search this blog

Monday, October 19, 2015

Reconstructing the genetic history of Siberia and Northeastern Europe


A new preprint on the genetic history of North Eurasia has just appeared at bioRxiv. Here's the abstract:

Siberia and Western Russia are home to over 40 culturally and linguistically diverse indigenous ethnic groups. Yet, genetic variation of peoples from this region is largely uncharacterized. We present whole-genome sequencing data from 28 individuals belonging to 14 distinct indigenous populations from that region. We combine these datasets with additional 32 modern-day and 15 ancient human genomes to build and compare autosomal, Y-DNA and mtDNA trees. Our results provide new links between modern and ancient inhabitants of Eurasia. Siberians share 38% of ancestry with descendants of the 45,000-year-old Ust-Ishim people, who were previously believed to have no modern-day descendants. Western Siberians trace 57% of their ancestry to the Ancient North Eurasians, represented by the 24,000-year-old Siberian Malta boy. In addition, Siberians admixtures are present in lineages represented by Eastern European hunter-gatherers from Samara, Karelia, Hungary and Sweden (from 8,000-6,600 years ago), as well as Yamnaya culture people (5,300-4,700 years ago) and modern-day northeastern Europeans. These results provide new evidence of ancient gene flow from Siberia into Europe.

My initial thought was that the MA1 or Mal'ta boy-related admixture estimate of 57% for Western Siberians (in fact, Mansis) was way too high. However, I checked it using qpAdm and the Mansis from the Human Origins dataset, and apparently it does make sense as a three-way model including MA1, Han Chinese and Georgians. The full qpAdm output is available here.

Mansi
Han 0.367
MA1 0.561
Georgian 0.073

chisq 0.278, tail prob 0.870155

Curiously, I get a very similar result with Eastern European hunter-gatherers (EHG) in place of MA1, and the fit is almost as good. The full output is here.

Mansi
Han 0.402
EHG 0.537
Georgian 0.061

chisq 0.916, tail prob 0.632459

In any case, we can safely assume that Mansis harbor a lot of MA1/EHG-related ancestry. Perhaps as much as ~60%. However, I have to say that this quote from the paper makes no sense whatsoever:

Our findings also point to Western Siberians Mansi as a likely source of the ANE [aka MA1-related] ancestry among northeastern Europeans.

Really? Mansis? Not even proto-Mansis, or some sort of Mansi-related population?

Also, as far as I can see, the authors consider Y-haplogroup N1c1 as an EHG paternal marker, simply because they dated its main expansion to 7,100-4,900 BP based on present-day samples. Please note that Karelia_HG and Samara_HG are classified as EHG.

The Western Siberian admixture into the Eastern Europeans likely began before the Yamnaya culture period (5.3-4.7 kya), since the admixtures with Mansi are also very strong among hunter gatherers from Northeastern Europe from 6.6-8 kya (Karelia HG, Samara HG and to lesser degree Motala HG and Hungary Gamba HG; Fig. S21f-q) that predated the Yamnaya people. Therefore Western Siberian admixtures into northeastern Europe likely began prior to 6,600 years ago, coinciding with the expansion of Y-DNA haplogroup N1c1 among Siberians and northeastern Europeans (7,100-4,900 years ago). Since haplogroup N likely originates in Asia or Siberia, its presence among eastern Europeans likely reflects ancient gene flows from Siberia into Eastern Europe.

The problem is that this isn't yet supported by any direct evidence from ancient DNA. Thus far, we know that EHG carried Y-haplogroups J, R1a and R1b, but no N1c1. Later populations, with significant EHG ancestry, such as Corded Ware, Khvalynsk and Yamnaya, carried mostly R1a and R1b, as well as I2a and Q1a, but again, no N1c1.

That's not to say that N1c1 won't ever turn up in EHG remains. But in my opinion the major subclades of N1c1 can't be associated with EHG, but rather with later populations of more complex origin, such as early Uralic-speakers with significant levels of East Eurasian admixture.

By the way, the claims about Ust-Ishim in the paper are interesting, but in my opinion not very parsimonious based on the data we have at the moment. I could be wrong though. Let's wait and see.

Citation...

Valouev et al., Reconstructing Genetic History of Siberian and Northeastern European Populations, bioRxiv preprint, Posted October 18, 2015, doi: http://dx.doi.org/10.1101/029421

160 comments:

Rob said...
This comment has been removed by the author.
Nirjhar007 said...
This comment has been removed by the author.
Karl_K said...

They seem to think hypothesize that maybe the spread of N1c1 was limited to the very far North initially (domestication of reindeer)?

"Such rapid movement of people may be explained by a technological breakthrough, such as invention of sleds or domestication of reindeer..."

Davidski said...

I'm sure N was in Europe well before 1,000 BC.

IR1 from Gamba et al. belongs to N, and is dated to 830–980 cal BC.

Chad Rohlfsen said...

What was the timeframe of the N in W Russia, from those one samples? Wasn't that group a mix of R1a and N1?

Nirjhar007 said...
This comment has been removed by the author.
Davidski said...

Ah, yeah, I forgot about these results.

http://eurogenes.blogspot.com.au/2015/02/ancient-r1a1-and-n1c-from-western-russia.html

Shaikorth said...

@Nirjhar

As the article in David's link says N1c1 was in Smolensk near Belarusian border 2500 BC.

Around 1000 BC more east eurasian influenced steppe groups appear, but they don't have N1c1, but C and Q in addition to R1a.

Chad Rohlfsen said...

Yeah. I think that is the starting location of Corded Ware too. Likely before the N1c arrived though. But, you never know. N1c at 2500 BCE is pretty interesting.

Davidski said...

That's not the starting location for Corded Ware. Not even close.

Nirjhar007 said...

Sorry, I also forgot about it :P....

Nirjhar007 said...

BTW David, they were just R1a1 ? like Karelia?

Davidski said...

I don't know what they were exactly. But to get a population like early Corded Ware we need something like late Khvalynsk near the Volga.

Chad Rohlfsen said...

I still think Belarus makes more sense than Ukraine. They lagged behind on copper working and didn't have a lot of EEF.

Nirjhar007 said...

I see thx, and of course one thing all CWC tested so far have the M-417 mutation, is this because of the lack of markers for further downstream testing?

Davidski said...

Chad,

Belarus makes no sense at all. Most Corded Ware have more Caucasus than EEF.

They came from somewhere near the Volga or Don.

Nirjhar,

Plenty of info online about the Corded Ware R1a, and the quality of the sequences. No need for me to go over it here.

Nirjhar007 said...

David,
That's not the answer of my query :P just say ''yes'' or ''no'',
Yes, i know they are online but i thought that *you* may give a better/reliable answer now as i wanted to be sure , just that...

batman said...

Y-dan N goes with the Combed Ware - if anything.

Then one have to consider that the combed-ware people didn't bury their dead, but cremated them. Mostly the ashed were deposited, but occassionally they made tumuli with the ashes collected in a jar. Sometimes together with artefacts.

The age of the first hg N is unclear - but the spread seem to follow the general direction of the foragers ("Hunter-Gatherers") that came out of the Baltics.

The migrational signals that moves in the oposite direction seems to be of female origin, as the mt-dna C - seemingly from the Sino-Mongolian tribes - was inter-marrying with the immigrants of the western y-dna N and R. Thus we got mt-dna C in Carelia at 7.500 BP already.

A later arrival of CZ seem to have occured around 4.500-5.000 BP, which made it into the north of Fenno-Scandia, where it still exist - along with the western mt-dna U5b1 - amongst the Sami. Amongst this group the y-dna N is sligthly dominant today, but hg 1 and R are frequently present amongst the southern Samis.

Better check the works of Willems and Rootsi on this. Here's what Willems have been saying about N and N-tat for the last 14 years:

"The peculiar Y chromosomal DNA variant known as Tat C, is dominant in almost all the indigenous peoples of Siberia, from the nomadic Yakuts right across to the Chukchis and Siberian Inuits living on the shores of the Bering Strait - regardless of what language they may communicate in.

Among the Finno-Ugric races of Europe this genetic inheritance is much more diverse, more multibranched, and hence apparently older than among any of the Siberian peoples.

It is characteristic of the European Finno-Ugrians both in the area of the Baltic Sea and in the Volga region (the Mari Republic east of Nizhni-Novgorod and west of the Urals). Thus it is an original Finn-Ugric feature. We have observance of between 35 and 60%, or about half of the paternal inheritance, and it has penetrated to some extent into Norway and rather less to Sweden."

"The figures for Poles, Slovaks, and Hungarians are already very very small, however, and in Western and Mediterranean Europe they go right down to zero."

idurar said...

"Poor wording" as someone said above. What about this?

These observations are consistent with
previous findings that ancestry of Eastern Europeans such as Mordovians, Finns, Russians, Saami and
Chuvash
cannot be explained by a mixture of three early European groups: the ancient northern Eurasians
(ANE, represented by MA-1), the West European hunter-gatherers (WHG, represented by Loschbour), and
the early European farmers (EEF, represented by Stuttgart). These Eastern European groups are more
related to East Asians
, which agrees with our observations of strong Eastern Siberian admixtures
contributing to the ancestry of Northeastern Europeans.


I'm guessing what the paper meant is that the ancestry of these Northeast European populations can mostly be explained by the same ancestral populations as Northern Eurasians and "East Asians" like Eastern Siberians and cannot be explained by the same ancestral populations as other Europeans because of too minor EEF-related ancestry and significant Han-related ancestry.

Coldmountains said...

@Chad

Makes no sense. I don't understand why so many ignore the teal/Caucasian ancestry of Corded Ware. How should they get it if they are not from the Eneolithic steppe? They also had just low amounts of R1b so they don't got it from Yamnaya.

Chad Rohlfsen said...

I'm not ignoring it. I'm saying the low EEF and the lack of much copper working doesn't fit well with a group recently from the Black Sea. I think the border region of Belarus and Russia seems more likely. EEF reached Belarus later and with less of an impact. No mega-cities there.

Coldmountains said...

@Chad

The Don region is enough far away from EEF centers of CT. Late Proto-CW is from the "West" where they got significant EEF and where ancestors of Sintashta became so much EEF admixed but early/archaic Proto-CW must be from the steppe else it is impossible to explain how early Corded Ware in Germany was 75% Yamnaya-like and so rich in teal. There was just one source for teal ancestry and archaic Proto Corded Ware impossibly got it living north of the steppe

Dmytro said...

@Coldmountains

Even on your terms, please don't forget that the Western steppe area of the Pontic has not yet been analyzed. This includes the PostStog area, the west-of the-Don Yamna, and the southern portions of the Dnipro-Donetsk cultures. The "teal/Caucasus" stuff may also be found there. And also the EEF. And the actual "start point" or "start pointS" of the various CW groups is no longer certain since the RC datings are being constantly reviewed.

German Dziebel said...

@Davidski

Strangely, Fig S16 in SI shows that it's Andeans and not Mansi are a better fit for MA-1. Mansi are way down the list of populations closest to MA-1 when Chimps or Mandenka are used as outgroups. So, I'm a bit puzzled by the disparity.

Also Mezen Russians are Y-DNA E1b1... forming a clade with Mandenka. This is interesting considering that Kostenki (Y-DNA C, which is more basal than Ust-Ishim NO) had the 'Basal Eurasian" component that - now we know - back migrated into Africa with Neolithic populations.

Chad Rohlfsen said...

Aside from the genetics. Corded Ware pottery, geometric designs, lack of much copper working, burial orientation to the West, does not match with groups stretching from the Don to the Dnieper. The groups there have lots of connections with the Caucasus and the Balkans..

Chad Rohlfsen said...

The area I'm talking about is west of Samara. It's not any further north.

Karl_K said...

