search this blog

Monday, October 22, 2018

Y-haplogroup P1 in Pleistocene Siberia (Sikora et al. 2018 preprint)

Over at bioRxiv at this LINK. Below is the abstract, emphasis is mine. Two of the (unrelated) males from Yana RHS belong to Y-haplogroup P1 and mitochondrial haplogroup U2. Note that P1 is ancestral to Y-haplogroups Q and R.

Far northeastern Siberia has been occupied by humans for more than 40 thousand years. Yet, owing to a scarcity of early archaeological sites and human remains, its population history and relationship to ancient and modern populations across Eurasia and the Americas are poorly understood. Here, we report 34 ancient genome sequences, including two from fragmented milk teeth found at the ~31.6 thousand-year-old (kya) Yana RHS site, the earliest and northernmost Pleistocene human remains found. These genomes reveal complex patterns of past population admixture and replacement events throughout northeastern Siberia, with evidence for at least three large-scale human migrations into the region. The first inhabitants, a previously unknown population of "Ancient North Siberians" (ANS), represented by Yana RHS, diverged ~38 kya from Western Eurasians, soon after the latter split from East Asians. Between 20 and 11 kya, the ANS population was largely replaced by peoples with ancestry from East Asia, giving rise to ancestral Native Americans and "Ancient Paleosiberians" (AP), represented by a 9.8 kya skeleton from Kolyma River. AP are closely related to the Siberian ancestors of Native Americans, and ancestral to contemporary communities such as Koryaks and Itelmen. Paleoclimatic modelling shows evidence for a refuge during the last glacial maximum (LGM) in southeastern Beringia, suggesting Beringia as a possible location for the admixture forming both ancestral Native Americans and AP. Between 11 and 4 kya, AP were in turn largely replaced by another group of peoples with ancestry from East Asia, the "Neosiberians" from which many contemporary Siberians derive. We detect additional gene flow events in both directions across the Bering Strait during this time, influencing the genetic composition of Inuit, as well as Na Dene-speaking Northern Native Americans, whose Siberian-related ancestry components is closely related to AP. Our analyses reveal that the population history of northeastern Siberia was highly dynamic, starting in the Late Pleistocene and continuing well into the Late Holocene. The pattern observed in northeastern Siberia, with earlier, once widespread populations being replaced by distinct peoples, seems to have taken place across northern Eurasia, as far west as Scandinavia.

Sikora et al., The population history of northeastern Siberia since the Pleistocene, bioRxiv, posted October 22, 2018, doi:

See also...

Ust'-Ishim belongs to K-M526


Davidski said...

Also check out the discussion under the Ust-Ishim blog post...

epoch said...

Chad Rohlfsen stated that ANE was Sunghir + East Asians. He was right. They model ANS as roughly 25% Tianyuan and 75% Sunghir and ANE as descendant.

Santosh Rajan said...

Clearly looking at the distribution of Haplogroup K, we can see that it and some of it's decendants migrated around the tibeten plateau.

Clearly C, K, and H passed through South Asia.

I and G probably formed the basal eurasions and moved northwards from Saudi Arabia or Iran.

J mostly remained in Iran or in the middle east.

Looking at a map most decsendents of K went east out of India.

The other group that went south or east was C. But I am not sure. Maybe C went first and K came later.

What came back around the tibetan plateau was P and its decandants.

Matt said...

@epoch, I don't think they're actually saying it's specifically from Tianyuan and Sunghir (indeed it could not be in the case of Sunghir as actually being the ancestral population to these Siberians, as they are roughly contemporary and the divergences they model are deeper than this). Sunghir's just the high quality "Ancient West Eurasian" Upper Paleolithic samples they chose to use in their qpGraph.

Reconfirming ENA->ANE+WHG lineages (to the exclusion of EUP) is not so new, and really seemed almost certain to anyone looking closely ;) as soon as Kostenki and Ust Ishim were both published, and present in Lipson+Reich's trees and Fu's paper pretty soon after than. It's not saying anything new about Sunghir and ANE specifically.

The interesting bit in the trees seems like the 12% CHG-non-Basal part to Malta_UP, and then the 22% CHG to EHG (should leave CHG as 52% Yana related, 24% WHG related, 16.3% CHG_HG related, 7.5% Basal Eurasian). But of course there are probably multiple viable trees.

Matt said...

Although actually even then, I don't see that they use Tianyuan in the qpGraph.

Arza said...

Sup. page 72
The presence of CHG-related ancestry in EHG predicts that EHG should also contain “basal
Eurasian” ancestry, albeit at a reduced amount compared to CHG. We used f4 -statistics of the
form f4(Mbuti,Ust_UP;EasternHG_M,X) to test this hypothesis, as this statistic is expected to
be positive in the presence of basal Eurasian ancestry that pre-dates the divergence of
Ust’Ishim. We find a weakly positive signal for all test groups X (Table S6.6), suggesting
that EHG indeed harbours basal Eurasian ancestry in a diluted form.

Davidski said...

So the idea of the Mesolithic south > north cline in CHG/EHG ancestry from the Caucasus to Northeastern Europe looks more and more plausible.

But as I said before, I don't expect this cline to be perfectly clinal, probably with CHG-heavy groups between the Black and Caspian Seas giving way sharply to EHG populations with very little CHG.

epoch said...


