.biorxiv.org/content/early/2017/11/06/167551 - and how they could be used to predict a first approximation of ancient heights.

Results of this, and some estimated ancient population height scores, in this imgur gallery: https://imgur.com/a/4Aayj

For method, first I duplicated the Figure S12 from Berg et al, giving present day populations genetic (male) height and waist scores in cm (Images 1+2 in my link).

Then using the scores and populations I matched these to Global10 dimensions, and used Past3 to generate regression equations on Global10 scores, to predict the height and waist scores from Berg.

Validating this against the real scores in Berg (Image 3 in my link) showed that the Global10 dimensions can pretty much almost perfectly reproduce the scores from Berg et al for height, while waist scores are messier.

(If I had to interpret this, it would be that the deep ancestry Global10 captures is a pretty good first approximation for the forces selecting height, and that relatively little selection has happened outside these dimensions... but predicted is worse for waist, which is less related to ancestry as captured by Global10 and more to selection outside it.

Interestingly, both the Saami and Sardinians are outliers who have much lower height scores in Berg's data than in the Global10 prediction. They have also in the past suggested to both have been under selection for shorter height. Conversely, Icelanders, and to a lesser extent Western and Central Europeans seem like an outlier in terms of having slightly taller height in Berg's data than predicted.)

If anyone wants these regression equations I can upload to pastebin btw.

Now, feeding ancient Global10 scores back in, then should finally provide an approximation of ancient heights.

Sample results show that, as a first approximation, Euro HG would be tall (should have been slightly taller than present day Europeans under similar conditions) at around 183-181 cm, as expected, and that WHG may have been very slightly taller than EHG. Yamnaya show comparable height to present day Northern Europeans (as in Berg) at 180cm, while MNChl in Iberia and Germany are slightly smaller at 178cm, and Barcin_N at 176.5cm. Iran_N and Levant_N are predicted to be roughly 174cm, Natufian 173cm.

(All subject to these genetic height scores being less accurate the less related to Europeans a population is... e.g. Maasai, Biaka are extremes).

Matt said...

Btw, re: the last post, I also quickly just repeated the same methodology, but took out Global10 dimension 6, since I remember Davidski had concerns this could inflate similarity between recent Europeans and WHG.

No change to the predicted scores via using either a regression on Berg's height scores and full Global10 or Global minus dimension 6 (https://imgur.com/a/Qsa7H).

Olympus Mons said...

you are definitely a mental case.

"we already have Shulaveri..." -- you really live in lala land don't you, beta?

Chad Rohlfsen said...

Way to edit my line. Armenian samples between 4300-4000BCE. That's the tail end of S-S timeframe. Not mental, Mr African migration of R1b lololol.

Ric Hern said...

Did Globular Amphora and TRB adopt Iberian Bell Beaker Culture ? If not then I wonder why only Steppe Nomads adopted some aspects of the Bell Beaker Culture ?

Chad Rohlfsen said...

Dating might be an issue.. still needs to be resolved.

Ric Hern said...

Yes Chad. It is just strange that I do not see any Bell Beakers in Central Europe without a R1b connection. Or did I miss something ?

Aram said...

Rob

"" At some point (C. 3200 BC0 there seems to have been another migration or formation at NW Caucasus, Dolmen people, which eventually took over former Majkop territory. The are probably proto-NWC speakers. ""

Agree. Genetic data favours this. NWC people's Y DNA expanded into that region after 3000 BC. There is also increasing consensus among Russian archaeologists that they came after Maykop with the apparition of that Dolmens. Their ultimate origin could be the Wester Georgia lowlands Neolithic farmers of Odishi Anasueli type.

Concerning CHG in Yamna. I don't rule out Meshoko type people from NW Caucasus playing a role into formation of Yamna component. But then how to explain that it is the Yamna _ Kalmykia that has higher CHG/EHG ratio than those from Kuban region?

Davidski said...

There aren't any Yamnaya samples from the Kuban region available yet.

Rob said...

Ah yes the CHG is slightly higher in Kalmykia than Samara, right ?
So that sort of answers Anthro's question

Olympus Mons said...