@German

"Kostenki (Y-DNA C, which is more basal than Ust-Ishim NO) had the 'Basal Eurasian'"

Actually, we do not know that Kostenki's "basal eurasian" was the same as the "basal eurasian" in EEF.

Both were identified by observation of an outgroup to other populations.

Do they have a higher relation to each other than to other existing populations? Did the paper show a strong EEF afinity to Kostenki that required it coming via the basal component?

I have not seen that.

German Dziebel said...

@Karl_K

"Do they have a higher relation to each other than to other existing populations? Did the paper show a strong EEF afinity to Kostenki that required it coming via the basal component?"

I assume you mean the Seguin-Orlando Kostenki paper... No, Kostenki, while having BE (look at their ADMIXTURE run Fig. S20 in SI, it's the same EEF component), was still closer to Lithuanians than to Stuttgart, which should mean that "Basal Eurasian" is a subset of West Eurasian variation and not an early branch out of Africa. This is consistent with Y-DNA hg E being a subset of the Eurasian CT clade.

Karl_K said...

@German

...
"No, Kostenki, while having BE (look at their ADMIXTURE run Fig. S20 in SI, it's the same EEF component), was still closer to Lithuanians than to Stuttgart,"


Wow... pretty clear.

Not the same.

Thank you for clearing that up for me.

German Dziebel said...

@Karl_K

"Not the same."

Could you clarify? It must be either the same component or a closely related BE component, per ADMIXTURE.

Alberto said...

@Davidski

"My initial thought was that the MA1 or Mal'ta boy-related admixture estimate of 57% for Western Siberians (in fact, Mansis) was way too high."

I think that K8 would agree at about 55% WHG+ANE (probably around 30-35% + 25-20%?). Your K9 with EHG cluster mostly puts them close to 55% too (EHG+WHG, since EHGs themselves have 20%+ WHG):

https://docs.google.com/spreadsheets/d/19dLSYTTibNTOllihAaYaI0bBpGr2K0FMKQ4iSPci-Jc/edit?usp=sharing

I didn't read the paper carefully yet, but this strange confusion of Mansi (who have some ~40% East Asian admixture) as a possible source of ANE in Europe (and Yamnaya?) when we already have EHGs and MA-1 without East Asian admixture to explain it and the equally confusing references to KO1 having Western Siberian affinity don't make me too willing to dive much deeper into it. If they didn't figure out that East Asian admixture in Finns or Russians is not directly related to their ANE or that Yamnaya didn't have East Asian admixture it seems they have to work a bit more before formally publishing it (not saying the paper might not have some other merits, maybe it does, but it seems to have enough rough edges to make it less attractive to try to find those possible merits).

Petter Johansson said...

Like others pointed out, N in Europe is older than Uralic languages for sure.

Also, there are probably many layers of ENA admixture in Northern Europe, mostly coming from the Arctic region. I just think it would be a huge coincidence, if it would be shown that Uralic speakers were responsible for spreading all of it.

Nirjhar007 said...

PJ,
I agree.

Shaikorth said...

PJ, agreed. The inferred admixture events in northern and eastern Europe from Busby et al. support the idea. Look at all the src1-columns for sources of mixture.

http://www.cell.com/cms/attachment/2037871150/2052320367/mmc2.xlsx

D stats work a bit differently and perhaps can't so easily do the separation, but there may be something there. For instance D(Veps, Stuttgart; Andean, Chimp) and D(MezenRussian, Stuttgart; Andean, Chimp) are similar while D(X, Stuttgart; Buryat, Chimp) is clearly higher (with Z>3) for Mezen than for Veps. Of course, there never were Buryats in Mezen so the cause for this has to come from elsewhere.

Coldmountains said...

@Chad

What are you suggesting here? Corded Ware from Balkan or thr Caucasus? Not really and Corded Ware is a culture bringing single graves, pastoralism and new steppe ideas to North Europe. Of course Corded Ware culture contained more foreign elements because it was born by steppe folks moving out of the steppe in farmer and Hunter gather territories but the same kind of arguments were used some years ago to claim that Bell Beakers and R1b is from Iberia and this was obviously rubbish. Don't forget that Sintashta tribes are from Corded Ware and Sintashta tribes revolutionized equestrian warfare and created war chariots .

batman said...

@ Peter Johanson:

"Like others pointed out, N in Europe is older than Uralic languages for sure."

Proven by what?

"Also, there are probably many layers of ENA admixture in Northern Europe, mostly coming from the Arctic region. I just think it would be a huge coincidence, if it would be shown that Uralic speakers were responsible for spreading all of it":

All of it - perhaps not. Most of it, for sure.

What other y-dna than NO should it be that existed in the vast areas covered by Uralic speakers, that carried the Uralic languages along the Eurasian boreal-zone, from Balkan and Lappland - eastwards all the way to Mandchuria?

Then, what happened to them, as the N-group settled with the Uralian toungue across the woodlands and the northern taigas?

German Dziebel said...

@Alberto

" didn't read the paper carefully yet, but this strange confusion of Mansi (who have some ~40% East Asian admixture) as a possible source of ANE in Europe (and Yamnaya?) when we already have EHGs and MA-1 without East Asian admixture"

Agree. Russian geneticists are just obsessed about trying to find a Siberian population that beats Amerindians in terms of proximity to MA-1.

Krefter said...

I'm very skeptical about the legitimacy of the study that found 2500 BC N1c in Russia. I'm going to try to find an email of a author and find what their methods were. It reminds me of old-mtDNA studies which always found huge amounts of mtDNA rCRS(prob. H). I've read that a guy who sampled ancient mtDNA from Korea, said it was 100% H.

Krefter said...

@German Dziebel,

K14's relation to moderns follows the pattern of frequency of WHG/ANE ancestry. Whatever is creating his Basal Eurasian signal is differnt than what exists in EEF(ultimately has a Middle Eastern and maybe SE Euro source).

Karl_K said...

@German

"Russian geneticists are just obsessed about trying to find a Siberian population that beats Amerindians in terms of proximity to MA-1."

What is to be beaten? Where exactly do you think MA-1 lived? What world do you live in?

Shaikorth said...

@Krefter

SNP testing for Tat, which is the most legitimate way of finding N1c1. If it was a 6-STR test or something there would be a reason for scepticism.

batman said...

@ Davidski:

"The problem is that this isn't yet supported by any direct evidence from ancient DNA. Thus far, we know that EHG carried Y-haplogroups J, R1a and R1b, but no N1c1."

Please don't make "abscence of evidence into evidence of abscence". That have to often - in various branches of science - proven to be a grave mistake, crating unessesary short-circuits and copnsequent confuison.

As noted in another post - people who cremate thei dead doesn't leave samples for sequencing. Please observe, also, that cremation was the main traditition of ancient cultures. The few burial sites discovered are fractional compared to the amount of people that have lived, tens and even hundreds of generations in the various areas.

The overwhelming majority of graves/tumulis/kurgans/burial-nounds are from agricultural societies - and even they used cremation as the standard rule. Burial-mounds are probably just for the elected few, as men and women of rank.

Davidski:

"Later populations, with significant EHG ancestry, such as Corded Ware, Khvalynsk and Yamnaya, carried mostly R1a and R1b, as well as I2a and Q1a, but again, no N1c1."

Where I2, R1b, R1a and Q1a "hunter-gathers"?!

Kristiina said...

A few comments! :-)

The paper itself shows in Figure 4a that Evenk and Even N1c1 is more recent compared to Uralic N1c1. I think that it is obvious that the rapid expansion of N1c1 in Siberia is related to Steppe inventions: pastoralism and animal-drawn wagons. N1c1 men introduced the Northern version of it which included reindeer herding and sleds drawn by reindeers to Siberia and hence quickly replaced earlier yDNA. I do not think that the expansion of a certain y line is arbitrary.

On the basis of N1c haplotree (http://genome.cshlp.org/content/suppl/2015/02/18/gr.186684.114.DC1/Supplemental_Figures.pdf) my presumption is that N1c1 spread from Volga Ural area to its surroundings. YDNA N itself may have arisen in East Asia or not. The existence of the oldest branches in the West and the older coalescence age of Western branch compared to Chinese branch do not point to an origin in East Asia, but I let ancient yDNA to decide on this.

There is a telltale error in the main text where they claim that Saamis have 80% N1c1 and Finns 58%. The Finnish figure is correct but Saamis have less: Swedish Saami 37%, Finnish Saami 55% and Kola Saami 39%. Typically, Saami groups have high frequencies of I1: Swedish Saami 31% and Finnish Saami 41% (!). I presume that researchers instinctively think that because Saamis have a high Siberian percentage, they must have a high percentage of N1c1 but, in reality, even Latvians (38%) and Lithuanians (42%) have a higher amount of N1c1 than Saamis in general.

I really hope that we get ancient yDNA from Baltic rim/Estonia and West Siberian Baraba Steppe and Comb Ceramic cultures and from Oleni Ostrov in Kola peninsula. That will probably settle the origin and ancient distribution of Uralic yDNA N.

Romulus said...

The fact that the Motala samples contain the EDAR gene which is not present in Steppe people or R1 EHGs is strong evidence that this admixture in them did not come from them, neither is EDAR present in Loschbour or Labrana so it couldn't have been part of their original makeup. I believe the article is correct. I think what we are seeing is an immediate recolonization of Europe after the LGM from the south west by dark skinned, blue eyed, HGs belonging to hap I and from the North East by EHG types containing SLC45A2, SLC24A5, various MC1R mutations, and EDAR. mtDNA U5a seems to have a very strong correlation to EHG types and has been found as early as 8,000 B.C. in Germany.

batman said...

@ Davidski:

"...in my opinion the major subclades of N1c1 can't be associated with EHG, but rather with later populations of more complex origin, such as early Uralic-speakers with significant levels of East Eurasian admixture."

Can't all y-dna be viewed systematically - as branches from a common origin? In this case as macro-dad F - that obviously survived ice-time in northern Eurasia - whee his descendats later came about - as F-> GH, IJ and NO?

How xan a split from a stem, as in 'branching', be anything like 'complex' - if Y-dna N is a branch from NO - and NO is a subgroup to F/KL?

According to the old maths of the human genome y-group N are of the same age as F->GHIJK/NO, which fits pretty well with the first settlers of the Eurasian continent, as of 11.500 years ago.

That, again, fits very well into the branching east of the Swidrien-Kunda-cultures, that created Suomousjärvi and Komsa to the north, Ladoga-Volga-Ural to the east adn pre-Yamna to the south, later known as "Venedae".

All of them brought their first icons from the Baltic and their flint-tools from central Poland and Valdai. All of them remained "forest-cultures" - herding reindeers, goats and/or sheep.

The southern belt of these population connected frequently with the boat-culture of hg I and northen form of cattle-farmers, known as R1a. As climate improved some of these farmlands could be taken over by R1b.

That didn't make these "foreest-finns", who also practised slah-burn-cultivation, with rye and barely, to change their burial-traditions. Nor did it refrain them, or of having "Asian wives". Per consequnece their cattle-farming neighbours and nephewa would have brides with asian mt-dna - too.

As noted, the asian mitos reached all the way to Carelia already 7.000 years ago, entering northern Fenno-Scandia no later than 4.500 yrs BP, producing children with high cheeks, moon-yellow skin, shiny-black hair and eloquent brown eyes - thanks to mito C/Z, still apparent amomngs Samis.

Ultimately this north-western branch of herdes the N-group became known - among their southern cousins - as "Sa-manni" (later Sami), finnish for "gathering-men". By then they had already mixed, as some R1a and I also escaped into Lappland, becomming part of the 'sami-nation'. Similar mixes can be traced all along the Boreal forest-zone.

Avoiding submission, taxation and gospels from the Greek-ortodox or the Catholic churches the Sami -like some of their eastern cousins - were able to keep their traditions unto the 18th century. That's why we can't get any ancient DNA from theese populations, because they simply didn't bury their dead before the chritianization forced them to.