You're right. This is the part:

"Symmetry tests using f4 statistics reject tree-like clade relationships with both Early West Eurasians (EWE; Sunghir) and Early East Asians (EEA; Tianyuan); however, Yana is genetically closer to EWE, despite its geographic location in northeastern Siberia (Extended Data Fig. 3d, Extended Data Table 2; Supplementary Information 6). Using admixture graphs (qpGraph) and outgroup-based estimation of mixture proportions (qpAdm), we find that Yana can be modelled as EWE with ~25% contribution from EEA (Figure 2; Extended Data Fig. 3; Supplementary Information 6). "

EastPole said...

There is also linguistic supplement “The Formation of the Siberian Linguistic Landscape”by MichaĆ«l Peyrot and Guus Kroonen.

They didn’t go deeper into speculation about Yana people language and religion, only metioned Kortlandt’s and Greenberg linguistic theories:

“On a still deeper level, Kortlandt sees Uralo-Siberian related to Indo-European and to Altaic, according to him consisting of Turkic, Mongolic, Tungusic, Korean and Japonic (2010a, 2010b). A similar so-called macrofamily, “Eurasiatic”, has been posited by Greenberg (2000).”

Linguistic influences and relationships are as interesting as religious and cultural. What was the origin of IE religion and culture and what was it’s relation to Siberian religions and cultures?

There is a well known fact that at the root of Greek and Vedic religion and philosophy lies a shamanistic religion and culture whose origin is variously explained by different people.

Very common explanation is that it came from Siberia.
It is probably true but the question is when and how.

Siberian East Asian genetic influences in Slavic populations, Sintashta “Indo-Iranians” and ancient Greeks are not visible, yet they share many similarities in language, religion and culture and some elements of them can be called shamanistic. So how do we explain it?
I have a theory.

We know that shamanism is a prehistoric tradition dating back to Paleolithic. There are many similarities between Siberian shamanism and North American. Shamans eat special magic mushrooms, have special experiences of soul traveling to heaven and explain the world in a similar way etc.
It is reasonable to assume that Yana people could practice shamanism. Then we have population practicing shamanistic tradition which is not admixed with East Asians.
We can then suspect that R1a EHG deriving from Siberia and distantly related to Yana and Malta people were also practicing shamanism.
When R1a EHG admixed with CHG and became pastoralist some elements of shamanic religion was probably also present among them.
And now the interesting thing:

Siberian shamans as a result of interaction with Slavs switched from eating magic mushrooms to drinking vodka. Why did they do it? Probably because it is safer, healthier, more predictable and when connected with proper rituals, singing and dancing produces desirable effects of contact with loving and caring spirits or gods.

I suspect that the same could happen with our ancestors. In Sredny Stog they interacted with EEF of Tripolye and switched from eating magic mushrooms to drinking alcohol, mead or beer. So some elements of shamanistic believes were preserved but new rituals were invented to reach the goal of contact with loving and caring spirits or gods.

CWC was probably a culture in which rituals involving drinking of alcohol, singing, dancing, experiencing the love God who dwells in heaven and in our hearts were very important. It resulted also in the development of language and poetry.

I wrote about poetry here:

This would explain the links with shamanism and at the same time the existence of many common elements in poetry, religion and language among Vedic Aryans, Greeks and Slavs.

epoch said...

Another very interesting tidbit:

" Intriguingly, a signal of Australasian ancestry that has been observed in some Amazonian groups 24,28 is not evident in any of the ancient Siberian or Beringian samples sequenced here, or in previous studies"

The very interesting fact is that Tianyuan actually *did* have that ancestry:

"No trace of this connection is observed in present-day East Asians and Siberians, but unlike them, the Tianyuan man also possesses genetic similarities to the same South Americans, in a pattern similar to that found for the Papuan and Onge."

Sofia Aurora said...

I find a bit odd that they have not found any non-Sapiens archaic homo contribution.
No Denisovan, no any probable ghost li eage, nothing at all!

And what about Japan? It has the oldest homo sapiens (Ishigaki) in the upper Pacific dated at 27000 BP.
The phenotype of the Japanese homo is almost identical to the much latter Yomon and Ainu.
With the Kennewick man and all should not they also check out these Japanese specimens?

Sofia Aurora said...

I know that Kennewick man's DNA did not show any non-Sapiens or generally unexpected haplogroups etc.

But Kennewick man was not the only Jomon (or Yomon) looking individual that went to the Anericas right!

Thus why not looking at the oldest representatives of that type in case we find something.
Ishigaki should be checked on definetely as soon as possible!

Ebizur said...

Santosh Rajan wrote,

"Clearly looking at the distribution of Haplogroup K, we can see that it and some of it's decendants migrated around the tibeten plateau.

Clearly C, K, and H passed through South Asia.


Looking at a map most decsendents of K went east out of India.

The other group that went south or east was C. But I am not sure. Maybe C went first and K came later.

What came back around the tibetan plateau was P and its decandants."

Are you sure anything about those haplogroups' routes of dispersal is clear?

