search this blog

Tuesday, December 11, 2018

How did Y-haplogroup N1c get to Bolshoy Oleni Ostrov?


Y-haplogroup N1c probably entered Europe from Siberia during the Bronze Age or the Eneolithic period. It first appears in the European ancient DNA record in two samples from a burial site at Bolshoy Oleni Ostrov, in the Kola Peninsula, dated to 1523±87 calBCE (see here). These individuals also harbor significant genome-wide Siberian ancestry, but it's possible that this is in large part a coincidence, and that N1c spread into the Kola Peninsula from the south in a population of overwhelmingly European ancestry.

Crazy, huh? Not really. Consider the qpAdm mixture models below for BOO002 and BOO004, the two males from the Bolshoy Oleni Ostrov site belonging to N1c, and BOO006, a female and the most Siberian-admixed individual from the same site. Although BOO002 and BOO004 show a lot of Nganasan-related and thus Siberian ancestry, they also require significant input from a source closely related to Baltic_BA, a fully European Bronze Age population from the East Baltic region. On the other hand, BOO006 doesn't need Baltic_BA for a successful model.

BOO002_&_BOO004
Baltic_BA 0.124±0.029
EHG 0.406±0.032
Nganasan 0.469±0.017
chisq 10.847
tail prob 0.286316
Full output

BOO006
Baltic_BA 0.065±0.043
EHG 0.265±0.084
Nganasan 0.517±0.033
West_Siberia_N 0.152±0.074
chisq 8.847
tail prob 0.355397
Full output

BOO006
EHG 0.367±0.049
Nganasan 0.544±0.031
West_Siberia_N 0.089±0.063
chisq 9.878
tail prob 0.360451
Full output

Keep in mind that N1c is very common in the East Baltic today in populations with minimal Siberian genome-wide ancestry. Indeed, Latvians and Lithuanians can often be modeled with no Siberian input. Thus, it's likely that by the time N1c arrived in the East Baltic, probably during the late Bronze Age or early Iron Age, it did so with populations with heavily diluted Siberian genome-wide ancestry. Such groups may also have taken N1c north of the Baltic and into the Kola Peninsula.

See also...

On the trail of the Proto-Uralic speakers (work in progress)

85 comments:

Huck Finn said...

Very interesting, much appreciated. Re BOO002 & BOO004, did you try to include WSHG into the model?

Davidski said...

Re BOO002 & BOO004, did you try to include WSHG into the model?

Yes, and it didn't work.

Huck Finn said...

It may well be the case that those gentlemen just did not have any WSHG, however I still wonder if EHG of BOO002 & BOO004 is, in this case, partly just more WHG biased WSHG.

Leron said...

To me it looks like BOO6 is a relatively recent arrival from Siberia while the other 2 have been living in the west for more generations and picked up local genes.

JuanRivera said...

N1c seems to look like it was ultimately brought by a Mongolian-Baikal population, the increased admixture of them in both Baltic samples and Botai explaining the ocurrences of N1c in them. R1a and R1b seem to look like EHG and WSHG (West_Siberia_N), as both of them were found in the former as early as the Mesolithic and found in descendant NE European and Steppe populations, and R1b in Botai, which was still mostly WSHG. It appears that the Mongolian-Baikal admixture went through the taiga, explaining its lack in Steppe and the low-moderate levels of extra admixture in Botai.

JuanRivera said...

The Ymyyakhtakh culture springs to mind, given its timing and expansions both east and west.

Mouthful said...

Does Baltic_IA sample have any detectable Siberian ancestry I wonder? Knowing that's it's currently the earliest sample with N1c from Baltic area that we have so far, before the upcoming Tarand graves samples.

Slumbery said...

@Huck Finn

Under the earlier post I mentioned that all FU groups with significant Siberian ancestry have more BHG than WSHG in nMonte. I used EHG + Sintashta + WSN + Shamanka then, because I focused on Volga-Ural populations. This is how it looks like for the modern Saami (first with Sintashta, then with Baltic_BA):

Saami
EHG: 19.17%
Shamanka_BA: 22.5%
Sintashta_MLBA: 58.33%
West_Siberia_N: 0%

and

Saami
Baltic_BA: 61.67%
EHG: 11.67%
Shamanka_BA: 25.83%
West_Siberia_N: 0.83%

Sintashta probably had some local ancestry that is why there is less Siberian in that test, but the point is: the Saami have pretty much no WSHG ancestry.

Interestingly BOO behaves somewhat differently:

Bolshoy_Oleni_Ostrov
Baltic_BA: 14.17%
EHG: 23.33%
Shamanka_BA: 45%
West_Siberia_N: 17.5%

And the Saami take extra Baikal-like ancestry when modelled with BOO.

Saami
Baltic_BA: 50.83%
Bolshoy_Oleni_Ostrov: 40.83%
EHG: 0%
Shamanka_BA: 8.33%

(Levanluhta_IA behaves similarly to modern Saami.)

Davidski said...

@Mouthful

Does Baltic_IA sample have any detectable Siberian ancestry I wonder?

I can't see any, but this might be due to the low coverage.

Nevertheless, this sample does seem to show affinity to Baltic Finnic populations.

Huck Finn said...

@ Slumbery: If you're happy modelling Saami partly based on BOO, there is WSHG in Saami as there apparently is WSHG in BOO, if I'm right.