Whatever we think about these matters - under the old paradigm it was a sacrilage to bury the dead - keeping their "soul" from leaving the (dead) matter and get into the "higher realm" of the 'atmos'.

Re-incarnation or not, try finding a pre-christian gravefield - with skulls and bones - from the areas of todays American Indians, Inuits, Mansis, Kets, Permians, Finns or Samis. Even amongst the agriculturalists of pre-Christian Europe and Asia they are - indeed - rare exceptions.

Consequently - to keep a debate about real issues down to earth, we need to consider what we do not know without our own prejudisms, based on presumtions and consents from other sciences, rather than proof from ones own.

What we do know about the y-dna N-groups is that it is closely linked to both the Uralian language and the Combed Ware. As pointed out by Dr. Willems already 14 years ago - the diversity seems to prove that the spread of the N-groups started out in the Baltics. Thus we have N1-N2 and N3, along the way to Mongolia.

So far there's no evidnce to contradict that. Using the present mix within the Mansi, along with the dead-end genome of MA-1, can't change that, for obvious reasons.

Kristiina said...

As for Mansi mtDNA, I counted the percentage of their West Eurasian and East Eurasian mtDNA.
C 20.7%
D 19%
G2 1.6%
U4 14.3%
H 14.3%
J 14%
T1 3.2%
U2 4.8%
U5 1.6%
U7 3.2%
N2a 3.2%

East Eurasian 42%
West Eurasian 58%

That distribution is astonishingly close to their West - East autosomal ancestry distribution according to this new paper, i.e. 57% Western and 43% Eastern. Again, mtDNA is a reliable indicator of the autosomal composition of an ethnic group.

German Dziebel said...

@Krefter

Thanks. But at K=7 and up in ADMIXTURE (S20 in Sequin-Orlando) the same color component shows up in K14 and then at Basques, Ethiopians, Stuttgart, Iceman, Druze, etc. It's at very low frequencies but still visible in Lithuanians, too. This is "EEF" that has European Paleolithic (and not Sub-Saharan African roots). It was brought into Africa with Eurasian Neolithic.

What am I missing?

German Dziebel said...

@Karl_K

"What is to be beaten? Where exactly do you think MA-1 lived? What world do you live in?"

Earth to Pluto: MA-1 is an Amerindian back migration into the Old World for a simple reason that it's best modern representatives live in America and not in Asia. Raghavan et al's data is overwhelmingly supportive of this interpretation. And they suggest that if it's Amerindian gene flow, it must be older than 24,000 YBP because we find ANE all over Eurasia.

Do you have any problem with that?

Davidski said...

Chad,

Yamnaya has less EEF than Corded Ware (in fact, it has none). This must mean it came from the Arctic Circle.

Chad Rohlfsen said...

Just wait until we have Neolithic and chalcolithic stuff from Ukraine, Belarus, Lithuania, etc. I didn't suggest the Balkans or Caucasus. I suggest looking west of Samara, more towards those R1a remains, near Belarus. They wouldn't have picked up a lot of EEF through that region, into Germany. That is pretty much how it looks. Anyone from Ukraine should have a good amount of EEF like ancestry. The pottery of CW most resembles GAC and FB, again suggesting movements through that area, north of the Carpathians and certainly not in the southern half of Ukraine or Russia. The pits and mounds of Corded Ware are also different than the North Pontic, as they tend towards more a crescent shape and smaller, and again, the burial position is different. As far as Yamnaya lacking EEF, I still think it's an issue of reference populations. You can't have an Armenian pop, without 30% EEF, maybe 40% Anatolian, within that population. Teal is simply a product of Admixture using modern references, same as the Euro HG that sits opposite of the teal.

There is no serious connection to those regions. Here is one of several papers I have.

https://drive.google.com/file/d/0B962TtPkX1YnVXZJcUZ1MWw3WE0/view?usp=sharing

Davidski said...

Even Khvalynsk overall has too little Caucasus to be a good fit for early Corded Ware. Only late Khvalynsk more similar to Yamnaya works.

There's no way such a population was present directly west of Khvalynsk until the Corded Ware expansion.

You're confusing yourself with copper and pottery, when what we're discussing here is the geographic origin of the proto-Corded Ware population, which was basically like that one Khvalynsk individual that clusters with Yamnaya.

Chad Rohlfsen said...

It just looks way more like a fusion zone between Samara, GAC, and Southern Ukraine and Russia.

Davidski said...

I'm talking about the early Corded Ware people. They were basically like that one Khvalynsk individual that clusters close to Yamnaya (ie minus the EEF that Corded Ware have).

You need to take a closer look at the Corded Ware and Khvalynsk PCA and Admixture results.

Rob said...

@ Chad

" I'm saying the low EEF and the lack of much copper working doesn't fit well with a group recently from the Black Sea."

The Black sea steppe didn't exactly have much of a develop metallurgical industry ,either, even as late as 3000 BC. The metallurgical centres were the Caucasus, lower Danube and Carpathians, with unequal permeation to centres on the Caspian and Black sea steppe.

idurar said...

That Mandenka someone mentioned above and clustering with the E-V13 Russian Mezen in the Y haplogroup tree has haplogroup E-M78 (E-V12), which is a clear sign of North African ancestry (thus Eurasian ancestry). It's in line with ADMIXTURE and PCA plots where Mandenkas always seem more Eurasian-shifted than Yorubas.

This should be taken into consideration when choosing an outgroup for instance or when interpreting TreeMix.


Krefter said...

@German,
"Do you have any problem with that?"

Amerindians have lots of East Asian-affinity that MA1 lacked. They're a mixture of something closely related to Han and to MA1. There isn't actual Amerindian ancestry outside of Siberia and maybe surrounding areas.

Krefter said...

We might see lots of new Palaelithic-Late Neolithic genomes soon.

https://sites.google.com/site/pinhasierc/home

Chad Rohlfsen said...

Metallurgy of the pre-Yamnaya period was imported from the Caucasus, and only increased as time passed. This style does not lend itself to Corded Ware. This eliminates them from being in that immediate region. Khvalynsk or not, we have a 1000 year and half of Europe gap.

I think all cultural traits coming from Samara, Ukraine, and Poland only show that the genesis of pre-Corded Ware happened between Samara and the South Baltic.

I get what everyone is saying about the Caucasus admixture, but the timeframe eliminates the need for it to be from exactly that area at that time. Flow from the Caucasus was a constant thing, not a one time deal, where we say, look, that's it. We need more samples between Eastern and Central Europe during that 1000 year period. We already know R1a was in the area that I'm speaking of and at the right time. If they're travelling from Western Russia, through Belarus and into Poland around 3200-3000BCE, then what's the problem? What happens if the R1b and R1a in Khvalynsk are all dead ends? Turnover on the steppes happened a lot over a 3000 year period.

I would urge a lot of caution, because archaeology doesn't support a migration from the Volga or Black Sea coast.

Rob said...

@ Chad

Good reply.

"Metallurgy of the pre-Yamnaya period was imported from the Caucasus, and only increased as time passed. This style does not lend itself to Corded Ware. This eliminates them from being in that immediate region. Khvalynsk or not, we have a 1000 year and half of Europe gap."

-> So you're saying majkop/ CMP type copper / arsenic alloys were missing from CWC ?

"I think all cultural traits coming from Samara, Ukraine, and Poland only show that the genesis of pre-Corded Ware happened between Samara and the South Baltic. "

-> I'd agree with that .


" I get what everyone is saying about the Caucasus admixture, but the timeframe eliminates the need for it to be from exactly that area at that time. Flow from the Caucasus was a constant thing, not a one time deal, where we say, look, that's it. We need more samples between Eastern and Central Europe during that 1000 year period. We already know R1a was in the area that I'm speaking of and at the right time. If they're travelling from Western Russia, through Belarus and into Poland around 3200-3000BCE, then what's the problem? What happens if the R1b and R1a in Khvalynsk are all dead ends? Turnover on the steppes happened a lot over a 3000 year period."

-> Agreed.

" I would urge a lot of caution, because archaeology doesn't support a migration from the Volga or Black Sea coast. "

-> in regard to where ? Northern Europe / CWC ? And what's your opinion about the ultimate source of BB ?

German Dziebel said...

@Krefter

"Amerindians have lots of East Asian-affinity that MA1 lacked. They're a mixture of something closely related to Han and to MA1. There isn't actual Amerindian ancestry outside of Siberia and maybe surrounding areas."

It's rather Han who are Amerindian, but from a different variety than MA-1. All of East Asians are closer to Amerindians than they are to West Eurasians and all of West Eurasians are closer to Amerindians than they are to East Asians. It's pretty obvious that West Eurasians and East Asians split from Amerindians and subsequently did not admix with each other. Different Amerindian groups may very well have mixed with each other inside the New World (because they all share a continent), but East Asians and West Eurasians must have gone separate ways without admixing. So, admixture does not contradict an Amerindian origin for all of Eurasians. On the contrary, it supports it. East Asians and West Eurasians are both too heterozygous to create such a homozygous population as Amerindians. Prior to the discovery of MA-1 DNA, a bottleneck from more heterozygous East Asians to less heterozygous Amerindians made some sense, but now that founding admixture between East Asians and West Eurasians is postulated, a bottleneck hypothesis just doesn't cut it anymore.

Rob said...

@ Rob

"So you're saying majkop/ CMP type copper / arsenic alloys were missing from CWC ? "

Corded Ware metals still employed pure copper, reminiscent of old Balkan traditions, which are also found in Balanavo settlements further east.

Davidski said...

Chad,

Archaeologists haven't been able to figure out where Corded Ware came from for the past 50 years, so I doubt you will based on archaeological evidence.

But genetic evidence shows the origin of the early Corded Ware population to be near the Volga. You can see this on any decent PCA if you care to look.

Anything significantly west or north of the Khvalynsk and Yamnaya sites sampled to date isn't a viable option because the proto-Corded Ware population had lower EHG and higher Caucasus than any of the Khvalynsk samples.

You can only be right if Late Neolithic samples from Belarus and surrounds show totally unexpected results. But of course they won't.

capra internetensis said...

@Davidski

Why not anything *west* of Khvalysnk? Khvalynsk is a hell of a long way from the Caucasus, most other steppe cultures were closer.

Nirjhar007 said...

But genetic evidence shows the origin of the early Corded Ware population to be near the Volga.
Then we may see a largely R1a dominated population from that area, aDNA should decide it...

Aram said...

And what about this early split branch that is present in Balkans?
http://yfull.com/tree/N-P189.2/

It's age is 22000 years and it is found in Balkans. Is he also an Uralic marker? I don't think.

Davidski said...

Capra,

That's right, Khvalynsk is a long way from the Caucasus, so how are we going to get the Corded Ware minus EEF genotype anywhere that is even further from the Caucasus?

Do you really believe that such a genotype existed in the forests west of that Khvalynsk cemetery?

When the Mathieson dataset is released I'll pick out the Corded Ware individuals with the lowest levels of EEF, and we'll see where they'll cluster in TreeMix and PCA.

I'm willing to bet they'll look very similar to one of those Khvalynsk samples and the Caspian Yamnaya.

I'm also willing to bet that Late Neolithic samples from Belarus will look very different, with a shitload of WHG, and maybe even no R1a, because those Russian results were obtained with PCR, which might mean they were contaminated.

Davidski said...

Aram,

Do we know that this N lineage was in the Balkans 3,000 years ago?

Aram said...