Haplogroup H seems to have probably reached South Asia very early, perhaps during the initial range expansion of F-M89-bearing AMHs with their Upper Paleolithic industry, but where is the evidence of further expansion beyond South Asia?

What evidence exists that would prove migration of C and K through South Asia as opposed to through Central Asia/East Asia?

Aniasi said...

What is the actual relationship of ANS groups to ANE? I'm also confused by the mention of bidirectional gene flow with Na Dene, when they stated that AP is ancestral to paleo-Americans.

Ric Hern said...

This study for me personally proves yet again that where we see populations of Older Haplogroups today is most probably not where their Ancestors lived and originated thousands of years ago.

For me personally the Caucasus the Western Himalayan/Altai Mountain ranges looks like Very Early diverging points and less so South/Southeast Asia....

Mountainous areas makes sense because it provided protection against predation. Massive predators roamed around during the Ice Age on the Plains mostly...

Ebizur said...

Of six Devil's Gate Cave specimens in this study, three have been assigned to D4m, and the other three have been assigned to D4. The one male specimen has been assigned to Y-DNA haplogroup C2b, which I suppose should be ISOGG's C2b-L1373, the typically northern branch of haplogroup C2-M217. Among modern populations, the D4m2 clade seems to be most common in the Nivkhs of northern Sakhalin. D4m2 and its subclades also have been observed in Kazakhs, a Kyrgyz, Turkmens, Uyghurs, an Uzbek, a Shor, Tubalars, a Tuvinian, Buryats, a Bargut, Evens, Yakuts, an Evenk, and a Yukaghir. The D4m1 clade has been observed in Japanese. D4m has been observed in two Kyrgyzes, a Uyghur, and a Tuvinian.

Of the mtDNA haplogroups detected in the UstBelaya_EBA samples, F1b1b and C4a1a3 tend to be found in Turkic people at present (but also a bit among their neighbors, such as Buryats, Mongols, Tajiks, Nganasans, etc.), while D4j1 has a distribution centered on Tibet but also extending to Mongol, Turk, Indian, and Thai populations.

Yakut or Siberian Tungus x 11
Uyghur x 9
Kyrgyz x 4
Kazakh x 2
Mongol x 2
China? x 1
Turk x 1
Altai-Kizhi x 1
Mongush x 1
Tajik x 1
Uzbek x 1
[F1b1 has been found in at least nine Chinese, two Azeris, an individual from Ladakh, an individual from Tibet, an Armenian, an Iranian, a Uyghur, and a Kyrgyz. F1b1a and F1b1d appear to be Japanese clades, although F1b1a also has been found in a Uyghur and in an African-American. F1b1c has been found in China, including a Yi, and in a Buryat, a Kazakh, a Mongush, and a Telengit. F1b1e has been found in three Uyghurs, a Kyrgyz in Artux, an individual in Sverdlovsk Oblast of Russia, a Buryat, and an Oroqen. F1b1e1 has been found in at least two Yakut. F1b1f has been found in a Stony Tunguska Evenk, a Buryat, and an individual from China. The MRCA of F1b1 appears to date back to some time between 16,000 and 26,000 ybp.]

Altai-Kizhi x 1
Uyghur x 1

Buryat x 3

Tofalar x 1
Nganasan x 1

Tubalar x 1
Bargut x 1

Siberian Tungus or Yakut x 3

Siberian Tungus or Yakut x 2

Thailand-Laos x 4
Uyghur x 1

Bargut x 3
Mongol x 2
Buryat x 1
Khamnigan x 1
Irgit x 1

India? x 7
China x 1
Buryat x 2
Bargut x 1
Khamnigan x 1
Tibet x 3
Lahu x 2
Thailand-Laos x 8
Thailand x 1
Uyghur x 7
Kyrgyz in Artux x 1

India? x 6
Tibet x 3

India? x 9

Ladakh x 1
Tibet? x 1

India? x 2
Nepal x 1
Tibet x 3
Thailand-Laos x 3
Kyrgyz in Taxkorgan Tajik Autonomous County x 2

India? x 2

The one tested UstBelaya_EBA male has been assigned to Y-DNA haplogroup Q1b1, so he may be related to many modern Ket, Samoyedic, and Turkic males.

The UstBelaya_N individual from approximately 2,000 years earlier has been assigned to mtDNA haplogroup D4b1a2. Among samples of modern populations, D4b1a2 has been found in a Yukaghir. D4b1a2a has been found in a Khamnigan and a Han Chinese in Beijing. D4b1a2a1 has been observed in eight Buryat, three Tubalar, a Mongol, a Uyghur, a Karakalpak, and an Inuit. (D4b1a2a1 also has been found in one of the present study's two specimens from Uelen_IA and two of the present study's ten specimens from Ekven_IA.) D4b1a2a2 has been observed in a Tuvan and a Buryat. The nearest outgroup, D4b1a1 (including its subclade, D4b1a1a), has been observed mainly in Japanese, although the D4b1a1a subclade also has been observed in one Kyrgyz.

Tesmos said...

I guess this is a huge blow to the out of SEA theory for haplogroup P/K2b?

Samuel Andrews said...

mHG U2 in Upper Paleolithic Europe & Siberia I think confirm that although mHG U2 has the highest frequency and basal diversity in India today it did not originate in India.