@Chad.
You are talking about people that dressed,eat, made pottery, made houses ,lived in diferent settlements in the same geography.

Different architecture, diferent way of live in the same place means diferent people dumb ass.

Stop being such a beta for Davidski.

Olympus Mons said...

.... The only Shulaveri dna we have is in aratashen, arknashen which yield mtdna I1, H2, H15. And since are early SS its good. However the shulaveri were actually in georgia and Azerbaijan. Not really armenia!!

Armenia CHL mentioned are Y dna L1A most likely from south Iran so I figure not very representative of the armenia (and much less geirgia) Adna.

Chad Rohlfsen said...

Wow.. you are something. First of all.. Aratashen and Arknashen are in Armenia. The Armenia Chl samples are just SE of there and across from Lake Sevan of other sites from S-S. There are lots of similarities and they are likely related.

Secondly, you once again demonstrate your ignorance of aDNA. Tepecik-Ciftlik is already getting similar to Armenia CA, but between 7500-5500BCE. So, even Central Anatolia is very different from Balkan farmers from the outset of farming in Europe. How in the hell would Bulgarian and Serbian farmers be implicated in S-S? Halaf is generally pointed to.

Seriously, you have no knowledge of geography, history, or aDNA. Give it a rest already.

Rob said...

OM, do you have any links for Balkan -SS connnections ?

Matt said...

Off topic again: As a followup to my above post, on using a regression between Berg et al 2017's genetic height predictions and the Global10 scores to predict ancient height, I thought I'd test another way, using male heights from Grasgruber 2016 (https://www.ncbi.nlm.nih.gov/pubmed/26948573 - "Major correlates of male height: A study of 105 countries") and Global10.

Crossover was not too bad between Grasgruber 2016 and Global10, with 59 populations in common, and I added 8 of my own best guess values for populations outside of the large swathe of Eurasia studied in that paper, mainly African (Yoruba, Biaka, Ju_hoan_North, Masai, Ethiopian_Jew, Nganasan, Australian_Bougainville, BedouinB). None of the supplemented values should have had any impact on the West Eurasian regressions, particularly in Europe.

Results: https://imgur.com/a/LyDzR

Final results are non linear for much of Africa and to a lesser extent East Eurasia, however there's essentially a 1:1 correlation between regressions based on Berg's genetic height and Grasgruber's observed height for West Eurasia (albeit with a slightly higher slope in regression on Grasgruber). Both seems to agree that:

a) WHG slightly taller than EHG, with larger magnitude in Grasgruber, and that Euro HG would likely have had male genetic/observed height above 181 cm.

b) Yamnaya would have had genetic/observed height around 177-179cm. Data differ in that regression on Grasgruber's observed heights place Yamnaya as slightly shorter than Europe MN, while regression on Berg's genetic heights places them as slightly taller than Europe MN. This may be because observed heights in Grasgruber have small bias towards more well nourished West European populations? But in any case, the overall effect only leads to a 2cm (0.8 inches) difference in estimates.

Both regression based on Grasgruber observed data and Berg's genetic heights place EuropeMN at male height of 178cm (5'10"), while direct Berg's direct genetic score would place at 176cm (5'9").

c) Barcin_N and Levant_N are also roughly identical between both, however CHG and Iran_N has slightly larger difference of 5cm - 7cm shorter height based on Grasgruber's observed heights.

Overall, I'd say repeating the same exercise confirms to me that it's likely that the genetic height estimates for West Eurasians in Berg 2017 and the regressions I've based on them are probably mostly accurate and that the ranking of HG>>Yamnaya=>EuropeMN>Barcin>LevantN/CHG/Iran, and the rough genetic heights are fairly likely. Berg's height scores are probably not underestimating genetic height differences between ancient ancestors to Europeans by very much at all (within fractions of centimeters) and even still not very much for CHG and IranN.

(maybe of interest to no one but as I did it for my own interest, thought I'd share!)

Guy Tipton said...

Hi Matt,

I at least found it interesting. Thanx for running those regressions.

Cheers, Guy

«Oldest ‹Older   1 – 200 of 234   Newer› Newest»