Otherwise, I still have difficulties in understanding why all three IA N1c1-samples we have for the being from the areas next to Gulf of Finland lack anything like a more meaningful Shamanka_BA element if N1c1 is a late immigrant from Baikal area. Admittedly the samples are not yet available but the communication by the teams has otherwise been very clear.

Besides, it is very difficult to connect the European N1c1 clades to BA Baikal by looking at N1c1 tree.

Them meee said...

I’m still kinda confused. I mean, the Baltic_BA-like ancestry is from around Estonia right? And the N1c group upon their arrival was mostly EHG-like by then or nah?

Davidski said...

@Them meee

Baltic_BA is from Latvia and Lithuania. But this doesn't matter, because there were probably similar populations living all over western Russia as well at the time.

Also, based on the models I posted, there are lots of possibilities for the type of population that may have brought N1c to Bolshoy Oleni Ostrov from the south.

You can basically pick any ratio of Baltic_BA, EHG and Nganasan, and you might be right.

But I think the most realistic solution is a population very similar to the early N1c Finnics from the Tarand graves in Estonia, except with more EHG and Nganasan-related ancestry.

They probably moved at around the same time into the East Baltic, where they lost most of their EHG and Siberian ancestry, and up north, where they gained it. And yeah, they were probably Uralic speakers, so I think it's likely that some lost Uralic language was spoken at Bolshoy Oleni Ostrov.

Toby_P said...

It seems people (outside this blog) are increasingly finding the idea of modelling the expansion of FU groups based on Nganasan as problematic. These are drifted hunter-gatherers, and might represent grops which language-switched to FU. Otherwise, all the Finno-Permic side of FU are clearly western, but with some variable eastern admixture which trickled down.

Davidski said...

@Toby_P

It doesn't matter that Nganasans are "drifted hunter-gatherers". This doesn't affect the results from qpAdm. And you can see that because they're interchangeable with Glazkovo_EBA in my models.

And if you're referring to Carlos when you say people outside of this blog, then I couldn't care less what he says. See here...

Indo-European crackpottery

Anthony Haken said...

@JuanRivera

N-F2905 is the only modern clade of N1c1 common amoung the people of lake baikal or Mongolia and even then it isn't found at more then 5% in the latter. N-F2905 is also clearly younger then its western counterparts.

There was only one BHG positive for N1c1 and it was one of the younger, more western Ust-Ida samples and tested negative for N-L708 (up stream of L1026).

The rest of the BHG were N-L666 or N-L729* so they diverged about 15900ybp
from the ancestors of the western line according to yfull.

Davidski said...

It's very unlikely that the Proto-Uralics formed anywhere in or near Mongolia.

But it looks certain that they were rich in N1c and had significant levels of Siberian ancestry of the Nganasan type.

JuanRivera said...

Uralic's homeland looks like it was in the taiga of western Siberia.

Anthony Haken said...

@JuanRivera
As I and others have said, the Tarand graves came from the Netted Ware cultures of the Volga which in turn were derived from Volosovo/Garino-bor. Therefore the bronze age Volga-Kama is where we should look for PU. If some sort of BOO like Siberians spread N1c to the Baltic we would see higher Nganassan like ancestry in IA samples.

PPU was probably from Neolithic West Siberia and the ancestors of Samoyedic speakers probably spread east quite early.after PU.

Davidski said...

Here's a freely available PDF that might be useful for some background reading...

The problem of Samoyed origins in the light of archaeology: On the formation and dispersal of East Uralic (Proto-Ugro-Samoyed)

Samuel Andrews said...

David can you add the new Neolithic samples to G25

Davidski said...

I can't because it's just gigs of BAM files. It would take hours to do it properly.

There are Neolithic and GAC samples from East Central Europe in the Global25 datasheets anyway.

Kristiina said...

I noticed that T2d1b1 which was detected in Bolshoy is now detected in Gulyukovo and Bolshie Tigani burials in Volga Ural and somewhat east of it in the areas where Hungarian roots were traced in this new paper:
MATERNAL GENETIC COMPOSITION OF EARLY MEDIEVAL (6TH-10TH CENTURY AD) POPULATIONS LIVED IN THE CIS- AND TRANS-URAL AND VOLGA-KAMA REGIONS

Bolshoy is older than Gulyukovo and Bolshie Tigani burials. However, we can presume that there was a continuity of T2d1b1 and N1c in Volga Ural from the Bronze Age to the Middle Age.

Davidski said...

Yeah, I think it's pretty clear that N1c didn't hop over from Siberia to Bolshoy across the Arctic, but moved there from the Ural region from the south, probably along with mtDNA T2d1b1.

Baltic_BA in my model is a proxy for some of the ancestry from the Volga-Kama region and from populations along the way.

mono said...

Davidski

There are a lot off talks about Nganassans, could You please clarify what is Nganassan from a genome wide perspective.
Are they similar to Glazkovo? Do they have much Afontova gora or that Paleosiberian (Kolyma M)

Davidski said...

@mono

Nganasans are extreme Siberians, and similar in this respect to Shamanka_EN, but with some minor EHG.

That's probably why they're interchangeable in my models with Glazkovo_EBA (aka Shamanka_EBA), because Glazkovo_EBA also has some minor western ancestry.