IMO, mHG U2 arrived in India with the Neolithic Iranian farmers who got U2 from their "West Eurasian" or ANE ancestors.

epoch said...

Razib Khan, in his comment section, on the Australasian admixture:

"it shows up in 10kya brazilian higher fraction. soon to be published. might be artifactual, but it’s robust"

That's interesting.

Ryan said...

@Tesmos - no really, since this sample still comes back as part ENA.

Davidski said...

Yeah, I guess there's a slim hope that the P1 in the ANS is from Southeast Asia. LOL

AP said...

P1 is the only one of the P (disregarding M and S) lines that moved north. Yana was expected to be P1 and ancestral to ANE.

Kristiina said...

The Southeast Asian theory argues that yDNA P travelled from Sunda to Central Asia via India. If the ENA ancestry in Yana UP is related to this trip, then it should be Onge like, but it is instead Devil’s Gate like. Of course, one could argue that P travelled through China to the north. However, Ust Ishim is K2a and in West Siberia c. 45 kya and now there is P1 in Northeast Siberia 31.6 kya. The age of K2 in yfull is 45 kya. It is quite convoluted to argue that K2 arose in Sunda 45 kya and an erratic K2a managed immediately to make it to West Siberia.

Ryan said...

@Davidski, @Kristiina - Does it have to be geographically from SE Asia, or just from a population genetically related to SE Asians?

@Kristiina - The literal SEA theory also argues for a slower Y chromosome clock though.

I think Oase is a bigger problem for the literal (ie geographic) theory anyways.


Kristiina said...

@ Ryan

Do you mean that if the paleolithic Central Asians are Onge-like / sitting on the same branch with Onge, that would resolve the problem?

In any case, most people seem to believe in an out-of-Africa c. 50 kya, while the Southeast Asian theory for K2, which also seems to be mainstream, requires ancient modern humans to be in Southeast Asia significantly earlier than 50 kya. These ideas are not mutually compatible.

Jaydeep said...


Onge is not AASI. If you read the Narasimhan et al supplement, AASI supposedly split up with Onge around 47 kya. That is almost as old as East-West EurAsian split. Hence Onge is not an ideal stand-in for AASI whatsoever. It's more closely related to SE Asian HG as far as I can recall.

Ust-ishim is quite closely related to South Asians among modern populations. You can refer to the figure 4a of this very paper. Among other things, the closest related modern y-DNA subclade to that of Ust-Ishim was found with a Telugu individual in India by Poznick et al 2016.

Hence a route out of India makes imminent sense.

On a wider scale, one may envisage an Out of Africa migration with some population leftover in the Near East while the rest of them move to South-SE Asia where they split up around 50 kya and migrate in different directions. The one that moves into South Asia splits up again with some reaching Siberia, other Europe while some into the Near East mixing up with the Basal Eurasians.

Davidski said...

Nah, South Asia has nothing to do with this, and neither has Southeast Asia.

Obviously, what we're looking at here are expansions from Siberia.

Kristiina said...

@ Jaydeep

If K2 arose in Southeast Asia 50 kya ago and both K2a and K2b quickly move back to India and beyond, I would expect Ust Ishim and Yana to show at least an admixture edge from Onge/Papuan branch.

You seem to suggest that the movement of yDNA K from Southeast Asia was only a movement of yDNA without the movement of the autosomal ancestry.

Hector said...


The Telegu guy is Y-philogenetically no closer to Oase or Ust-Ishim than a typical O-M122 Chinese guy is to them. All modern Y chromosomes are about equally derived from a common ancestor. There is no such thing as modern "ancestral" lineage. What you have is a relic or rare lineage and only when it adds to the philogenetic diversity the homeland can be sometimes, but not definitely, inferred.

The case in point for example would be K2b-P2 among SE Asian islanders in addition to other K2b's. This results in a nested structure that is highly suggestive of either the homeland or the native population that the lineage originated from.

India does not have much K-M526 except those that are of certifiably recent SE Asian origin such as O-M95. So you don't really have a case.

Critias said...
This comment has been removed by the author.
Hector said...

@Kristiina Neither Ust-ishim nor Oase has any admixture edge from any modern population. So there. I hate to be harsh but your reasoning is too amateur.

Davidski said...


Since you're such a pro mate, why don't you tell us again how R1a is native to India and Siberia, and couldn't have arrived there from Eastern Europe during the Bronze Age.

Ric Hern said...

The High Altitude adaptation of Denisovans tells me that Mountains played a significant role early on in Eurasia and maybe played a part in the Migration routes of Early Modern Humans...This makes the Himalayan Mountain Range particularly important as an early distribution point I think....?

Ryan said...

@Kristiina - "Do you mean that if the paleolithic Central Asians are Onge-like / sitting on the same branch with Onge, that would resolve the problem?"

Yes. Or derive part of their ancestry from an East Asian branch at least.

I suspect a lot of what we're seeing in terms of Y-chromosome structure is segregation by time rather than geography - ie K, C and probably D came from an earlier dispersal across Eurasian than F(xK).

"I would expect Ust Ishim and Yana to show at least an admixture edge from Onge/Papuan branch."

Except they do, or at least from an ENA branch. The Onge are probably the worst proxy for whatever deep East Asian ancestry was in the area given the geography.