Nations who don't speak Uralic languages but have N

Yakuts Turkic 88%
Chukchis 58 %
Tofalars Turkic 59%
Yupiks (Eskimo-Aleut ) 50%

I just listed the cases who have absolute majority. There are also others. Baltic nations also.
https://en.wikipedia.org/wiki/Y-DNA_haplogroups_in_Central_and_North_Asian_populations

Of course I am not so hypocrite to completely deny the obvious fact that the N is associated with Uralic languages. But I expect from an academic scholar some cautiousness and rigorousness when dealing with linguistic issues. Declaring a whole branch N1c1 that is 14000 years old an Uralic marker certainly is not helping to our better understanding of the ethnogenesis of Siberian nations.

Davidski said...

Aram,

Sorry, who declared a whole branch of N1c1 that is 14000 years old an Uralic marker?

Aram said...

Davidski

///Do we know that this N lineage was in the Balkans 3,000 years ago?///

Of course we don't know and I doubt about an old presence of N in East Europe. Because there is no much solid aDNA to prove that. And we all see that the modern distribution is not much helpful for understanding the past. So my comment was just a call for cautiousness when linking linguistics with genetics.
After all the same is constantly demanded for IE studies and I think the same should be applied to other linguistic families.

Rob said...

Ha ha Aram
Glad to see you picked up your own irony : not all N1c is Uralic, but all R1 is PIE...

Nirjhar007 said...

oh Comon Rob!, PIE is a special case, there is a reason that it originated the most successful language family on the planet!....

Aram said...

Davidski

Ok I see that the authors of this study don't make such claims. My excuses for that.
I was confused by some comments. But I keep my words because there are addressed to those who wants to make such claims.

Rob

I never claimed R1 is PIE, You will not find a single such a comment by me.

Rob said...

Aram

It's ok I'm yanking your chain old friend

Nirj

Please elaborate

Kristiina said...

@Aram "I doubt about an old presence of N in East Europe. Because there is no much solid aDNA to prove that."

There is a big black hole in the area where Uralic speakers traditionally lived (east, north and northwest of Moscow), so the paucity of N1c samples is not a surprise.

https://drive.google.com/file/d/0B9o3EYTdM8lQS2xlMkVPU0k5Y2M/view

https://drive.google.com/file/d/0B9o3EYTdM8lQdzRlWXBQaXlnMGc/view

http://dienekes.blogspot.lu/2015/02/a-story-of-69-ancient-europeans.html

Currently, researchers often focus on presumed IE cultures and that surely has an effect on how we perceive the ancient yDNA distribution. There were probably also other groups in the surroundings that are not analyzed.


rozenfag said...

@Davidski: "Do we know that this N lineage was in the Balkans 3,000 years ago? "

Didn't IR1 from Gamba's paper had N? He is 2900 years old and from Hungary. http://eurogenes.blogspot.com/2014/10/genetic-continuity-and-shifts-across.html

Davidski said...

Almost in that case, but no cigar.

Aram said...

Kristiina

Ok I see the black holes. Just the two Karelian EHG didn't bring much clarity to this question. And Karelia is quite North.
I wish they find N with more digging, so we can have an idea where and when it was evolving.

Kristiina said...

The paper says that "Nenets appear as an early diverging group related to other European populations. The tree model suggests that 43% (95% CI: 38-47%) of the Western Siberian ancestry can be attributed to an admixture with a group related to modern-day Evenki people. Furthermore, Nenets share 38% (95% CI: 31-46%) of their ancestry with a group related to Even people. Consistent with this prediction, we observed particularly high affinity between Mansi and Evenki as well as between Nenets and Even people based on the D-statistic."

In this context, also Even and Evenk yDNA should be considered.
All Evenks (127): yDNA C-M217 97/127, 76%; N1b 18/127, 14%; N1c 18/127 14%; I 5/127, 4%
All Even (89): yDNA C-M217 46/89, 52%; N1b 13/89 14.6%; N1c 30/89, 34%

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0083570

Uralic groups do not have C-M217, so there is no need to explain Northeast Asian admixture in Nenets and Mansi with N1c.

George Okromchedlishvili said...

@German Dziebel
"So, admixture does not contradict an Amerindian origin for all of Eurasians. On the contrary, it supports it. East Asians and West Eurasians are both too heterozygous to create such a homozygous population as Amerindians. Prior to the discovery of MA-1 DNA, a bottleneck from more heterozygous East Asians to less heterozygous Amerindians made some sense, but now that founding admixture between East Asians and West Eurasians is postulated, a bottleneck hypothesis just doesn't cut it anymore."

Nope. Amerindians are abottlenecked population that has nothing to do with Crown Eurasians. They can not be at the root of Eurasians simply due to the fact that their shared ancestry with West Eurasians postdates East-West Eurasian split. This is why tree-mix, admixture and etc show them as a mix of East Asians and ANE. ANE is distinctively West Eurasian and Amerindians have a portion of genome that they obtained from fully diverged West Eurasians.

German Dziebel said...

@ George

" They can not be at the root of Eurasians simply due to the fact that their shared ancestry with West Eurasians postdates East-West Eurasian split."

No. As I wrote earlier, ALL and ANY East Asians are closer to Amerindians than they are to West Eurasians and ALL and ANY West Eurasians are closer to Amerindians than they are to East Asians. If a "fully diverged" (this means already divided into daughter populations) West Eurasian population admixed into would-be Amerindians we would have seen only some (easternmost) West Eurasians being close to Amerindians, but not all of them. Those West Eurasians that did NOT show an Amerindian affinity would have been closer to East Asians. But the data at hand (both ancient and modern is unambiguous): even westernmost La Brana is closer to Amerindians than it is to east Asians (or to SSAfricans).

"his is why tree-mix, admixture and etc show them as a mix of East Asians and ANE."

TreeMix is designed to model admixture. So it's an artifact of the method. It can't model East Asians as a mix of Amerindians and West Eurasians, or West Eurasians as a mix of East Asians and Amerindians because they did not mix with each other, so they model an ancestral population (Amerindians) as an admixed population. But if you look at almost all STRUCTURE and ADMIXTURE runs (starting with Rosenberg et al. 2002), Amerindians are unadmixed from K=2 up.

George Okromchedlishvili said...

"No. As I wrote earlier, ALL and ANY East Asians are closer to Amerindians than they are to West Eurasians and ALL and ANY West Eurasians are closer to Amerindians than they are to East Asians. If a "fully diverged" (this means already divided into daughter populations) West Eurasian population admixed into would-be Amerindians we would have seen only some (easternmost) West Eurasians being close to Amerindians, but not all of them. Those West Eurasians that did NOT show an Amerindian affinity would have been closer to East Asians. But the data at hand (both ancient and modern is unambiguous): even westernmost La Brana is closer to Amerindians than it is to east Asians (or to SSAfricans)."

You're not making any sense here, sorry. Amerindians are a mixture of West and East Eurasians. Hence all West Eurasians will be closer to them than to East Eurasians by default. Of course La Brana is closer to Amerindians than to East Eurasians since it is much closer to ANE than to East Eurasians.

Tobus said...

@German: If a "fully diverged" (this means already divided into daughter populations) West Eurasian population admixed into would-be Amerindians we would have seen only some (easternmost) West Eurasians being close to Amerindians, but not all of them

Can you explain something to me German? I've heard you make this argument before and I'm wondering what it is you expect an admixed population to look like. For example, what do you think would be the situation if, say, an English man and a Chinese woman had a baby... would the baby be closer to some East Asians than his father is, but not others? Do you think the baby would show increased affinity only to certain European populations compared to its mother, or would it have a general affinity to all of them to some degree?

German Dziebel said...

@George

"Amerindians are a mixture of West and East Eurasians. Hence all West Eurasians will be closer to them than to East Eurasians by default. Of course La Brana is closer to Amerindians than to East Eurasians since it is much closer to ANE than to East Eurasians."

How can a supposedly derived population such as Amerindians be closer to all of West Eurasians who supposedly branched off tens of thousands of years earlier? Did it absorb only the ancestral West Eurasian alleles, while leaving all of the derived (locally developed) alleles behind? My model, on the other hand, fits the data well. East Asians are closer to northern Amerindians, while West Eurasians are closer to Southern Amerindians, but all of Amerindians are forming a clade of their own.

"Of course La Brana is closer to Amerindians than to East Eurasians since it is much closer to ANE than to East Eurasians."

"ANE" is a misnomer for a back migrated Amerindian population, hence your statement is tautological.

German Dziebel said...

@Tobus


Can you or someone tell me if those SS African populations (Mandenka, Yoruba, Mbuti), who show affinity with Sardinians vs. East Asians, show a similar shift toward Amerindians or not?

Thanks!

German Dziebel said...

@Tobus

The reason I'm asking this question is because I want you to start thinking about human prehistory in terms of real data, not your "peanuts in a bag" examples.

Karl_K said...

**And do not ever, under any circumstances, reply to Marnie, or comment moderation is back on.**

German Dziebel said...

I agre. On my website the comments policy is even stricter. Karl_k Andrew Capra and maybe Tobus won't qualify to post there at all.

Matt said...

Davidski: Curiously, I get a very similar result with Eastern European hunter-gatherers (EHG) in place of MA1, and the fit is almost as good.

Maybe not so curious. IRC, the relatedness of MA1 and Karelia_HG to Native Americans is not statistically different, nor East Asians, as per Haak. So the difference is really that MA1 is less related to Anatolian Neolithic and WHG (without really being more related to anyone, save perhaps South / South Central Asians). And if neither of those are (Early Neolithic or WHG) in the groups in qpAdm, qpAdm can't really tell the difference.

(One check on this could just be to check whether Mansi share more drift with EHG or MA1. If EHG, then probably EHG is a better proxy for their ancestors.)

capra internetensis said...

Chimp Mansi EHG MA1 -0.0307 -4.915

Not surprisingly.

capra internetensis said...

There is no point in explaining things to German, he is a troll who is not interesting in open-minded discussion.

German, since you refuse to learn anything about genetics, why don't you contribute something from your actual field of expertise? Uralic kinship systems or whatever?

Alberto said...

@Capra

Chimp Mansi EHG MA1 -0.0307 -4.915

Not surprisingly.


Yes, IIRC all modern populations are closer to EHG than to MA1 (except maybe Amerindians by a bit?).

But in any case, it's also not surprising because Mansi have a good amount of WHG (probably higher than their ANE). Maybe Nenets are a bit more MA1 than Mansi and a bit less EHG, but in any case they will be clearly closer to EHG too.

German Dziebel said...

@Capra

"German, since you refuse to learn anything about genetics"

You're funny. I did study population genetics at Stanford and continue to learn, as we all. If you'd like to learn about kinship, read my book http://www.amazon.com/gp/product/B011W9EPKI?keywords=dziebel%20kisnhip&qid=1445368779&ref_=sr_1_fkmr0_3&sr=8-3-fkmr0.

And you need to learn how to argue rationally and think scientifically. So basically, for one Capra, it's starting from scratch.

capra internetensis said...

That's nice German, I also studied undergraduate population genetics, and much more recently than you I suspect.

Since your appalling ignorance is clear to everyone, shameless appeals to your own authority won't cut it.

German Dziebel said...

@Capra

"Since your appalling ignorance is clear to everyone, shameless appeals to your own authority won't cut it."

But since you and others can't rationally express what this "ignorance" of mine is all about, any population genetic knowledge you acquired as an undergrad was a waste of (someone's) money. What's the point of teaching genetics to someone who won't be able to work with the data and get his thoughts together into an argument?!

capra internetensis said...

We are perfectly capable of rationally expressing it, and you, I am quite certain, are perfectly capable of understanding it; you simply choose not to, because you are a troll.

And now I am done feeding you.

German Dziebel said...

@ capra

I haven't seen a single rational argument from you. Calling someone a "troll" does not constitute a rational argument.

And thanks for sparing me of your hysterical fits from now on.

George Okromchedlishvili said...