And Nganasans are the easternmost Uralic speakers along the Uralic cline (the black stars farthest to the left).

The Uralic cline in the Global25

All Uralic speakers, except maybe most Hungarians, have Nganasan-like ancestry, which suggests to me that the Proto-Uralics living in the Volga-Kama region had some of this type of ancestry. I don't know how much though. I'll make up my mind when I see Bronze Age and Iron Age samples from the Volga-Kama.

M. Myllylä said...

@Davidski,

"I can't because it's just gigs of BAM files. It would take hours to do it properly."

Using Schiffel's tools, like researchers today do, it doesn't take hours. But you have to ensure that the data is cleaned of duplicates and the clipping is right. Clipping is a kind of state of art and depends on the sample quality. I have seen around here sample bundles made without clipping and doing it is essential. Downloading can take hours, though.

Ebizur said...

According to Malyarchuk et al. (2010), "Mitogenomic Diversity in Tatars from the Volga-Ural Region of Russia," there are only two potential members of mtDNA haplogroup T2d1b1 (or T2d in general) among their Volga Tatar samples (n = 126 Mishars from Buinsk in western Tatarstan and n = 71 Kazan Tatars from Aznakayevo in eastern Tatarstan). None of their 126 Mishars belongs to T2d, but two of their Kazan Tatars who have been predicted to belong to mtDNA haplogroup T potentially may belong to the T2d or T2d1b1 subclade, although they have not been reported to exhibit the C16296T mutation that is listed as a marker of haplogroup T2 on phylotree. (Note that one of the Mishars whose mtDNA has been fully sequenced and assigned to haplogroup T2b* also has not been reported to exhibit any mutation at 16296.)

So, there is a total of 0/126 = 0% T2d in one sample of Mishar Tatars and a potential maximum of 2/71 = 2.8% T2d in one sample of Kazan Tatars. These populations currently inhabit the very center of the Volga-Kama river basin. What were the frequencies of T2d1b1 in this region like before the hypothetical immigration of Turkic speakers?

On the Y-DNA side, a significant percentage of the Volga Tatars (17/133 = 12.8%) belong to N-Z1936 (TMRCA 4,400 [95% CI 3,600 <-> 5,200] ybp according to YFull), which is very common among Finns (16/39 = 41.0%), Vepsa (13/40 = 32.5%), northern Russians (24/106 = 22.6% Arkhangelsk Russians, 20/103 = 19.4% Pinega Russians), Karelians (28/139 = 20.1%), Saami from Sweden (7/35 = 20.0%), and Bashkirs (85/568 = 15.0%) according to Ilumae et al. (2016). It was also found in 10% (3/30) of northern Komis from Izhemsky District, 9.8% (5/51) of Russians from Kostroma Oblast, 9.5% (4/42) of Nenets, and 7.9% (5/63) of Khanty/Mansi.

On the other hand, N-VL29 (TMRCA 3,600 [95% CI 3,000 <-> 4,200] ybp according to YFull), which is common among Lithuanians (44/111 = 39.6%), Latvians (33/86 = 38.4%), Estonians (65/235 = 27.7%), Russians from Pskov Oblast (24/132 = 18.2%), Saami from Sweden (6/35 = 17.1%), Karelians (21/139 = 15.1%), Nenets (6/42 = 14.3%), Russians from Arkhangelsk (14/106 = 13.2%), Finns (5/39 = 12.8%), Maris (6/51 = 11.8%), Komis from "Syktyk distr.," perhaps Syktyvkar (2/17 = 11.8%), Russians from Belgorod Oblast (13/143 = 9.1%), Belarusians (21/267 = 7.9%), and Ukrainians (39/566 = 6.9%), was found in only 5/133 = 3.8% of Volga Tatars. N-VL29 is notably present among Scandinavians and Poles, too.

Apparently, significant percentages of Saami in Sweden, Finns, and Karelians belong to each of N-Z1936 and N-VL29, whereas nearly all Estonian members of haplogroup N belong to the N-VL29 subclade (like Latvians, Lithuanians, Russians in neighboring Pskov Oblast, Belarusians, etc.). Ilumae et al. (2016) found N(xVL29), all of which was N-Z1936, in only 7/235 = 3.0% of Estonians. According to YFull, the MRCA of N-VL29 and N-Z1936 is N-L1026 (TMRCA 4,800 [95% CI 4,100 <-> 5,500] ybp), which also subsumes the common Chukotkan (N-B202) and Buryat/Mongolian/Turkestani (N-F4205) subclades of haplogroup N-Tat.

Which (if any) SNPs downstream of TAT are exhibited by the BOO male specimens?

Huck Finn said...

@ Ebizur: N-L392 i.e. N-1026, TMRCA 4800 years bp. Unluckily it seems that the quality of the samples does not enable any further clarification of the sublade, at least according to some members of the team.

Kristiina said...

It is a pity that, apart from T2d1b1, the new ancient mtDNA paper does not specify the other mtDNA haplotypes of the early medieval Volga Uralians (https://ri.btk.mta.hu/archaeogenetika/images/kazan.pdf).

However, Bolshie Tigani cemetery in Volga Ural and Gulyukovo cemetery in Tatarstan both yielded T2d1b1, and on the basis of their mtDNA, they cluster together in the Ward clustering.