@David - You're talking about a dispersal of P1 from Siberia I assume, and not of K2?

AP said...

No point going on a tangent about R1a. Non South Asian origin of R1a has been clear for a while.
Hector is correct we have to look at P1 in relation to other K2b and that a South Asian origin looks implausible for P1. P1 is just a twig that became successful. There is no doubt that P1 & its immediate descendants show clear expansion from Siberia (or perhaps Beringia).

Ryan said...

Talking about a nested structure, I find it interesting that the OoA Y-chromosome gene pool is not nested. Or that it has multiple layers of nesting.

Consider that West Eurasian F(xK) is nested within what is otherwise mostly East Eurasian (or at least "Old Eurasian") CF. K though is nested within what is otherwise pretty much exclusively West Eurasian F.

That suggests to me at least that the Out-of-Africa process was probably complex, with different staging grounds and structure to the various dispersals into Eurasia.

I think some of David's work with Treemix similarly suggests a deep and complex structure to OoA too.

What I find really strange is that while West Eurasian F(xK) seems pretty diverse, suggesting that there wasn't some great bottleneck, the relatively low frequency of haplogroup M on the West Eurasian mitochondrial DNA side suggests a more severe bottleneck for West Eurasians.

What kind of weird process wipes out mtDNA diversity but not Y DNA diversity? Is this evidence that the first OoA tribes were matrilocal?

Slumbery said...

"The age of K2 in yfull is 45 kya"

I don't know, maybe these estimations became more accurate as they got somewhat calibrated to ancient DNA, but based on memories from a few years ago I'd say this number can be significantly off.
The same goes for the 50 kya out of Africa (actually I'd bet that groups with surviving genetic heritage moved to Eurasia earlier than that).

"Obviously, what we're looking at here are expansions from Siberia."

The expansion of what? P1? P? K2? K?
At some point the lineage had to come from the "South" (wherever it was). Almost certainly before P1 though.

Anthony Haken said...

What are your thoughts on the new Finnish IA sample? She has none of the Siberian ancestry found in the later Saami IA samples or present day Finns. Possibly Saami prior to mixing with Paleo-Laplandic populations or whatever BOO is?

Davidski said...

@Anthony Haken

What are your thoughts on the new Finnish IA sample? She has none of the Siberian ancestry found in the later Saami IA samples or present day Finns. Possibly Saami prior to mixing with Paleo-Laplandic populations or whatever BOO is?

I doubt it, because the Saami_IA samples have plenty of Siberian ancestry.

James said...

Did anyone else note how the chart shows Ust Ishim belonging to K2b and not K2a is this a mistake or a new discovery?

Samuel Andrews said...

"What I find really strange is that while West Eurasian F(xK) seems pretty diverse, suggesting that there wasn't some great bottleneck, the relatively low frequency of haplogroup M on the West Eurasian mitochondrial DNA side suggests a more severe bottleneck for West Eurasians.

What kind of weird process wipes out mtDNA diversity but not Y DNA diversity? Is this evidence that the first OoA tribes were matrilocal?"

It is an interesting question to look at. Paleolithic "West Eurasians", the branch that includes UP Europeans & Siberians & who contributed to Paleolithic southwest Asians, appear to have had almost 100% mtDNA U.

More recent mtDNA U founder effects dominated Mesolithic European mtDNA: U5, U4.

But, we also see Y DNA founder effects. Paleolithic Siberians (ANE-Yani) might have been 100% Y DNA P1 (later R & Q). Paleolithic Europeans might have been almost 100% yHG I & C1a2. A Paleolithic ancestor of CHG/IranNeo might have been close to 100% yHG J. Ancient Anatolians were largely yHG G2a2. Mesolithic east/central Europeans almost 100% R1b1a & I2a2a.

Samuel Andrews said...

Considering mHG U is the maternal signature of the Paleolithic "West Eurasian" genetic signature maybe its direct ancestor mHg R is the maternal signature of a Paleolithic genetic signature at the core of "Crown Eurasian."

Samuel Andrews said...

Btw, mtDNA gave a big NO! to the idea that WHG shared a recent ancestor with Paleo southwest Asians. Their share a distant Upper Paleolithic ancestor.

Hopefully, the Reich squad will stop saying WHG arrived in Europe 14ky from Turkey. There's a 19ky U5b2b in Italy, 20ky U5b in Spain.

Davidski said...


These samples are actually in the Global25 already...



Davidski said...

Finland_IA:DA236 seems to be Scandinavian (ie. from Sweden or Denmark), which is in line with the isotope analysis from the paper.

In summary, the 87 Sr/ 86 Sr data suggest, that individual JK1963/DA238 was probably a local relying on terrestrial foodstuffs, while individuals JK1968/DA234 and JK2067/DA237 might well have been locals incorporating a component of Baltic sea resources in their diet. Individual JA2065/DA236 could have been of foreign origin, or a local using almost exclusively Baltic Sea dietary items.

Nathan Paul said...

Slumbery wrote
"The expansion of what? P1? P? K2? K?
At some point the lineage had to come from the "South" (wherever it was). Almost certainly before P1 though."
You nailed the point.Lot of resistance for that. All groups need to be renamed as K1233456.