"How can a supposedly derived population such as Amerindians be closer to all of West Eurasians who supposedly branched off tens of thousands of years earlier? Did it absorb only the ancestral West Eurasian alleles, while leaving all of the derived (locally developed) alleles behind? My model, on the other hand, fits the data well. East Asians are closer to northern Amerindians, while West Eurasians are closer to Southern Amerindians, but all of Amerindians are forming a clade of their own."

Closer to whom? They are closest to East Asians. In fact in global Tree Mix analysis Amerindians are consistently placed with East Asians. They are after all simply 60% East Asian + 40% ANE (and that is for Karitiana. Northern Amerindians have even less ANE).
And yes Northern Amerindians do have more East Asian affinity than Southern ones. It is easily explained by the fact that Inuits and similar people have been coming to Northern America in relatively recent times. How do you think Greenland has been settled?

I think you don't understand population genetics enough to differentiate between some allele sharing and overall genetic similarity. Southern Amerindians indeed have the largest direct ANE ancestry in the world. However the remaining 60% of their ancestry is very different from ANE. While say Northern Europeans have s considerably smaller direct ANE ancestry but are overall much slower to Malta since ANE and WHG are very close to begin with.
Just read the papers on subject carefully. All reputable scientists and genome bloggers unambiguously state that ANE is a West Eurasian ancient genome. Amerindians (even South-Central ones) are a primarily East Asian population with a relatively large degree of ancient West Eurasian admixture. That's all.

Tobus said...

@German: I asked first.

@George:
... that is for Karitiana. Northern Amerindians have even less ANE
Southern Amerindians indeed have the largest direct ANE ancestry in the world.

Just to let you know, there's actually no difference between North/South Amerindian affinity to MA1 (see Figure SI 23 from the Mal'ta paper)... unless of course you take "North Amerindian" to mean only relatively recent migrants from Siberia (Inuit etc. over last 5000 years), and "Southern Amerindian" to mean everybody else, including Athabascans. I've never seen a similar chart for East Asian affinity, but I expect it's pretty much the same.

German Dziebel said...

@George

Now were talking! Capra, take notice: that's what a rational argument looks like. Good job, George, although you misunderstand the situation. Here's why.

First of all, I don't care what "reputable scientists and genome bloggers" say. I care what the data says and what it doesn't. I'm professionally trained to do that. Sometimes "reputable scientists and genome bloggers" repeat what the data says and sometimes they repeat what previous generations of "reputables" thought about human origins and the peopling of the Americas. Hence, they don't always have a "perfect pitch" to the data, even their own.

"ANE is a West Eurasian ancient genome."

MA-1 is a DNA sample from a material geographically located in East Asia, so your statement is already confusing. Yes, it's overall closer to West Eurasians (I know that, alright) than to Amerindians and it's geographically in Eurasia and not in America. Genetic distance is often proportionate to geographic distance. So, everything is logical. But what you are missing is that Amerindians do not share any of MA-1's other genetic components that make it closer to West Eurasians. MA-1 and Amerindians share only what's found at highest frequencies among modern Amerindians. (You know that, alright.) If everything that makes MA-1 closer to West Eurasians had been "there" before MA-1 admixed into East Asians to generate Amerindians, then Amerindians would have had it, too. But they don't. hence, there was no "West Eurasian" admixture from MA-1 to Amerindians. (MA-1's mtDNA is hg R, which is a West Eurasian lineage but it's not found in the Americas.) So your idea does not work.

My idea, on the contrary, works perfectly well. Genetically, MA-1 is basal to all of West Eurasians (mtDNA is R-basal, not R1 or else), hence all of its West Eurasian components postdate the ANE component that it shares with Amerindians. Hence, MA-1 is Amerindian-admixed or Amerindian-derived. Since ANE is found all over West Eurasia, it means MA-1 and all of the West Eurasian populations that carry ANE are Amerindian-derived.

"Northern Amerindians do have more East Asian affinity than Southern ones. It is easily explained by the fact that Inuits and similar people have been coming to Northern America in relatively recent times."

So, here's a problem with that. An East Asian population admixes with West Eurasians and thus the first, main wave of Amerindians is born. But then the later one(s) don't have that West Eurasian admixture! How is that possible? That's not science, that's mythology. If there was a West Eurasian admixture into an ancient East Asian pool, then with multiple migrations into the Americas we would have seen more ANE in the north (and obviously among modern Siberians) and less ANE in South America. The opposite is true, however.

Good try, Curious George, but no, no cigar! (The system asked me to select images with mountains to prove that I'm not a Russian robot, so I thought it was very appropriate before I post a response to a Georgian.) :)

German Dziebel said...

Sorry, Y-DNA R above.

German Dziebel said...

@Tobus

"@German: I asked first."

It doesn't matter who asked first. What matters is the quality of the question. Yours was irrelevant. Mine was appropriate. It will help you get an answer to yours as well. And no games, please. Just do some leg work, get the requested data. Don't try to dodge into your sophistry.

German Dziebel said...

@George

" In fact in global Tree Mix analysis Amerindians are consistently placed with East Asians. They are after all simply 60% East Asian + 40% ANE"

Re this: They just model it this way. Prior to all the recent buzz about admixture they would have modeled it as a simple tree in which Amerindians are phylogenetically more upstream than East Asians and West Eurasians but with West Eurasians sharing fewer alleles with Amerindians than East Asians do. It's like having a tree with 4 mutation steps for East Asians and 6 mutation steps for West Eurasians from a root shared with proto-Amerindians.

Rob said...

German

That's very interesting. So if Mal'ta boy (c. 22 kya) represents (or part of) a back migration from America, then scholars who date the Clovis culture and such to c. 14 kya must be horribly off. Ie humans must have been in the Americas since at least 30 kya ?

German Dziebel said...

@Rob

You are darn tootin. They are off. That's why geneticists (to please archaeologists) keep creating scenarios in which Amerindians were hanging around Siberia between Malta times and Clovis times and then expanding rapidly to Tierra del Fuego.

BTW, there's a new dissertation devoted to Topper (http://trace.tennessee.edu/utk_graddiss/3466/) that shows that humans were in the Americas 50,000 years ago and it's only because we were using an Upper Paleolithic measuring stick to identify ancient lithic activity in the Americas that we missed what are clearly human artifacts at that site.

Kristiina said...

@Aram “Just the two Karelian EHG didn't bring much clarity to this question”

In a way they did. I think that Ryukendo Kendow once said on this blog that judging by ancient yDNA, yDNA often changes when economy changes. These Karelian hunter-gatherers show that there is probably no yDNA nor linguistic continuity in Finland from 6 kya. In recent analyses on Finnish yDNA and mtDNA, it has been concluded that there is more U in the east and U5b is the oldest mtDNA but there has been no evidence for claiming that N1c would be old and originate in the eastern part of the country.

George Okromchedlishvili said...

"Re this: They just model it this way. Prior to all the recent buzz about admixture they would have modeled it as a simple tree in which Amerindians are phylogenetically more upstream than East Asians and West Eurasians but with West Eurasians sharing fewer alleles with Amerindians than East Asians do. It's like having a tree with 4 mutation steps for East Asians and 6 mutation steps for West Eurasians from a root shared with proto-Amerindians."

You can not "just model" things. Amerindians perfectly fit as a mixture of ANE and East Asians. Neither Kostenki, nor Ust Ishim, nor Oaze can be obtained with good accuracy by a similar fitting method.
What does "prior to all the recent buzz" mean? Yeah, prior to the discovery of ancient Neolithic farmers many models showed Mediterranean as a native European component. That is why models built with the additional help of ancient samples are better than those without them.
You have to admit that your initial claim about ANE being Amerindian is simply wrong. And additionally that there is no controversy in WHG and ANE distances fro Amerindian.

I will further add that Ust-Ishim who has a Y-DNA at the root of both Western and Eastern Eurasian lineages plus Amerindian ones. An he is indeed equidistant between WHG and ANE on one side and East Asians on the other. So here you have a Crown Eurasian in Siberia 45 thousand years ago who was among the cline ancestral to modern Amerindians. So your theory really does not make sense.

Tobus said...

@German: What matters is the quality of the question. Yours was irrelevant

Let me rephrase then. If a group of 40 wandering West Eurasians (WEu) bump into a group of 60 wandering East Eurasians (EAs), and they start wandering together and interbreeding, how would you expect their 40/60 admixed descendants a few generations later to appear? Would you expect the admixed descendants to be closer to some EAs pops than their unadmixed WEu ancestors were, but not others? Do you think they would show increased affinity only to certain WEu populations compared to non-admixed EAs pops, or would they have a general affinity to all of them to some degree?

This is an honest question so please don't be defensive - I've already decided not to correct your mistakes, and I'm asking because I'm genuinely interested in what you expect such a mixture of populations would look like (and why!).

Tobus said...

@George: So here you have a Crown Eurasian in Siberia 45 thousand years ago who was among the cline ancestral to modern Amerindians.

You are completely correct - UI certainly appears to be ancestral to all modern non-African populations, Amerindians included.

It's worth noting that the genetic affinities of the Kostenki sample also preclude an American influx into Eurasia - if this hypothetical mixture happened before K14 then we'd expect K14 to be closer to Amerindians than MA-1 is (and he isn't). But if the mixture happened after then he wouldn't show any "Amerindian shift" if the admixture came from America into MA-1 - only MA-1's descendants would show it and K14 would be too early to pick any of it up. The only solution that fits both the UI and K14 affinities is a post-K14 admixture from Eurasia into America, just like Raghavan etc. al. proposed.

Aram said...

Kristiina

I think Pit Comb Ware culture should be checked for the N haplogroup. In Finland there should be some remains.
Most probably it came from East.
It's most interesting feature is the fact that it has ceramics but it is a hunter gatherer culture.
https://en.wikipedia.org/wiki/Pit%E2%80%93Comb_Ware_culture

Aram said...

Btw that N1c from the Russian study is from Zhizhitskaya culture.

Krefter said...

I tend to think all Finno Urgics have a lot of Indo European ancestry. If you look at the Y DNA of Finno Urgics and Turks in Russia, besides N1c most have R1a-Z93, R1a-Z282, R1b-Z2103, R1b-L151, and I1.

Finnish and Estonians cluster closely with IE-speaking Balts and Slavs. Finno-Urgics have the same signal of Yamnaya-type ancestry. Were early Finno-Urgics the same Teal+EHG mix as early Indo Europeans? How likely is that? It makes more sense they're mostly or largely of Indo European-decent.

Dmytro said...

Years ago when I was reading a lot of Soviet era archaeological literature I remember a statement that after the initial IE push which created Fatyanovo, there were a number of back and forth movements by Uralics and IEuros, which mainly resulted in language shifts. The last "push" in historic times by Slavs versus Balts and Ugrofinns. So there were large swaths of territory where the basic population largely stayed in place while switching speech preferences.

German Dziebel said...

@Tobus

"t's worth noting that the genetic affinities of the Kostenki sample also preclude an American influx into Eurasia - if this hypothetical mixture happened before K14 then we'd expect K14 to be closer to Amerindians than MA-1 is (and he isn't). "

We've already discussed that and you're wrong. Patterns of genetic variation reflect both time and geographic distance. K14 is more westerly than MA-1, hence its Amerindian signal is weaker. But it has "Basal Eurasian," which MA-1 and Amerindians don't have. That's the component that likely evolved in West Eurasia after the initial Eurasian population had moved from Siberia into the Russian Plains.

"UI certainly appears to be ancestral to all modern non-African populations, Amerindians included."

Again, we discussed it in the past. See http://anthropogenesis.kinshipstudies.org/2014/10/ancient-ust-ishim-dna-as-seen-from-the-americas/. The Neandertal fraction of UI ancestry is most strongly expressed in Amerindians followed by East Asians and West Eurasians (again in a perfect geographic cline), so UI is another good example of Amerindian back-migration in Eurasia, maybe the one that illustrates a joint West-East Eurasian node but of Amerindian origin nonetheless.