These two sites with T2d1b1 are important in the history of Hungarians. The following is from the new paper:

The Bolshie Tigani cemetery in Tatarstan is one of the most famous sites connected with the ancient history of Hungarians. The early medieval cemetery lies on the left bank of the Santala river running near the Kama river. … Most researchers now believe that the cemetery started to be used in the second half of the 9th century, thus the site is likely connected to the Hungarians stayed in the east, not to those who migrated to the west. Although Bolshie Tigani is classified as a site of the Kushnarenkovo-Karayakupovo culture, its different geographical situation together with its material characteristic of the Volga Bulgars distinguish Bolshie Tigani from the typical Bashkir sites of the Kushnarenkovo culture. The partial horse burials, specific sabres, belt mounts and jewelry as well as the use of death masks and shrouds that are burial habits with Uralic origin, all represent connections with the heritage of the Hungarian Conquest Period.

The material of Gulyukovo cemetery of the culture was investigated. … The cemetery shows connections with the supposedly Ugric, semi-nomad population of the Ural area, therefore majority of the archaeologists in the region identify this material as the heritage of the Hungarians who stayed in the east and had later been found by Friar Julian in the 13th century. This assumption can be reinforced by the presence of kurgan burials, shrouds placed on the eyes and mouths, handmade pottery with stamped decoration and the sporadic remains of partial horse burials. This material appears in the Trans-Ural region as well.

Davidski said...

Here's a qpAdm mixture model for the Bolshoy Oleni Ostrov woman with T2d1b1. This is just about the only one that works.

Sintashta_MLBA_o1 are one of the eastern shifted outlier Sintashta groups.

I'm seriously thinking that Bolshoy Oleni Ostrov was partly settled by Uralic migrants from the Volga-Kama. So expect plenty of N1c in Bronze Age samples from this area.

BOO003
Latvia_MN 0.278±0.030
Nganasan 0.405±0.028
Sintashta_MLBA_o1 0.318±0.044
chisq 17.739
tail prob 0.0595347
Full output

Slumbery said...

Huck Finn
I am not "happy" modelling Saami with BOO. I did that test just from out of curiosity for this particular comparison and never claimed it to be a particularly good model.
Actually the fact that Saami show about an order of magnitude less WSHG than they should have if they are half BOO, suggests that this is not a good model.

Huck Finn said...

@ Slumbery: apologies, my misunderstanding.

@D: Re Sintashta_MLBA_01, which one did you use or maybe all of them? Related to this, some old models of Shaikorth:

Sintashta_MLBA_o1:I0983

Sidelkino:Sidelkino 35.05
CHG 22.70
West_Siberia_N 21.40
AfontovaGora3 9.90
Ganj_Dareh_N 5.85
Barcin_N 5.10
ShamankaEN:DA249 0.00

"distance%=5.0196 / distance=0.050196"

Sintashta_MLBA_o1:I1007

West_Siberia_N 47.4
AfontovaGora3 35.1
CHG 9.7
Ganj_Dareh_N 7.8
Sidelkino:Sidelkino 0.0
Barcin_N 0.0
ShamankaEN:DA249 0.0

M. Myllylä said...

@Kristiina

T2d1b1 is extremely rare in Finland, you can't find even single T2d in FtDna's Finland project among almost 4000 samples. If T2d was something in the history of Hungarians, it seems to mean nothing in Finland.

Parastais said...

Most likely Proto-Uralic at its latest formation stages lived nearby a NW IE speaking tribe (s) and PII speaking tribe(s). Judged by the early very archaic loanwords from those sources.
Also I believe some gene share took place, I mean a word for “woman” in Proto-Uralic is from PIE derived gwen something (with for example Russian cognate женщина), so some mixing must have happened :)
So Proto-Uralic person would look like main part ? + CW like adstrate.. before they spread all directions.

Anthony Haken said...

There is obviously no way to prove it without aDNA but I think the earliest expansions of N-L1026 can be explained by Seima-Turbino. Especially in the east but also west.

I think N-Y16323 (up stream of N-F2905 and N-B202) can be explained by Seima-Turbino in the Altai/Mongolia region, migrating north and undergoing a bottleneck explaining its modern distribution in East Siberia.

Netted Ware and Seima-Turbino artefacts also spread hand in hand into Fennoscandia as early as 1900BC which could have implications for Bolshoy Oleni Ostrov as well.

Kristiina said...

@ M. Myllylä "T2d1b1 is extremely rare in Finland"

Yes, inexistant in Finns and Saamis. Considering what Slumbery said about the lack of WSHG in Saamis, this is another reason not to think that Siberian in Fennoscandia must be from Bolshoy.

M. Myllylä said...

@Kristiina,

it is very possible that 3500 years ago migrations existed between Kama and Kola, we know that similar south-to-north migrations happened later in Scandinavia and Finland. But I need evidences to be convinced.

Davidski said...

@Parastais

I get what you're saying about the Indo-European, including Corded Ware, linguistic influences in Uralic.

But I wouldn't overemphasize the Corded Ware genetic contribution to the Proto-Uralic population.

That's because the Corded Ware people who migrated back east were already heavily admixed, and their descendants continued mixing with the locals in the forest steppe.

There were other West Eurasian populations, both mostly of forager and steppe origin, living around the Volga, and some of them probably had a much more profound impact on the Proto-Uralic gene pool.

epoch said...