Davidski said...

Well duh, of course there was a migration from the south.

That's how Ust'-Ishim got to Siberia. But Ust'-Ishim's ancestors weren't from Southeast Asia.

Jaygk said...

I think they were from Southeast Asia. I personally prefer a South/Central Asian route for the migration of ancestors of Ust-Ishim (probably an early East Eurasian) and a separate route through Northeast Asia for EEA (Early East Asian) ancestors of Yana.
And the direction of gene flow is clear. ANS and ANE are admixed with EEA, not vice versa.

Jaygk said...

"Chad Rohlfsen stated that ANE was Sunghir + East Asians. He was right. They model ANS as roughly 25% Tianyuan and 75% Sunghir and ANE as descendant."

What separates ANS apart from Sunghir is the presence of ~25% ancestry from EEA and y-dna P1. I think it makes most sense that these two suggest a migration of the same people.

Davidski said...


You should read the preprint again, this time more carefully. Here's the key part...

The first inhabitants, a previously unknown population of "Ancient North Siberians" (ANS), represented by Yana RHS, diverged ~38 kya from Western Eurasians, soon after the latter split from East Asians.

Where did this happen? In Southeast Asia? Keep dreaming.

Nathan Paul said...

Davidski wrote

"Where did this happen? In Southeast Asia? Keep dreaming."

where else ?

Davidski said...

@Nathan Paul

Obviously somewhere close to where the ~31 kya old ANS remains were dug up.

Realistically, there's no time for anything else.

Samuel Andrews said...

Yana doesn't seem to share a whole lot of drift with MA1. Yana might be a relative of MA1 not ancestor.

Leucuuo said...


That “K2b” is probably a mistake. One paper (Haber 2012) has reported a case of a Hazara in Afghanistan with Y-DNA hg M1 (M4/M186) a.k.a. K2b1b(!) This seems more likely to be an error/otherwise unknown subclade of K2a, O, N, Q, R2 etc. (Also, the Hazara are extremely admixed and there may just be a religious/trade connection as the Moluccas, which were an early stronghold of Islam, have some M1.)

As regards Ust'Ishim, Poznick et al. 2016 seem to confirm that both Ust'Ishim and Oase1 are the only known members of K2a*. The same authors untangled the Y-DNA line of K-M2308 > K-M2313 > NO-M214, all three of which were previously considered to be synonymous with K2a. For some reason Poznick's discoveries have not yet been incorporated in the quite different (and mutually confusing) trees published by ISOGG or YFull (which also list two parallel lines, probably branching from K-M2308, K-M2313, or NO-M214; YFull has K-Y28299 > K-Y28301 and ISOGG has NO1~CTS707/M2306). We also know of two living members of K-M2313* – a Telugu man in the UK and an ethnic Malay from Singapore – as well as a member of K-Y2829* somewhere in India and two members of K-Y28301*, from Andhra Pradesh and Arunachal Pradesh respectively. So, we have some hints that K2a, K-M2313 and NO originated somewhere around South Asia(?)

Conversely, I think the case for P1 a.k.a. K2b2a originating in South Asia is much weaker (see below).

@AP "we have to look at P1 in relation to other K2b and that a South Asian origin looks implausible for P1. P1 is just a twig that became successful. There is no doubt that P1 & its immediate descendants show clear expansion from Siberia"

Yes, although Tumonggor, Karafet et al., 2014 found that K2b*, P* (P295) a.k.a. K2b2* and the only known samples of the primary subclade P2 (B253) a.k.a. K2b2b occur together only among the Agta (or Aeta) of Luzon. In addition, Luzon, like many other parts of Island South East Asia, has significant P1* (M45) a.k.a. K2b2a. The present-day high concentrations of P1* are spotty and ambiguous, possibly representing founder effects, i.e. in addition to Central Asia, they include Pacific East Siberia and the Burma/India border region. Note also that P1 is negligible/non-existent in Pakistan and Afghanistan. So while it's possible that P1* in the Philippines represents back-migration from further west, the broader distribution may suggest that P1 originated in South East Asia?

Matt said...

Sam, yeah, this is the conclusion of the D-stats analysis in the supplement as well. The Yana samples seem something to MA-1 and AG3 as Goyet116-1 is to the Villabruna cluster. Probably structure and replacement *within* ANS related populations across Siberia just as within the West Eurasian dominant clade. It might be worth discussing that section, I'll have a look again at it later.

Ebizur said...

"The Young Yana specimen is directly dated to 862 ± 26 radiocarbon years BP (OxA-32884)"

"Yana_MED Yana_young N-M178 N1a1a1a1a4a1 7 7 1.00 M1993;M1987;M2077;M1991;M2103;M2108;M2122"

According to the current version of the YFull tree, N-M1993 is estimated to have a TMRCA of 1550 [95% CI 1000 <-> 2100] ybp. The samples analyzed include:

SRR1822619 Even/Sakha Republic N-M1993*(xM1979, M1991)
id:YF01684 Sakha Republic N-M1991 > N-M1988
SRR1822287 Evenki/Sakha Republic N-M1991 > N-Y25011
ERR1347705=HGDP00951 Yakut/Siberia N-M1991 > N-Y25011

According to "Yakuts#Origin and history" at Wikipedia:

"The ancestors of Yakuts were Kurykans who migrated from Yenisey river to Lake Baikal and were subject to a certain Mongolian admixture prior to migration in the 7th century. The Yakuts originally lived around Olkhon and the region of Lake Baikal. Beginning in the 13th century they migrated to the basins of the Middle Lena, the Aldan and Vilyuy rivers under the pressure of the rising Mongols. The northern Yakuts were largely hunters, fishermen and reindeer herders, while the southern Yakuts raised cattle and horses."

