search this blog

Wednesday, April 7, 2021

The Bacho Kiro surprise (Hajdinjak et al. 2021)


Over at Nature at this LINK. The paper focuses on Neanderthal ancestry in Initial Upper Paleolithic (IUP) humans from what is now Bulgaria. But, to me, much more interesting is the claim by its authors that present-day East Asians harbor ancient European, or, at least, European-related ancestry. From the paper, emphasis is mine:

When we explored models of population history that are compatible with the observations above using admixture graphs [28], we found that the IUP Bacho Kiro Cave individuals were related to populations that contributed ancestry to the Tianyuan individual in China as well as, to a lesser extent, to the GoyetQ116-1 and Ust’Ishim individuals (all |Z| < 3; Fig. 2d, Supplementary Information 6). This resolves the previously unclear relationship between the GoyetQ116-1 and Tianyuan individuals [13] without the need for gene flow between these two geographically distant individuals.

...

In conclusion, the Bacho Kiro Cave genomes show that several distinct modern human populations existed during the early Upper Palaeolithic in Eurasia. Some of these populations, represented by the Oase1 and Ust’Ishim individuals, show no detectable affinities to later populations, whereas groups related to the IUP Bacho Kiro Cave individuals contributed to later populations with Asian ancestry as well as some western Eurasian humans such as the GoyetQ116-1 individual in Belgium. This is consistent with the fact that IUP archaeological assemblages are found from central and eastern Europe to present-day Mongolia [5,15,16] (Fig. 1), and a putative IUP dispersal that reached from eastern Europe to East Asia. Eventually populations related to the IUP Bacho Kiro Cave individuals disappeared in western Eurasia without leaving a detectable genetic contribution to later populations, as indicated by the fact that later individuals, including BK1653 at Bacho Kiro Cave, were closer to present-day European populations than to present-day Asian populations [29,30].

Hajdinjak, M., Mafessoni, F., Skov, L. et al. Initial Upper Palaeolithic humans in Europe had recent Neanderthal ancestry. Nature 592, 253–257 (2021). https://doi.org/10.1038/s41586-021-03335-3

See also...

Ust'-Ishim belongs to K-M526

58 comments:

Davidski said...

Where are the Bacho Kiro samples from that are in the Global25.

Are they from the official dataset?

Arza said...

@Davidski
They are from the BAM files from ENA processed the same way as kra001. So not official, but you won't get official 1240k dataset anyway as:

We enriched the selected amplified libraries for about 3.7 million SNPs across the genome described in supplementary data 2 of ref. 19 (SNP Panel 1 or 390k array), and supplementary data 1–3 of ref. 7 (SNP Panels 2, 3 and 4, or 840k, 1000k and Archaic admixture arrays, respectively).

Another publication:

A genome sequence from a modern human skull over 45,000 years old from Zlatý kůň in Czechia
https://www.nature.com/articles/s41559-021-01443-x

Rob said...

I dont think IUP populations expanded from Bulgaria to China any more than the other way round. Rather, these affinities relate to IUP dispersing population (probably from western Asia) . Somewhat, paradoxially it was preserved to greater degree in subsequent Paleo populations in western Europe (Goyet) than in Eastern Europe or Siberia. The apparent East Asian affinities ('pseudo-East Asian' as I termed it) of early IUP groups is due to comparatively less turnover in East Asia c.f. western Eurasia

Simon Stevin said...

This is all starting to make some sense. I have thought for a while now that Tianyuan, Salkhit, and Yana’s common ancestry, is really a sort of undifferentiated common Eurasian/Central Eurasian, with one branch contributing ancestry to modern South and East Asians (hence Tianyuan’s mtDNA B), and the other (Yana/Salkhit) contributing ancestry to modern West Eurasians and Native Americans. Y-DNA haplogroup K is clearly this undifferentiated “central” Eurasian lineage, it spawned P, R, Q, and NO. This may also help explain the common Eurasian like character of mtDNA C, as well as the appearance of mtDNA X in Native Americans. Just my own amateur theory.

Archi said...

I don't agree with them at all. They are categorically wrong about the direction of gene flow. Neither Tianyuan came from Europe, nor did GoyetQ116-1 and IUP Bacho Kiro come from China. They have a typical lantern search error, they only see single appearances under the lantern and draw erroneous conclusions because they don't see the rest of Eurasia. Their assertion that settlement went from Europe to East Asia is nonsense, although they expressed themselves poorly here may be.

epoch said...