@Kristiina

I know you once suggested PIE could have been a mixed language based on the fact that it had so many roots for the same concept. However, I recall Agamemnon on AG stating that the almost mathematical way PIE was structured argued against it, as mixed languages we know have different ways to deal with both parts of the mixture - Pardon me if I explain this badly, btw.

How about Proto-Uralic being a mixed language, with an Indo-Uralic part and something else? Is that an option?

EastPole said...

I have combined David W. Anthony’s map of MBA cultures with Ilumae’s (2016) distribution map of hg N3a1 which seems to be a sister clade of N3a2’6 from which N3a3 and N3a4 originated:

https://i.postimg.cc/DzjP3Zrb/screenshot-463.png

It seems to me that most probably N3a3 entered Northern Balto-Slavic tribes via Fatyanovo culture:

https://i.postimg.cc/G2zwCvMx/Uralic2.png

Them meee said...

But what is the ultimate source of T2d1b1 in Uralic populations?

Davidski said...

@Them meee

But what is the ultimate source of T2d1b1 in Uralic populations?

Neolithic farmers to the west of the steppe.

But the fact that there's no T2d in any Corded Ware samples suggests that this mt-hg made it east via some other way than the Corded Ware expansion.

Davidski said...

@Anthony Haken

There is obviously no way to prove it without aDNA but I think the earliest expansions of N-L1026 can be explained by Seima-Turbino. Especially in the east but also west.

I'm thinking that Seima-Turbino might also be one of the main explanations for how the East Asian-like Baikal Hunter-Gatherer (BHG) ancestry became more important in Uralic speakers than West Siberian Hunter-Gatherer (WSHG) ancestry.

It seems to me like WSHG populations were pummeled from both the east and west, even if they were the ultimate source of N1c in modern populations.

Davidski said...

By the way, a real effort must be made to collect samples from the best Proto-Uralic candidate cultures, like Garino-Bor and even Volosovo, rather than from the ends of the early Uralic expansion, like the Baltic Tarand graves.

It's likely that a few good quality genomes from such burial sites from around the Urals would pretty much wrap up the Proto-Uralic controversy overnight.

M. Myllylä said...

@Davidski

"By the way, a real effort must be made to collect samples from the best Proto-Uralic candidate cultures, like Garino-Bor and even Volosovo, rather than from the ends of the early Uralic expansion, like the Baltic Tarand graves.

It's likely that a few good quality genomes from such burial sites from around the Urals would pretty much wrap up the Proto-Uralic controversy overnight."

Actually I don't exactly understand what we are searching for. How we can link the Proto-Uralic language and their genes if we don't try to connect them to the later existence of Uralic languages, to proven Uralic speakers, like Tarand graves, Hungarians, Finns, or even to Maris and Udmurds. It is impossible to see the spoken language in genes, if we don't try to track them from known speakers. Regarding IE-languages the situation much easier, because we have much older written evidenses.

Davidski said...

@M. Myllylä

I've got plenty of Uralic samples in my dataset. I'm sure you do too. So comparing these samples to those from, say, Garino-Bor might be very enlightening.

By the way, I wasn't implying that the samples from the Baltic Tarand graves aren't useful, but such samples won't be of much help by themselves because they're too distant from the Proto-Uralic homeland in space and time.

Them meee said...

Off topic but what is right now the most likely origin of I1?

Kristiina said...

@ epoch
Yes, a protolanguage spoken in a small area in a relatively simple society cannot have three-five words for such basic concepts as fire or water and even 15 words for binding, and more so, if taken into account that these roots are usually shared with non-IE languages in specific sub-areas. Therefore, a part of the lexicon may have enterd the IE languages from the early areal substrate languages.

Comparative method is based on logic, i.e. forms are reconstructed from daughter languages with a kind of an extrapolation. The method itself makes the result what it is - logical. It is more difficult to reconstruct illogical idiosyncratic features of a protolanguage spoken c. 7000 years or more ago or its immediate daughter languages, although such forms may have existed.

Kristiina said...

We have some ancient T2d samples that reflect the distribution of T2d haplotypes and give hints of its area of origin:
BMAC BA Gonur Turkmenistan Gonur1 I1792/Gonur1_BA_o(utlier) 2350 BCE T2d1a; Geoksiur Late Eneolithic Turkmenistan I8504/S8504.E1.L1 4000-2000 BCE T2d2;
BA Bolshoy Kola Peninsula BOO003/BOO72-4 T2d1b1

Modern mtDNA:
T2d1b Iran persian; T2d1b Haryana Brahmin Eastern Punjab; T2d1b Mordvins; T2d1a (16086C) India; T2d1b1 Khanties, Komis
T2d2 Iran; Pamir, Tunisia, Italy; Debrecen Hungary

Davidski said...

OK, so T2d isn't from farmers living west of the steppe. It's probably from Iran, and made it into the Volga-Ural and the Uralic gene pool via BMAC and Steppe_MLBA.

And then, with the Uralic expansion it went as far north as the Kola Peninsula. Never would've picked that.

Toby_P said...

@ Them meee

''Off topic but what is right now the most likely origin of I1?''

I have seen one one hypothesis recently. But too few data points to confirm anything about I1 at this stage, IMO.

Ebizur said...