We now have a specimen found in the Yana River basin, which is located northeast of the Lena River basin, radiocarbon dated to the 12th century CE who belongs to a typically Yakutian subclade of N-M178.

It appears that either some early Yakut must have migrated as far northeast as the Yana River by the 12th century CE, earlier than the Yakut colonization of the Middle Lena and the Aldan and Vilyuy basins that is alleged in the current version of the Wikipedia article to have occurred in the 13th century CE, or else the typically Yakutian subclade of N-M178 must not be of Yakut ancestry, but rather some sort of pre-Yakut (Ewenic? Yukaghir?) ancestry.

epoch said...


This is the second time the Kopenhagen group claims Basal Eurasian in pre-farmer Europe. He does it like this:

f4(Mbuti, Ust'Ishim; Karelian_HG, Kostenki_UP) 0.0006 2.44 2,418,348
f4(Mbuti, Ust'Ishim; Karelian_HG, Sunghir_UP) 0.0003 1.89 2,805,748
f4(Mbuti, Ust'Ishim; Karelian_HG, Bichon_LP) 0.0004 2.12 1,822,240
f4(Mbuti, Ust'Ishim; Karelian_HG, Loschbour_M) 0.0004 2.04 2,681,732
f4(Mbuti, Ust'Ishim; Karelian_HG, Yana_UP) 0.0007 3.84 2,792,248
f4(Mbuti, Ust'Ishim; Karelian_HG, Malta_UP) 0.0007 3.27 1,937,098

I wonder if this actually is BE. If so, then Samara_HG should be similar. Is there a way to see if this also works?

f4(Mbuti, Ust'Ishim; Karelian_HG, Malta_UP)

Also, does |Z| > 3 in f4 stats mean that D stats will also produce a Z score larger than 3?

Chad Rohlfsen said...

It is likely BE and much older. The comparison between GoyetQ116-1 and all other hunters makes it more obvious. SunghirIV is one that looks like they could be modeled as almost 1/3 as much basal as Natufians. There is a hypothesis that Gravettian is rooted in West Asia.

epoch said...

Sorry, meant to ask if this also works:

D(Mbuti, Ust'Ishim; Karelian_HG, Malta_UP)

epoch said...


Linked to the Ahmerian, isn't it?

I saw a paper where they stated that there was cultural continuity in the Carpatian basin from very early (Aurignacian) on. I also saw a paper which claimed that the Gravettian spread from there. Unfortunately I lost the link to both.

epoch said...

Ah, found one again:

"As Dr. Moreau summarises, "Not only did Gravettian hunter-gatherers accept higher costs of obtaining lithic raw materials when local rocks proved unsatisfactory, their social networks mitigated the risk of resource failure, more than in any previous period. Contrary to common narratives, strong technological similarities between Aurignacian and Gravettian assemblages in Romania reflect changing adaptive and mobility strategies, not a new incoming population."

epoch said...

And the other:

Davidski said...


I'm not quite sure what you're asking for?

Matt said...

@epoch, if you're asking for the f4(Mbuti, Ust'Ishim;Samara_HG,Malta_UP), it may also be worth asking for f4(Mbuti, Ust'Ishim;Sidelkino,Malta_UP) as well, for a third check, and maybe those Latvian MN who look like EHG transplants as well? I4435 and I4554. That gets you to about 5 genetic EHG to check out. You'd think if anything a CHG like admixture should be stronger in Samara than Karelia though, on the grounds of geography.

(AG3, Iron_Gates_HG, West_Siberia_N might also be worth looking at for cline patterns).

epoch said...

@David & Matt

Yes, that wasn't very clear. My only excuse is that I have a hectic family life. There is no such thing as a quite Sunday anymore ;).

Anyway> I was asking two things:

1) As Matt describes above: What does f4(Mbuti, Ust'Ishim;Samara_HG,Malta_UP) and f4(Mbuti,Ust'Ishim;Sidelkino,Malta_UP) do? He also suggested expanding that to Latvian MN.

2) Does D(Mbuti, Ust'Ishim; Karelian_HG, Malta_UP) come up with a similar result as f4(Mbuti, Ust'Ishim; Karelian_HG, Malta_UP)?

@Chad Rohlfsen

The two links are for Chad.

Davidski said...


I can run stats for you occasionally, but you have to list the stats in the simplest way possible, and drop all of the paraphernalia, like the f4s and brackets, like this...

Mbuti Ust_Ishim Samara_HG MA1 0.002035 2.984 338275
Mbuti Ust_Ishim Sidelkino MA1 0.00216 3.698 759528
Mbuti Ust_Ishim Latvia_MN MA1 0.001297 2.634 696700
Mbuti Ust_Ishim Karelia_HG MA1 0.001767 2.835 690129

Because first I have to figure out what you really want, and then remove all of those unnecessary extras. All of this takes time.

epoch said...