And another surprise is that a 45 k yo sample from Czechia is not related to anybody at all.

https://www.nature.com/articles/s41559-021-01443-x.pdf

Matt said...

@Rob, that is probably correct.

...

qpGraph is interesting in this paper; Tianyuan is a composite of 39% of this IUP Bacho Kiro group that is Basal to Kostenki, and then 61% part that splits with Kostenki.

Andrzejewski said...

Probably resolve the puzzle of why some forager and earlier ANF pops had ydna C1

Genos Historia said...

@Andre,

Yeah, thanks for mentioning this. I think it does.

Bacho Kiro Cave
CC7-335 and BB7-240. 45ky-42ky
Both carry Y DNA C1.

They must be where Paleo Euro and Anatolian C1 comes from.

Goyet had Y DNA C1a, and this paper says has ancestry from Bacho Kiro Cave. It is all making sense.

old europe said...


@all

What about the 5 R1b M 73 samples from the Kunda culture? It seems nobody is elaborating the implications for the R1b history in eastern europe.

Archi said...

@Genos Historia

"Goyet this paper says has ancestry from Bacho Kiro Cave."

It doesn't say that. It says that GoyetQ116-1 shares significantly more alleles with Bacho Kiro Cave, not that they are its ancestors.

Seinundzeit said...

Matt,

"qpGraph is interesting in this paper; Tianyuan is a composite of 39% of this IUP Bacho Kiro group that is Basal to Kostenki, and then 61% part that splits with Kostenki."

You should see what happens when a graph includes MA1, ancient Near Easterners, and ancient Africans!

In that particular vein, I've been working extensively with automated qpGraph, supervised qpGraph, qpAdm, f4 stats, etc. There are some consistent features ("consistent" despite differences in SNP selection and number, varying population choices, and a varying number of admixture edges) that I think would pique your interest

Speaking for myself, the most fascinating and consequential results pivot upon the "Basal Eurasian" concept:

All things considered, it appears rather certain that the notion is unnecessary. Despite (by now) countless attempts, the ghost population simply doesn't appear at the end of any automated graph explorations (the cool thing to note here is that you can compare multiple graphs, resample SNPs, lots of interesting things).

And good ol' supervised qpGraph models with a "Basal Eurasian" ghost yield worse fits (in terms of both either the likelihood score or the lowest worse residual; you can look at both, and choose to optimize for either/or) than ones without.

Instead, it seems rather more parsimonious to allow ourselves a conception of "African" admixture in southern West Eurasia (from the Levant to northwestern India, with levels of this ancestry decreasing as distance from northern/northeastern Africa increases. I imagine that modern Southern Europeans and South Central Asians would also have this sort of ancestry, but much closer to "trace" levels).

I understand that the semantics might get complicated; some Sub-Saharan African lineages are closer to Eurasians than they are to other Africans.

But based on the topologies, if Mota is "African", and if 60-70% of Iberomaurusian ancestry is "African", then the ancestry I'm describing in southern West Eurasia is best described as "African".

Another result of interest involves the subject of the paper under discussion:

It does seem that East Eurasians are heavily IUP, but with substantial "West Eurasian" admixture. This "West Eurasian" admixture involves the stream of ancestry which apparently connects MA1 to the Iranian plateau/southern Central Asia/northwestern South Asia. So the relationship between MA1 and the ancient eastern Near East/northern South Asia is definitely real... but that shared connection seems to also be the conduit for "West Eurasian" admixture in the deep ancestral makeup of ENA. I'd imagine that this "West Eurasian" pop which connects ANE with Tajikistan_Sarazm/Indus_periphery/Iran_N existed in Central Asia, and from there flowed into what eventually became the populations of East Eurasia. But I'm just spit-balling here; who knows the reality of it all. More aDNA is needed.

In the context of South Asia (which you and I often discuss), it could even be that the "AASI" conceptualization might also be due for retirement... but we certainly need actual Mesolithic aDNA from the region (who knows what we'll find). It'll all be a very complex situation though; that much is certain.

Rob said...

The common “west Eurasian” link between ~ Sungir & western part of ANE is probably from southwest western Central Asia / eastern West Asia

Archi said...