T2d2 may have a somewhat more westerly distribution than its sister, T2d1:

MtDNA Haplogroup T2d
TMRCA 13,584.7 [95% CI 9,647.6 <-> 17,521.8] ybp (Behar et al. 2012)

T2d1
Mongol [T2d1+C194T], Peru, USA

T2d1b
Wakhi (Tashkurgan)

T2d1b1
Khant

T2d1b2
Persian (Iran)

T2d1a
India

T2d2
Italy (incl. Toscani), Switzerland, Hungary (Debrecen region), Turkey, Bahrain, Tunisia, Iran, Pamiri (Gorno-Badakhshan, Tajikistan), USA

The distribution of T2d1* is curious. Mongol, Peru, USA...

M. Myllylä said...

@Davidski,

who are present-day Uralic people and where they live? The answer is next to Siberians. It is not a big surprise that people who have always lived nearby Siberians have Siberian admixture, no matter do they speak Uralic, Turkic or Sino Tibetan languages.

Linguists say that Tarand graves belong to Baltic Finnic speakers.

We should not be into cherrypicking. Saying this, I want to remind that alternatives are to strenghten the rule, not that I have made up my mind.

Davidski said...

@M. Myllylä

Who are present-day Uralic people and where they live? The answer is next to Siberians.

This is a misleading statement, because there's a lot more to this. The same admixture signals and uniparental markers from around the Urals are found in practically all Uralic speakers, and the Proto-Uralic homeland was located by linguistics and archeologists in the southern Urals even before anyone knew about N1c or Nganasan-related admixture.

It's very likely that ancient DNA from burials from the best Proto-Uralic candidate cultures will confirm that the Uralic homeland was located in the southern Urals.


M. Myllylä said...

@Davidski,

actually Finns, Karelians, Komis, Ingrians and most western FU speakers also live next to Siberian connected people who practice same livelihoods with Siberians. New Iron Age finds prove it. Estonians were during the historic time a target of large Finnish migration, which can be read from history books, no matter what Estonian geneticists say. Church registers are undeniable. Even 25% of the peasants in the North Estonia were Finns, of the artisans even more, at the time of Swedish military campaigns.


Prot Uralic home land is a linguistic paradigm. It can be true or not true, in genetics researchers have to follow own methods.

Davidski said...

@M. Myllylä

Prot Uralic home land is a linguistic paradigm. It can be true or not true, in genetics researchers have to follow own methods.

Well, obviously, genetics can corroborate or contradict a linguistic theory, and this can be an important addition to a multidisciplinary study.

If, in this case, ancient DNA shows a series of expansions from the southern Urals dating to the Bronze and Iron Ages that match the linguistic theory that the Proto-Uralic homeland was located there, then there won't be anything to debate other than the details.

Them meee said...

@Toby_P

So I see the article suggests I1 came from Unetice alongside R1b-U106.

Couldn’t I1 have been present in a Funnelbealer population and get absorbed by Corded Ware migrants? I’m not sure how more likely is it that it arrived later with a migration of Beaker offshoot people alongside R1b.

Them meee said...

Also isn’t R1a significantly more common in Estonia than Finland, and viceversa for I1? Similarly aren’t Estonians way more Baltic-like? Wonder what that means for that Finnish migration. If I’m missing any details (maybe they left or something) then feel free to correct me or add something.

Tesmos said...

@Them meee

The author of the article seems to forget that there was already an R1b-U106 sample in Scandinavia (RISE98) before the arrival of Unetice, plus there is no sign of R1b-U106 in Bell Beakers yet.

M. Myllylä said...

@Davidski,

"If, in this case, ancient DNA shows a series of expansions from the southern Urals dating to the Bronze and Iron Ages that match the linguistic theory that the Proto-Uralic homeland was located there, then there won't be anything to debate other than the details. "

If ancient DNA shows an expansion from the southern Urals to areas of Northern European FU speakers we can start the debate :)

M. Myllylä said...

@them mee

"Also isn’t R1a significantly more common in Estonia than Finland, and viceversa for I1? Similarly aren’t Estonians way more Baltic-like? Wonder what that means for that Finnish migration. If I’m missing any details (maybe they left or something) then feel free to correct me or add something."

The Finns are less Baltic like because of more Scandinavian like I1 and Saami like ancestry in Finland. Finnish expansions during and after Swedish military campaigns to the Baltic region are totally ignored by Estonian geneticists. Older Soviet era Estonian history books still tell about it, and of course Finnish tax and church registers. Maybe rising Estonian spirit is the reason.

M. Myllylä said...

@them mee,

forget to mention that northern Estonians get usually 20-40% Finnish ancesry in autosomal tests, made by third parties like 23andme. There is of course explanations for certain unreliability, but f.ex. 23andme tells that their tests are valid only in genealogical time frame (max. 500 years).

Ebizur said...