Thanks, I'll try to be more clear in requests. I appreciate the effort.

This is interesting as it might be the case that CHG mixed into EHG. So the time and place where EHG came to be, and how it came to be becomes interesting. If EHG really is a mixture it could mean that CHG reached far more north of the Caucasus at one point. I recall people linking Kapova Cave to the expansion of ANE westward.

So, we have Dzudzuana at 24.000 BC. We have Sidelkino at 9.000 BC. We have Satsurblia at 11.300 BC. What happened?

Lenny Dykstra said...

It's a shame they didn't include Dzudzuana in their analysis/topology, it could have potentially been quite interesting.

I also like the fact that Kets fall on the Ancient North Siberian-East Asian Cline (like Native Americans), not the Neo-Siberian-East Asian cline (as apparently most Turkic/Tungusic/Mongolic speakers).

I believe the Xiongnu were proto-Yeniseian speakers who had back-migrated from Beringia, and the Kets are maybe the only modern group that can claim perhaps the majority of their ancestry from this back-migration. Incredible.

Ric Hern said...

Apparently there was Hunter Gatherers living near the mouth of the Don River(Kammenaya Balka) at 20 000 years ago and archeologically they look related to the Imereti Culture of the Caucasus...

So maybe ANE like people kept to the North until the Mammoths died out and then migrated down the Volga to the mouth of the Don where they could have encountered the CHG people.

That area was apparently more forested and I can see how desperate Plains Hunters could have changed their survival strategy by hunting Red Deer and Aurochs in a more forested environment.

gL said...

@Santosh Rajan
"Clearly C, K, and H passed through South Asia."
wiki really helps on this subject, if you read and analyze it

C and D passed - some of C stayed in west, too for some time - even if there is not that much left in modern population.
There was not much of H migration going on in ancient times, if compared to modern times. H is mainly India and Iran and Middle East(in earliest times) and over time it lost territory to other groups.

Origins of K are in SE Asia. At least for K2 - not sure if LT is backmigration through India or they developed there.

C is coastal migration - from Africa till Australia and even a bit further.

Current Yakut ancestry is linked to a very recent assimilation of Yukagir people, who probably populated also Lena basin, which essentially is the size of Saha. I believe, that wiki has info about this - maybe not in english, but at least maps about spread of Yukagir in 18th century should give some ideas, that they were even more widespread before russians encountered them. Actually I remember some interesting read about early contacts of Russians and Yakuts and how they at that time expanded under russians, too. Cossacks at the same time changed a lot of ethnic map - by extermination of locals, so ethnic map of Siberia before expansion of Russia looked very different.

@Lenny Dykstra
Yenniseian could not have be that backmigration, because Na-Dene who are proposed to be most related to Ket have huge amounts of C from locals of Asia across Bering strait, where Ket are nearly 100% Q. That doesn't mean, that there was no back and forth migration across Beringia - water is best transportation route that doesn't require that much energy and Inuit people at some time did migrate to Siberia from and then again back to America.
The truth is that R in North Asia for a very long time period was lesser brother of more populous Q brother, that went on to America. From what I understand, then earliest R1 samples are near Baikal(20 000) and stayed there for a very long time - it was bottleneck population for many thousand years - probably they did not hunt mammoths, so had to hunt something else, that couldn't sustain huge population growth. There is still remnant of R2 in India, so Q and R split from each other before migrating to north(from SE Asia). I understand, that P can still be found in Siberia and SE Asia, so that looks, that there were at least 3+ waves in different times, that made to north - P, Q and R or that some of those groups took along neighboring groups.

Xiongu is very late culture and is mainly related to Mongolians. Part of Yenisseans(who at that time were main inhabitants of Baikal might have been included in this aglomeration of tribes. Hovewer, split of Na-Dene and Yenisseans most probably happened 10000BP. And even if we take, that Yenisseans were split from Na-Dene, because of spread of N(4000 - 11000), that still was very far away in time from Xiongu.

The article takes a look on a very wide time in history. Yeah, that is the name of paper, too. There is no data that would challenge already know facts and there is no breakthrough - it just makes something as a foundation to already proposed or known information.

Ric Hern said...

@ gL

So your whole theory hangs around R2 in India ? How about Ust Ishim way up North ? So K2 ran all over the place and P1 just sat and wait to incubate R and Q in Southeast Asia for thousands of years before all of them suddenly got a urge to migrate together ?

Davidski said...


R2 isn't from India. It arrived there with Neolithic farmers from Iran. Two samples from early Neolithic Iran belong to R2/pre-R2.

Davidski said...

And the earliest R1 is from Europe, not from near Lake Baikal.

At least try and get the basics rights.

Ryan said...

If K2 passed through a northern route, then that must mean East Eurasians are a mix of two distinct streams of ancestry. You had a Treemix that suggested that a while back didn't you David?

gL said...


Damn... I've just realized, that somehow I've mixed P* branch with R* and that's been imprinted in my mind for some time and raised questions how they made from SE Asia. Anyway, disregard anything I've mentioned before about P branch, as it is pure error.