An open question is what fraction of UI Neandertal ancestry is found in the Sahul. Paabo didn't show that data. And it's possibly even stronger than the Amerindian link but regarding East Asians (without Papuans/Australians) and West Eurasians the situation is very clear.

Karl_K said...

@German

"An open question is what fraction of UI Neandertal ancestry is found in the Sahul."

Actually... an even better question is how these people in Oceania have Denisovan admixture from at least 50,000 years ago, yetva similar component was found in Sourh Americans without Denisovan admixture. That must mean that people were in South America for over 50,000 years. But then they were swamped by newcomer ANE/EastAsian people (Eurasian immigrants to the Americas).

German Dziebel said...

@George

"What does "prior to all the recent buzz" mean?"

It used to be that populations were modeled as "pure" and a tree was supposed to represent progressive splits without mergers. During those times Amerindians were modeled as an offshoot of East Asians. When people saw Amerindians' long branches, they interpreted them as "genetic drift". Then an "admixture turn" happened in population genetic studies. So they created another phylogenetic link for Amerindians - now linking them with West Eurasians - in addition to linking them with East Asians. They call those additional phylogenetic links "admixture edges". So the deviation of Amerindians from an East Asian source began to be interpreted not as "genetic drift" in the New World but as admixture with a non-East Asian population. So, it's like having two "normal" trees mapped onto each other. But if you just stay with the original phylogenetic model, you just elevate Amerindians to a higher node than East Asians and West Eurasians and you get a better solution to the puzzle because you just have progressive splits with no admixture between East and West Eurasians (until Mansi times).

Now the new paper (http://anthropogenesis.kinshipstudies.org/2015/07/amerindians-are-even-more-genetically-diverse-and-older-than-we-thought/) actually showed that 1-2% of Amerindian ancestry (in the very same Karitiana) has an autonomous link to Papuans (alleles not shared with either East Asians or West Eurasians). This confirms my model nicely, as it's extremely unlikely ("unlikely" as in "fantastical") that East Asians merged with West Eurasians and with Papuans to generate Amerindians.

"You have to admit that your initial claim about ANE being Amerindian is simply wrong. "

Why is it that scientists are always pressured to admit that they are wrong in the face of mounting empirical evidence to the contrary?! I labeled Tobus "cryptocreationist" at some point because he dismisses all the evidence and hides behind vacuous phrases such as "German's mistakes" and "German is wrong."

German Dziebel said...

@Karl_
"Actually... an even better question is how these people in Oceania have Denisovan admixture from at least 50,000 years ago, yetva similar component was found in Sourh Americans without Denisovan admixture"

Denisovan admixture is found in Amerindians as well at lower frequencies than in Papuans. Amerindians tend to have more "archaic (both Neandertal and Denisovan) ancestry" than East Asians or West Eurasians. But since there were no "archaics" in America (or Papua New Guinea), I think this is in fact a trace of common descent of modern humans from a Eurasian hominin related to Denisovans and Neandertals, not admixture. Admixture may have happened on top of that primal ancestry but it can't explain it all. Archaic admixture is what we have in Sub Saharan Africa.

But it's a rather swampy area where more data, ancient DNA and direct comparisons are needed.

German Dziebel said...

@George

"Amerindians perfectly fit as a mixture of ANE and East Asians."

Look at the STRUCTURE run in Rosenberg 2002 (http://anthropogenesis.kinshipstudies.org/2012/09/how-europeans-got-to-be-10-american-indians/). All the subsequent ADMIXTURE runs are not much different but I like to start with Rosenberg because I attended a closed-doors "early unveiling" of this new method by Marcus Feldman at Stanford in 2001 and he showed that very graph.

In it, all world populations "perfectly fit" as a mixture (at different proportions) of Amerindians and Sub-Saharan Africans at K=2 and Amerindians stay largely unadmixed all the way up to K=6. East Asians are mostly Amerindian with some SSA and West Eurasians are mostly SSA with some Amerindian, Papuans are in between.

Karl_K said...

@German

"I think this is in fact a trace of common descent of modern humans from a Eurasian hominin related to Denisovans and Neandertals, not admixture."

Really? Common descent from a Eurasian hominin? Then why all the stuff about Amerindians all the time? So one branch of Eurasian went to America for 10,000 years and then went back again? And that's when they became 'modern' or something? Honestly, this is all really hard to follow.

German Dziebel said...

@Karl_K

Well, they really had to speciate into "us" somewhere, in isolation from other hominins. Out of Africa does not even ask the question: where could modern humans speciate in a rather densely populated and well explored by archaics Africa. America is a new continent for hominins, hence it was conducive to the evolution of a rather peculiar new species. So yes, I postulate descent from a Eurasian hominin, a migration to the New World, speciation into "us," with a subsequent back migration to the Old World and replacement of hominins. A similar development is hypothesized for woolly mammoths (http://www.cell.com/current-biology/abstract/S0960-9822(08)00970-6?_returnURL=http%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0960982208009706%3Fshowall%3Dtrue).

I just follow the evidence: linguistic diversity is world highest in the New World followed by Papua New Guinea. Linguistically, America to the rest of the world is like Taiwan to Austronesians. Language, in my model, is a defining behavioral trait of modern humans. Out-of-America was originally inferred from this pattern in "deep" cultural data, which I think explains genetic variation well, too.

Tobus said...

@German: You failed to answer my question, again. It seems you are avoiding giving detail of how you think admixture works, so I guess we'll just have to leave it there - with me thinking the vacuous "German is wrong" again.

Alberto said...

Karitiana share very little drift with Ust Ishim. From ancient genomes they share most with Bronze Age Okunevo. And they share more with MA1 than with Kostenki, and more with Kostenki than with Ust Ishim. The oldest the Eurasian genomes, the less drift they share with Amerindians:

https://docs.google.com/spreadsheets/d/1MBElgax5_AQGQb-hOvPgEROsrr2Iokae-rxO10ymS3Q/edit?usp=sharing

German Dziebel said...

@Tobus

I'm not going to repeat again what I already wrote to you regarding your questions. They are irrelevant to this discussion. What I want you to work on is pull up D stats and Z-score for those African populations that show West Eurasian affinities. I suspect Mandinka, Yoruba and Mbuti are closer to West Eurasians than they are to East Asians. But what about Amerindians? Are those SSAfricans with West Eurasian admixture also closer to Karitiana than they are to East Asians? Then we can talk about your interests in what's going on in my mind. Maybe...

German Dziebel said...

@Alberto

Thanks.

"Karitiana share very little drift with Ust Ishim. From ancient genomes they share most with Bronze Age Okunevo. And they share more with MA1 than with Kostenki, and more with Kostenki than with Ust Ishim. The oldest the Eurasian genomes, the less drift they share with Amerindians."

At the same time Amerindians tend to share more with archaic Eurasians (Neandertals and Denisovans) than modern Eurasians. Also in Fu et al. SI S10.1B (PC2, SSAfricans excluded) Amerindians and Oceanians are opposed (in their own distinct way) to all other non-Africans including UI. So, UI forms a cluster with West Eurasians and East Asians, while and Amerindians and Papuans (precisely those populations that likely have extra long chunks in common with Neanderatls that "admixed" into UI) are their own clusters.

German Dziebel said...

@Alberto

Sorry, what's baRem in your spreadsheet?

Alberto said...

Bronze Age Remedello. Those get low scores with all for some strange reason (very drifted or genome quality, I'm not sure). The data is from Allentoft et al. 2015 (supplementary table 11).

Tobus said...

@Alberto: From ancient genomes they share most with Bronze Age Okunevo. And they share more with MA1 than with Kostenki, and more with Kostenki than with Ust Ishim

So it would seem the admixture may have happened sometime after MA-1... you don't have EHG results to include?

Alberto said...

@Tobus

No, Allentoft et al. didn't have EHG yet in their data set. hunterN should be SHG, which is lower than MA1. But Afontova_Gora is just above MA1, and it's younger. So yes, admixture probably happened quite after MA1. Siberia got depopulated after MA1, so it got populated again by a similar ANE population a few thousand years later, probably from South-Central Asia (anywhere from the South Caspian to SE Kazakhstan, where ever populations could survive after the LGM) and from around China from the East Eurasian side.

Kurti said...

Regarding this study and the problem I wrote few things on Eupedia.

is basically a joke, it isn't often that I and David agree 100% but this time definitely. The papers usage of wording is so confusing and they often contradict not only themselves but also known science(Northeast Europeans being close to East Eurasians than mainland Europeans really?) It really seems like those "Russian" scientist try to find as hard as possible a Siberian proxy population which beats Amerindians in their ANE scores.

They call the Mansi, not even something like "Proto Mansi" but modern Mansi ANE source population. They claim the first ANE came in combination with East Eurasian admixture but completely ignore the fact that Mal'ta didn't show any affinities to East Eurasian components. Not only that they also failed to explain how there is ANE in Yamna, Corded Ware and SHG without any evidence of East Eurasian admixture.

We have in the Near East populations such as Assyrians with zero East Eurasian admixture but over 13% ANE. Those are too many factors speaking against their Mansi theory.

Also if you take a look at Mansi y and mtDNA they have like 55% West Eurasian mtDNA and ~25% yDNA.

Mansi as a Ugric speaking population (whoms ethnogenesis by the way can't be older than some thousand years) are said to be ancestral to Mal'ta(20000 years) according to this paper.

Sorry but I stopped reading about it here. If anything Mansi represents a Population mix of a ANE like group which was absorbed into an East Eurasian like group.

The point is that Mal'ta predates Mansi and its genetic components are much closer to each other (as it should be) however with Mansi we are dealing with a population which is, a two way mix of East Eurasian and ANE admixture, Those components are quite distant. Based on fst distance and ancestral tree ANE and WHG are allot closer to each other anyways. So there is absolutely no way that ANE might have arisen, together with a component it is more distinct from, in a modern population which obviously has mixed ancestry.


to be continued.

Kurti said...

That most modern populations with ANE display some Mongoloid type features doesn't necessary mean it is connect to their ANE ancestry. More rather that Mongoloid type ancestry took over former ANE land, which obviously seems to be correct if you look at the history of these regions.

Contrary if you just pay close attention and compare the "East Eurasian" groups with more ANE to East Eurasian populations with less, one thing falls into the eye. Those groups with more ANE display more of the "pseudo" Caucasoid features than those with non or less.

Didn't most of the people and even scientist bet and said that the Kennewick man is going to turn out as West Eurasian autosomally because physically he looked so CCaucasoid? Turns out he is closest match to Amerindians. Why does an ancient Amerindian individual looks so much Caucasian compared to modern East Eurasians? Is it a because of their Han like ancestry they share with East Eurasians or B because of their ANE ancestry they also share with West Eurasian populations? Isn't the answer obvious? Just take a look at modern Amerindians, why is it that even isolated Amerindian groups and even Native Americans of old images look to have much more Caucasoid physical features (high nose bridge no Epicanthic fold), DESPITE ~40-50% of their ancestry being actually East Eurasians, compared to modern East Eurasians? Now imagine a full blooded ANE population, Is this also just coincidence? Obviously regarding ANE , we are dealing here with a predominantly West Eurasian population which was mixed or absorbed by a Han like group in most of Eurasia. Alone this is enough for me to come to the conclusion that Mansi can never be a source population of ANE. But even them display signficantly more Caucasoid features compared to Han Chinese for example. This is no coincidence but fits with their East and West Eurasian mixed origin based on y and mtDNA as well aDNA (ANE vs Han Chinese like ancestry)

German Dziebel said...

@Alberto

"Siberia got depopulated after MA1, so it got populated again by a similar ANE population a few thousand years later, probably from South-Central Asia (anywhere from the South Caspian to SE Kazakhstan, where ever populations could survive after the LGM) and from around China from the East Eurasian side."