Them meee,

Even the frequencies of subclades of N-L1026 (TMRCA 4,800 [95% CI 4,100 <-> 5,500] ybp) appear to differ between Finns and Estonians as I have noted previously in this comment thread:

N-VL29
44/111 = 39.6% Lithuanians
33/86 = 38.4% Latvians
65/235 = 27.7% Estonians
24/132 = 18.2% Russians (Pskov)
6/35 = 17.1% Saami (Sweden)
21/139 = 15.1% Karelians
6/42 = 14.3% Nenets
14/106 = 13.2% Russians (Arkhangelsk)
5/39 = 12.8% Finns
6/51 = 11.8% Maris
2/17 = 11.8% Komi (Syktyk district)
13/143 = 9.1% Russians (Belgorod)
21/267 = 7.9% Belarusians
39/566 = 6.9% Ukrainians
3/45 = 6.7% Russians (Kursk)
6/100 = 6.0% Russians (Smolensk)
5/110 = 4.5% Russians (Oryol)
4/93 = 4.3% Polish
4/103 = 3.9% Russians (Pinega)
5/133 = 3.8% Volga Tatars
2/54 = 3.7% Dolgans (Taymyr)
3/96 = 3.1% Russians (Voronezh)
8/277 = 2.9% Crimean Tatars
3/114 = 2.6% Chuvashes
1/40 = 2.5% Vepsa
1/43 = 2.3% Selkups
1/51 = 2.0% Russians (Kostroma)
2/119 = 1.7% Karanogays
1/60 = 1.7% Komi Permians
3/221 = 1.4% Siberian Tatars
2/203 = 1.0% Mordva
4/568 = 0.7% Bashkirs

N-Z1936
16/39 = 41.0% Finns
13/40 = 32.5% Vepsa
24/106 = 22.6% Russians (Arkhangelsk)
28/139 = 20.1% Karelians
7/35 = 20.0% Saami (Sweden)
20/103 = 19.4% Russians (Pinega)
85/568 = 15.0% Bashkirs
17/133 = 12.8% Volga Tatars
3/30 = 10.0% Komi (Izhemsky district)
5/51 = 9.8% Russians (Kostroma)
4/42 = 9.5% Nenets
5/63 = 7.9% Khanty and Mansi
4/54 = 7.4% Dolgans (Taymyr)
3/45 = 6.7% Russians (Kursk)
1/17 = 5.9% Komi (Syktyk district)
12/221 = 5.4% Siberian Tatars
3/60 = 5.0% Komi Permians
3/60 = 5.0% Mordva
2/50 = 4.0% Nganasans
5/132 = 3.8% Russians (Pskov)
4/114 = 3.5% Chuvashes
3/96 = 3.1% Russians (Voronezh)
7/235 = 3.0% Estonians
4/203 = 2.0% Mordva
1/86 = 1.2% Latvians
6/566 = 1.1% Ukrainians
1/100 = 1.0% Russians (Smolensk)
1/119 = 0.84% Karanogays

Data from Ilumae et al. (2016).

Them meee said...

@Tesmos

Yeah, that said I’m skeptical of any R1b-L151 in Corded Ware which is almost totally R1a-M417 and looking very Eastern IE while Germanics seem a bit more Western so R1b-L151/U106 May have something to do with that.

M. Myllylä said...

@them mee,

is it that L1026 includes Karelian and Savonian branches? If so, the explanation of Finnish high frequences is so called Savonian expansion, which happened only 500 yearsago.

http://weallfinland.fi/wp-content/uploads/2016/12/24.11.2016-Markku-Mattila.pdf

Actually the Baltic branch, L550, is older in Finland than in Baltic countries. I am not expert in this question, you or other readers can correct me, I'll search more information if needed.

Davidski said...

@M. Myllylä

We already know that ancient DNA shows an expansion of N-L1026 into Estonia during the Iron Age at the latest, along with the Tarand grave culture.

This is when N-L1026 starts replacing the R1a-M417 in the region, one way or another. So the recent expansions of various N-L106 subclaces in the East Baltic and Fennoscandia as inferred from modern DNA aren't all that relevant.

The Tarand grave tradition is derived from the ancient cultures of the Volga and Urals that have been posited to have been early Uralic-speaking cultures, like the Garino-Bor culture.

So it's pretty much all over now. But basically what we're still waiting for is confirmation from ancient DNA of the link between the Tarand graves and Garino-Bor. When that comes, that's when the debate ends.

Mouthful said...

@M. Myllylä "Actually the Baltic branch, L550, is older in Finland than in Baltic countries."

Basal is found in Swede and in general Baltic N1c falls under N-L1025 which is downstream from N-L550.

https://www.yfull.com/tree/N-L550/

Davidski said...

Modern DNA can be very misleading indeed.

Mouthful said...

Indeed it can, but the TMRCA of Baltic clades specifically Latvian and Lithuanian ones coincides with the N entry into there which is Iron age, which we know was absent during Bronze age.

Mouthful said...

Also sorry for posting again, but forgot to include this in previous post, here's an image of modern N clade distribution among various populations.

https://imgur.com/a/fSPuAkA

M. Myllylä said...

@Davidski

"This is when N-L1026 starts replacing the R1a-M417 in the region, one way or another. So the recent expansions of various N-L106 subclaces in the East Baltic and Fennoscandia as inferred from modern DNA aren't all that relevant."

I agree every word. This explains my lack of interest in ydna haplogroups without simultaneous auDna support.

Anthony Haken said...

@M. Myllylä

Like Davidski said the Volosovo/Garino-Bor cultures were candidates for PU well before aDNA. This is based on the need for Indo-Iranian loanwords in PU. The Tarand graves are connected to these cultures by the Netted Ware culture which is the bridge between Estonia and the Volga-Kama you are looking for.