This is consistent with the finding that a mtDNA hg X2 lineage re-entered South Siberia from West Eurasia in the past "few thousand years" and is not directly related to Amerindian X2 (http://anthropogenesis.kinshipstudies.org/2012/04/more-on-amerindian-haplogroups-x2-b2-and-c4-evidence-from-siberian-tubalars-tuvans-evens-and-ulchi/).

The mtDNA X-A nexus (if connected by a single mutation 152, per Maca-mayer 2001 http://www.biomedcentral.com/1471-2156/2/13) is a possible mtDNA correlate to ANE. X is non-existent in Siberia (save recent gene flow, per above) but is found all over Europe at low frequencies and in West Asia/northern Africa. A is very rare in Siberia but is found all over America.

Roy King said...

What do you all think of this Armenian paper, showing present day Armenians being an admixture between "Iceman" and Central Asians? Could the same phenomena that happened in Western Europe: a "return of the repressed" WHG in the Late Neolithic Era have occurred in Armenia and Anatolia, namely local hunter-foragers remained silent during the Earliest Neolithic and returned via adopting farming later bringing the Teal component to West Eurasia? Here WHG and Teal represent hunter-foragers in Western Europe/West Eurasia, respectively. Perhaps Y Haplogroup J is analogous to I in that both represent the local Mesolithic, geometric microlithic traditions.

Roy King said...

The paper is here:
http://www.nature.com/ejhg/journal/vaop/ncurrent/pdf/ejhg2015206a.pdf

Davidski said...

Armenians generally show a higher level of ancestry from Anatolian farmers than other Near Easterners.

This makes sense because they've been shielded in various ways from more recent population movements from the east, like that of the Turks.

But they're nowhere near as ancient Anatolian as Sardinians or even Basques, which just shows how massive the population turnover in the northern Near East has been since the Neolithic.

batman said...

@ Davidski;

"Armenians generally show a higher level of ancestry from Anatolian farmers than other Near Easterners."

I guess this statement builds on ancient dna rather than the present mosaic of the Anatolian populations. Or old, archeological hypos, for that matter.

Does aDNA prove that Turks and Armenians have a c-o-m-m-o-n ancestry - or that the Armenians have received their y- and mt-dna, f-r-o-m the Anatolians?

Matt said...

Massive replacement and/or quite genetically divergent in flow, e.g. Lithuanians who were estimated to be around 50 Yamnaya:30 WHG:20 LBK EN seem to share more of that f3(Yoruba,Iceman,X) drift with Iceman than Cypriots in the above linked paper (albeit marginally, and Iceman is probably more WHG from the Anatolian EN). Will this end up being because Cypriots have less % ancestry from the EN population, or because what admixed into them was more divergent from EN than WHG / Yamnaya were (relative to distance from Africa)? Ancient DNA will be interesting...

German Dziebel said...

@Alberto

Gisele Horvat has informed me that "The Altaian/Buryat/Tubalar/Teleut X sequences belong to X2e; a branch which is quite common in Georgia." So there indeed seems to be a good fit between autosomal and mtDNA, as Kartvelians is one of the few ANE hot spots in West Eurasia.

Simon_W said...

In the comments section on the latest Mathieson et al. paper („eight thousand years of natural selection...“) I explained to Nirjhar what the use and importance of autosomal DNA is, and I said that if the autosomal structure of Sintashta people cannot rule out a SC Asian origin of R1a-Z93, then the autosomal structure of modern S Asians cannot rule out an origin of their R1a-Z93 in Sintashta people either. So more arguments are needed to prefer one hypothesis over the other, and this is what I'm going to provide now, because I don't consider both explanations equally plausible at this point.

Obviously the origin of R1a-Z93 is closely tied to the origin of R1a-M417, because according to yfull their TMRCA is only 700 years apart.

While it's entirely possible and even somewhat likely that sheep herding teal people from SC Asia started to admix into steppe HG as early as 6000 BC, this is far too early for R1a-M417, which according to yfull dates to about 3500 BC.

But I think archeologically there is no evidence for a strong cultural impulse from the central Asian Kelteminar culture into the steppe around 3500 BC (the TMRCA of R1a-M417).

Furthermore, physical anthropology clearly suggests that the main vector of „teal“ admixture into steppe people was from the Caucasus and Transcaucasus area. Because the central Asians of that time were very eurymorphic people (having a very broad shaped face), whereas the tendency on the steppe was clearly an increase in leptomorphy, from eurymorphic HG to mesomorphic Yamnaya, which can be plausibly explained by West Asian admixture.

Then, somewhat later, we have Corded people from central Europe with R1a-M417 whose autosomes were similar to Yamnaya autosomes, but with an EEF admixture. Reasonable interpretation: The EEF admixture was from central Europe (or east central Europe) and originally they were like Yamnaya people, so the same problems regarding central Asian influence applies.

Then we have, much later, in the second millennium BC, Sintashta and Srubna people with various subclades of R1a-Z94. Their autosomal DNA is closely related to the Corded people. Though a very weak „teal“ impulse from SC Asia cannot be ruled out a priori, and maybe this might be responsible for their South Asian-like yDNA?

However, no, this can be ruled out a posteriori: The earliest finding of R1a-Z94 is in the Poltavka outlier. The Poltavka people were still very close to Yamnaya and carried R1b variants. Therefore this outlying intruder must be a very strong hint pertaining to the origin of R1a-Z94. Especially since this outlier lived between 2900 and 2500 BC, and the TMRCA of R1a-Z94 is 2800 BC, according to yfull, so he must have been one of the earliest R1a-Z94 people. If this guy was shifted towards SC Asians in the PCA, this could only mean that Nirjhar is right. However, this guy is not at all shifted towards SC Asia, but towards LNBA central Europe!

Nirjhar007 said...

To Prove that ''Nirjhar'' was right or Wrong, you have to wait a bit, but not for long AFAIK :)....

Simon_W said...

;)


Simon_W said...

On an unrelated note, did you people notice there is a Corded Ware male with R1b-M269 in the revised Mathieson et al. paper? It's the sample I1534, also called ESP14. In Krefter's list it's grouped as LN/BA Europe "unknown", but according to the paper "the individuals from this site (Esperstedt ref. site 4) could be unambigously assigned to the Corded Ware culture, both by accompanying pottery and by characteristic orientation of the burials. And ESP14 is a perfectly normal male from this site. The dating suggests that the single Vucedol R1b male from southern Hungary was a little earlier though, and there was also a slightly older possible R1b male from the Polish Corded Ware, but his sequence was poor, so it's not quite certain.

In any case ESP14 proves beyond doubt that R1b-M269 was present in the Corded Ware of central Germany. And since the early to middle Bronze Age Hungarians tested so far are 3/5 I2a, 1/5 G2a and 1/5 R1b, it's safe to say that the amazing expansion of R1b-L51 didn't start in Hungary, but in the Bell Beaker culture of Germany and the Czech republic. And although it's possible that this was descended from Vucedfol R1b, it's certainly more parsimonious that it was descended from local R1b.

Grey said...

"Perhaps Y Haplogroup J is analogous to I in that both represent the local Mesolithic, geometric microlithic traditions."

dna pushed into the mountains which later bounced back as hired herders seems plausible to me

batman said...

@ Grey

It does seem that the first HG carried F-GHIJ who eventually developed the first herding and small-scale, slash-and-burn agricutlure. This may incur that the descendants of the paleolithic Eurasians avtually had a basic form of domestication and agriculture in place already during the last 20.000 years of ice-time.

According to the archeological discoveries from the last two decades we do know that the ice-age Eurasians made a number of organized, seasonal tradtitions - where sawing, harvesting and domestication happened - following the rythms of nature and the seasons of the year.

Consequently we have to modify our term when it comes to "hunter-gathers" within the Mesolithic context. Besides harvesting, gathering and preservation of a vareiety of fruits, beside the fluor of grinded corns and seeds - they obviosuly went from trapping to domestication of wolfs, reindeer, horses, goats and sheep obviously have a solid background in the paleolithc.

The later, so-called "agricultural revolution" came about as the climate made the larger animals and farms to grow permanent settlements.

For some reason, like Lactose persistance, this seem to have favoured some specific haplogroups - like y-dna R and mt-dna U5b1 and H. For this reason we need to get a clue to how y-dna came about - as a presumed, downstream result of F->NO/P. Thus it is clear that both Y-dna G/H, I/J and N/O did inherrit some of these cultural traits. Lazter some of them picked up cattle-farmingm, too - seemingly confirmed by the lactose Persistance within the older y-dna groups, such as I and N.

batman said...

@ Simon W et al.

AFAIK, the oldest sample of R1a is 7.500 years - found at the border between occidental and oriental Eurasia - in Karelia - belonging to R1a-M459. The oldest traces of agriculture in this area is today dated 7.399 BP.

The highest frequencies of the downstream from R1a-M459 are now found in Northeastern and oriental Europe. Besides, the Carelian male seemed to have had three wives of mt-dna H, H1 and H4, apparently found also in the agri-cultures of Baalberge, Karsdorf and Gokhem.

Consequently we have to deem the R1a from Carelaia a early agriculturalist, rather than the presumtious title "hunter-gatherer".


Same thing with the 7,500-year-old R1b-sample from the oriental Samara, that belonged to pre-R1b-M478 from which a row of agriculturalists branched. The highest frequencies of the downstream from R1b-M478 today also seem to be found in Europe, though in the occidental areas - which clearly indicate that the Samarian individual had some close cousins out west.

To have a check on the "bifurication-area" bewteen R1a and R1b we may adress the following:

5650-5555 BC Samara R1b1a (M-297?)
5178-5066 BC Spain, El Trocs: R1b1a2 (M-269)

3516–3362 BC El Portal√≥n: R1b1a2 (M-269)
2910-2875 BC Samara R1b1a2 (M-269)
3300-2700 BC Lopatino II R1b1a2a* L-23

2887- 2634 BC Yamna, Russia R1b1a2
2857-2497 BC Volgograd, Russia: R1b1a2

2865-2578 BC Oblachzova, Poland: R1b1-L1345+

2500-2400 BC Esperstedt/CWC R1b1a2
2600-2500 BC Kromsdorf/BB R1b1b2
2296-2206 BC Quedlingburg R1a1a1


So the question remains of what these facts may require of honest logic and clever reason;

Is the R1b-269 from Le Trocs less ancestral to European R1b than R1b-297 from Yamna?

Which one of them - if any - is responsible for the R1b-269 samples from El Portalon, Esperstedt and Kromsdorf?

Then, again - from which area could a common source of R1* be placed to explain the source of R1b of El Trochs, Samara, Lopatino and Oblachova - as well as the R1a of Carelia and Quedlingburg, that made it to Scotland and northern Scandianvia - as well as to Andronovo, Tarim and Xiongnu?

Where could an area of common origin R1*, R1a and R1b be explained, according to the most simple and sound logic?

Krefter said...

@batman,

You should safe this. I've posted it before.

https://docs.google.com/spreadsheets/d/12G2cfjG0wHWarsl5bB99ridFmvUWzqlZfZ6_e_R6oIA/edit#gid=1870266760

Grey said...

@batman

"It does seem that the first HG carried F-GHIJ who eventually developed the first herding and small-scale, slash-and-burn agricutlure."

I don't know about that bit. You may be right but I was thinking more about HGs being pushed into the mountains by the expanding farmers but then in places where the farmers were mostly herders being recruited back down the mountain again as stockmen and eventually becoming a part of the farmer group hence I and J often seeming to be connected with a second ydna haplogroup.

Simon_W said...