Also I know without aDNA it isn’t too relevant but L550 and certain sub branches do seem to be basal in Sweden. It may not be a coincidence that Mälaren Sweden lends its name to the Akozino-Mälar axes originating from the IA mid Volga.

M. Myllylä said...

@Anthony Haken,

"Like Davidski said the Volosovo/Garino-Bor cultures were candidates for PU well before aDNA. This is based on the need for Indo-Iranian loanwords in PU. The Tarand graves are connected to these cultures by the Netted Ware culture which is the bridge between Estonia and the Volga-Kama you are looking for."

I do agree, I have not disagreed.


"Also I know without aDNA it isn’t too relevant but L550 and certain sub branches do seem to be basal in Sweden. It may not be a coincidence that Mälaren Sweden lends its name to the Akozino-Mälar axes originating from the IA mid Volga."

You know that the basal is in Sweden, but you also know that Finland follows Sweden in this matter. You have to know, because this is readable from same sources. Why it is so hard to say?

Tesmos said...

@Them meee

Sure, there is overwhelming evidence that P312 is strongly associated with Western IE, based on the results of Bell Beakers and it's successors (and R1a being associated with Eastern IE). However, this cannot be said for U106 as there is still no sign of U106 in CWC and BB, but it does show pop up in the Nordic Late Neolithic, Czech Unetice and Bronze Age Netherlands cultures. So yeah, it's quite puzzling.

Them meee said...

@Tesmos

True, but U106 isn’t associated with Eastern IE either. And somethings got to explain that northwest European pull.

That said, there is a certain presence of P312 and seemingly unique subclades in Germanic peoples, so that is also part of the puzzle.

Germanic seems intermediary but it has R1a-M417 and clear Corded Ware ancestry so I don’t think it’s very likely R1b-U106 came from a completely unique migratory movement rather than being derived from a Beaker offshoot or a group closely related to them. I’d say we need more samples from around the South Baltic and Elbe to start solving this mystery and follow the trail that leads to U106.

Toby_P said...

@ Tem Mee/ Tesmos

No I didnt udnerstand it to read that U106 or I1 necessarily came directly from Unetice, but during that period (the post Corded', 2200 ->) previously rare lineages came up, and it proposes some connections/ mechanisms why. You could be right that I1 is from FB/ TRB, but it hasnt turned up there so far AFAIK. IMO the earliest split of I1 is the Finnish branch, so Im not sure I agree with the suggested path, but there is only one pre-Bronze Age I1 found so far (LBK), and one possible ''pre-I1"" in a mesolithic sample from Stara Forvar, acc. to some posters.


''The author of the article seems to forget that there was already an R1b-U106 sample in Scandinavia (RISE98) before the arrival of Unetice, plus there is no sign of R1b-U106 in Bell Beakers yet.''

There is no Unetice culture in Scandinavia in anycase. But the Scand ''LN'' dates to same absolute as the start of the ''Bronze Age'' central Europe (a/w Unetice), that was kind of the crux of the post (diferent terminologies by region).
Also, have a look at BB Hungary I4178. Therefore the sketch proposed is BB/ LVA Hungary 2500 -> Unetice & LN Sweden 2200 .

Toby_P said...

Coincidentally, the MBA Netherlands U106 also connects there, with the post-Beaker changes. Some guy called 'Finn'' on AG posted a good entry about the Elp culture and its southeastern connections.

Tesmos said...

@Them meee

Agreed.

@Toby_P

I4178 is not confirmed as R1b-U106. Alex Williamson,the owner of ytree.net, mentioned that he has not any confidence in calling I4178 as R1b-U106, he is not even sure this specific sample belongs to R-L51.

''
I don't have any confidence in calling I4178 as R-U106 > Z18. There's only one read and it is a G > A mutation at the last position. This sort of mutation is typical of damaged aDNA. I can provide you a list of a dozen other calls for this kit that are downstream of L51 that are also supported by just one read and they're all from different branches. As it is, I'm not even sure he belongs to R-L51.''

ryukendo kendow said...

ADMIXR, a package for simple one-liner ADMIXTOOLS directly from R

Also, great to see the evolution of your views on Uralic, David.

Davidski said...

@ryukendo kendow

Also, great to see the evolution of your views on Uralic, David.

I don't think they've changed much. I don't expect Garino-Bor/Volosovo samples to be anything like Corded Ware or Steppe_MLBA, unless we're talking about the Sintashta outlier groups.

Bob Floy said...

@Davidski

Goofy question, have you ever done a PCA or anything like that for Sintashta_MLBA_o1?

Davidski said...

@Bob Floy

Sintashta_MLBA_o1 samples are in the Global25 datasheets, so they can be plotted and modeled in any which way.

But yeah, I have featured them in a number of PCA. In the PCA in this post the Sintashta_MLBA_o1 samples are the two that I circled, plus the other two immediately below them (and just above Yamnaya). The two Sintashta samples just left of them are from the Sintashta_MLBA_o3 cluster.

The mystery of the Sintashta people

Bob Floy said...

@Davidski

I see, thanks.

Them meee said...

@Davidski

Just for clarification, how relevant could Sintashta_MLBA_o1 be to later Uralics, if at all?

Davidski said...

@Them meee

Just for clarification, how relevant could Sintashta_MLBA_o1 be to later Uralics, if at all?

It's hard to say. Take a look at the comments under that post.