Some people like Chad and Colin think there must have been a massive invasion of R1b people coming from Yamnaya / the western steppe which gave rise to the eastern Bell Beaker people with their complete predominance of R1b. The problem's just that there is no increase in autosomal steppe ancestry after the Corded Ware culture, but a decrease. So I'd rather favour founder effects in a late Corded Ware environment. The latest dates provided by yfull back me up: L51 must have come from the steppe, its TMRCA is 3700 BC. But underived L51 is very rare. Most of the R1b in western Europe is P312, which is dated to 2500 BC! Precisely the beginning of the Bell Beaker era in central Europe! That means there was no wave of P312 people and we'll never find a P312 male that predates the Bell Beaker culture. But the most recent common ancestor was one of the first Bell Beaker males of central Europe. The other common clade, U106, is dated to 2900 BC. That would be perfectly in line with an founder in the central European Corded Ware.

Krefter said...

@Simon_W,
"The problem's just that there is no increase in autosomal steppe ancestry after the Corded Ware culture, but a decrease."

The incoming R1b tribes may have picked up a lot of EEF/WHG ancestry while migrating to Germany. And we can't assume East German Bell Beaker got their Steppe or any of their ancestry from Corded Ware. They may have gotten their Steppe ancestry from a totally new group of people who traveled along the Carpthian Basin. For a few hundred years BBC/CWC were contemporary and had differnt autosomal makeups along with differnt Y DNA.

Simon_W said...

@ Krefter

Yes, they might have diluted it on their way west, that's a possible hypothesis. Or they might have acquired some western admixture that diluted the steppe ancestry, or maybe it was even from local farmer relics. In fact they were quite diverse, some had very eastern autosomes, others were very Iberian-like, they were not yet homogenized. Bell Beakers and Corded Ware had different autosomal makeups, but a lot of overlap, and naturally the conservativism of those Corded people who didn't adopt the Bell Beaker package may have also prevented them from diluting their steppe ancestry.

The big problem I have with the Carpathian Basin theory is the lack of R1b in the early and middle Bronze Age cultures there. It's somewhat hard to imagine that they simply marched through without leaving any trace. Tough the autosomal DNA of Hungary was quite strongly changed, compared to the Copper Age and Neolithic.

And don't forget, EBA Unetice apparently had lots of I2, although their culture was most of all grounded on a Bell Beaker basis, and their autosomes were very similar to eastern Bell Beakers. So what do we make of this? Do we want to suppose another invasion, this time of I2 people? No, I would rather think it was another founder effect associated with a new culture, a new fashion and the spread of tin bronze technology.

So regarding the linguistic problem of Italo-Celtic versus Germanic and Balto-Slavic, I tend to think now that Rob was basically right and there was just Western early IE, associated with the Corded Ware, and the different branches crystallized from this background over time, and with different mutual contacts, rather than imagining Italo-Celtic as a distinct branch that migrated through the Carpathian Basin and Germanic as a later mix.

Simon_W said...

I mean, why should the purported "Yamnaya-Beaker folk" leave the Carpathian Basin in the first place? It's a landscape similar to the steppe. If they were suffering from overpopulation, they would just have sent forth groups of colonists, instead of leaving that good place altogether.

Krefter said...

@Simon_W,

We have only 3 Unetice Y DNAs and all are I2(apart of very differnt branches). This doesn't mean Unetice was I2 rich. Plus a "Proto-Unetice" in Poland had R1a1a1.

We have R1b in Hungary that predates Bell beaker. And another R1b in Hungary from 2000 BC. However most had EEF Y DNA. And they(2000 BC, we don't have genome of 2800 BC R1b guy) had much less ANE than Bell beaker.

I do agree with you though. The Bronze age Hungarian genomes don't suggest the Steppe ancestry in Bell Beaker came from Hungary.

I highly doubt Bell beaker got their Steppe from Corded Ware women because they don't ave enough Steppe mtDNA to account for their on average 50%+ Steppe ancestry.

Rob said...

Krefter & Simon

Let's also remember that all BA Hungarian samples to date are from Transdanubia, and not from the Kurganed areas in the Tisza plain. It would be interesting to see what they show..

Simon_W said...

@ Krefter

Ah yes, you're right, RISE431 from Poland is from the transitional horizon to Proto-Unetice and has R1a1a1. He shows that there was some local continuity with the Corded people, as was expected from archaeology. The question remains how typical his haplogroup was in the fully developped Unetice culture. Admittedly 3 samples are not enough, but at least they seem to show that the predominance of R1b had been broken and that Unetice wasn't completely dominated by R1b.

Indeed we have one R1b from Vucedol. Well there were Yamnaya-related steppe people in eastern Hungary, and it's not unlikely that they had brought some R1b with them. And after getting assimilated, they may have transmitted this to Vucedol. But there's a difference between introducing some R1b to the gene pool and complete predominance of R1b.

The other Hungarian R1b you mentioned (according to the original paper dated to 1950 – 1760 BC) is from the Gata-Wieselburg culture located on the western end of the Carpathian basin, adjacent to the eastern Alps, in what is now eastern Austria and western Hungary. Archaeologically its derivation is obscure, but at any rate it was in an area where the fully developped Bell Beaker culture had been before, and they also had crania of typical Bell Beaker type. To me this suggests that they were biologically Bell Beaker derived, and therefore it's no surprise that they had R1b. But they cannot be considered ancestral to Bell Beaker people. To the contrary, I think Bell Beaker people from central Europe had migrated eastwards along the Danube.

With regards to the Bell Beaker mtDNA, it has been argued that it was utterly different from Corded Ware mtDNA. Partly this is correct, I guess that's mainly because of their H haplogroups, which may be from the west as Haak et al. had originally suggested (? but I'm no expert on mtDNA). But partly they also had steppe haplogroups. As regards their ANE/steppe ancestry, the Corded people didn't have plenty of steppe mtDNA either. In the end the autosomal steppe admixture got more or less homogenized in the population because of recombination, and in consequence was no longer correlated with specific mtDNA haplogroups – women with EEF haplogroups ended up having similar steppe admixture as their males and women with steppe haplogroups. I think this can happen quite quickly. And if then a new population arrives and displaces the men, it can pick up steppe admixture from women with EEF haplogroups. But if my hypothesis is correct, the R1b wasn't from an intrusive population anyway.

Simon_W said...

@ Rob

That's not quite correct, BR1 and BR2 were not from Transdanubia, but clearly from east of the Danube. However, BR1 as a female had no yDNA, and BR2 was LBA. Indeed both were from west of the Tisza, and hence not exactly from the steppe influenced area. But the Mako culture, to which BR1 belonged, had spread on both sides of the Tisza, thus incorporating a large part of the steppe influenced area.

Then, as for the Allentoft samples, the Maros and MBA samples were not from Transdanubia either, but from far southeastern Hungary, and Maros has yielded us a G2a. The remaining samples are from the Vatya culture near the Danube in central Hungary. It arose as a blend of Nagyrev and Kisapostag influences, thus merging together the two largest cultures on both sides of the Danube.

Simon_W said...

To my knowledge the idea that Beaker folk = Yamnaya folk has never been seriously suggested by archaeologists, it's more of an urban legend. But what has been suggested by some specialists is Vucedol influence in eastern Bell Beakers.

If you check the Vucedol map on wikipedia
https://upload.wikimedia.org/wikipedia/commons/thumb/1/16/Vucedol_culture_map.png/430px-Vucedol_culture_map.png
you'll note it would make sense. Judging by this map, the Bell Beaker culture in eastern Austria and all the Czech Bell Beakers have Vucedol influence.

The map also suggests that the contemporaneous culture in Hungary east of the Danube, which is more commonly called Mako culture ( = the culture of BR1!) is also kind of Vucedol-like. This means: Although we do not yet have an autosomal sample from Vucedol proper, BR1 is probably a good surrogate.

And I had noticed before, when analyzing the Bell Beaker samples from Allentoft et al. with Eurogenes K15 and Dodecad K12b, that at least the Czech Bell Beakers deviate from German Corded people into a Vatya direction – with a strong WHG-rich Baltic component and very low West Asian and Gedrosia. So this is perfectly in line with some admixture from Hungary. The same cannot be said however about the German Bell Beakers.

Also noteworthy about the Vucedol map: The early core area is immediately adjacent to the Yamnaya-influenced area.

So, what does it all mean? The autosomes of the German Bell Beakers don't suggest appreciable Hungarian admixture, but they surely had contacts with their Czech comrades which might possibly have spread one or the other R1b westward.

Simon_W said...

In fact, all Bell Beaker R1b found so far was either from central Germany (the vast majority), from Bavaria, and from the Czech republic. A rather narrowly contoured area, likely with lots of contact and exchange within. To what extent Bell Beaker groups outside this area had R1b is still unknown. (At least in Northern Italy it cannot have been very common, see the EBA Remedello I2.)

Rob said...

Simon

Good points. Again about Hungary- the fact still remains that non actual kurgan people have directly been excavated.
I think youre onto something with the diversity we can expect from BB - with several 'factions' and genetic signatures, especially about the role of BA Hungary - which Alberto has also touched on.

Soon there will be a good paper coming out (I hope) which attempts to link BB with yamnaya on socio-cultural grounds. IMO , Kurgans reach as far as the eastern tip of Austria, so it isn't too hard to fathom they then did a dash further west up the Danube.

Simon_W said...

What regards Vucedol, since they had R1b it seems possible that they were responsible for the R1b in Bell Beakers in central/southeastern Germany and Bohemia. However, there were 2 yDNA samples from Vucedol in the paper that found it, and one was I2a2, only one was R1b. So even with Vucedol as a source we can't get around a founder effect.

But personally I doubt that people from the Carpathian Basin were that influential on the eastern Bell Beaker culture and on the longer run. Because the Unetice crania from Moravia were very similar to Corded Ware crania from central Germany (as demonstrated in a multivariate Penrose analysis I've seen). Even though culturally they were mostly Bell Beaker derived. Apparently they were biologically linked with the central European Corded Ware horizon. And in fact the Corded Ware had its strong presence in Moravia too, it wasn't just Vucedol material that has been found there.

And then, as I said, the Gata-Wieselburg crania suggest admixture from central European Bell Beakers, while the Nagyrev culture in central Hungary was based on the Bell Beaker Csepel group. So at any rate there wasn't just a one way vector from Hungary to central Europe, but also one into the opposite direction. And since the Unetice culture, which was IE for sure, was related with the Corded people, and the EBA in southern Germany had strong roots in the south German Bell Beaker groups, I'm now prefering the scenario that all western IE languages, including Italic and Celtic, are rooted in the Corded Ware area. That's where I'd put my money on.

@ Rob

Yes, we certainly don't have DNA from Hungarian kurgans, but IMO the autosomal evidence of Mako and Vatya give reliable clues, because what else should have caused the change from Baden (CO1) to Mako (BR1) and Vatya (Allentoft samples) if not the Hungarian kurgans? It has to be kept in mind that actually Mako isn't even real Bronze Age, at least not in the German terminological system, it's late Chalcolithic. So there wasn't any invasion from southeastern Europe that might have caused the change.


Rob said...

Simon
Indeed the Mayko etc genomes are good autosomal proxies, but maybe not for Y markers. Because Mako and Vucedol were probably just Balkan Copper Agers who've shifted a little eastward due to marriages etc with eastern pastoralists. I'd bet that the Yamnaya guys in the west (ie Hungary, Servia, etc) were L51.

Ie L51 made little inroad into the Balkans because the Yamnaya only occupied certain niche areas of the Balkans before retreating back to the steppe or moving westward c2500 BC when conditions dried up.

Whatever the case I accept it's equally plausible that L51 came via CWC; or both !

Simon_W said...

Yep, the main rationale for connecting Yamnaya with Bell Beaker is the R1b, and undoubtedly the fact that they all had R1b (at least in the above mentioned area) can't be a coincidence, but is testimnony to a common origin, otherwise shown by the cranial affinities among eastern Bell Beaker groups. The future has to show where exactly that origin has been; I doubt it was in Hungary, but we'll see! (hopefully)