Interestingly, they claim that BK-1653 is a female, so they don't give a definition for her Y-haplogroup!


Tigran said...

So K/.K2b/P are ultimately from an IUP like crown eurasian population? And this is the source of K2b in Tianyuan, K2a in Ust-Ishim and Tase and P in Yana?

George said...

Is the idea that these people came from the distant east or that they left the west and pushed east? I'm partial to the idea that they came from the east and most populating of Europe in general pre-neolithic happened from the NE. Personally, I don't know what I'm talking about at all, but this would align with the dearth of UP sites in Anatolia (I might be wrong about this + might just be under-excavated/studied + coastlines have changed dramatically + populations were tiny/mobile) and the idea that lots of early OoA haplogroup diversification seems to have happened in SE Asia.

Davidski said...

These people expanded from the Middle East into Europe, Central Asia, Siberia and East Asia.

Southeast Asia is a sink, not a source.

Its relatively high Y-chromosome diversity is an artifact of the multiple large scale population turnovers that happened elsewhere in Eurasia.

Rob said...

''It doesn't say that. It says that GoyetQ116-1 shares significantly more alleles with Bacho Kiro Cave, not that they are its ancestors.''

Given that Goyet is 10,000 year younger, it is likely that the deeply archaic ancestry in Goyet arrived from BK

Cy Tolliver said...

@ Seinundzeit

Can you elaborate a bit more on this potential African > West Eurasian gene flow? Are you sure there isn't any deeper Eurasian > African movement?

Rob said...

In fact, looking at fig 1, it does almost seem that IUP populations dispersed from Europe, across Siberia and into Eastern Asia. The data does seem to support it too.
There arent any IUP assemblages in the Zagros/ Iran region, interestingly (where para-Aurignacian sequences start after 40, 000 BP)

But there would have been a southern dispersal route too, toward southern India/ Sri Lanka; where the promising skeletal remains should shed light hopefully

Matt said...

@Sein, I think African related admixture models with a split of some population in Africa are plausible, difficult to say more without the qpGraphs themselves!

I do have some reservations about trying to model Eurasians and Africans on a graph as the 1240k capture panel inflates, non-equally, differentiation between Africans and non-Africans: https://arborbiosci.com/wp-content/uploads/2021/03/Skoglund_Ancestral_850K_Panel_Design.pdf

Simon Stevin said...

@Rob, How does IUP affect the relationship between Tianyuan, Salkhit, and Yana? We could be looking at a IUP-like, crown Eurasian population, who spread Y-DNA K and subclades of mtDNA M and N, from eastern West Asia/northern South Asia, into Central, North, and East Asia. This would help explain some of the more ambiguous genetic links between West and East Eurasians, both modern and ancient.

Archi said...

Bacho Kiro is not direct ancestral to GoyetQ116-1, even they on Admixture graph showed that.

The most interesting is the IUP cluster in the Altai-Baikal area. This is where we should look for the truth, because for example in the Altai, more typical Aurignyac needles were found in the IUP than in Europe. There dates of human appearance are more ancient than in Europe, there it is necessary to investigate genetically.

Seinundzeit said...

Cy Tolliver,

In a sense, it's fair to say that African > West Eurasian gene-flow could explain much about the uniparental data (primarily regarding the ancient Near East).

Essentially, the Natufians being 40% "African" is rather exactly what we would expect.

But I think you're absolutely right to note the possibility of deeper Eurasian affinities in Sub-Saharan African populations. In fact, it's almost expected (gene-flow is nearly always demonstrated upon further analyses done in conjunction with new aDNA; perhaps one could say that it's all mixture, all the way down).

But although the possibility is there, it seems that more statistical certainty can be placed on African > West Eurasian gene flow, rather than Eurasian > African movement (barring contemporary East Africans, of course).

Matt,

I have 31 independent graphs (and I've done multiple comparisons with quite a few of these) from which I could choose any single topology for the purposes of demonstration.

But it's the graph that I'm doing right now which I enjoy the most, because it's the simplest one yet (and despite that, the best-fitted). I'll let it run for a few more days, and then I'll post the topology. I think that you'll find it to be of some interest!

Now on the question of the 1240k capture panel, I'd note two things:

1) I have graphs lacking any ancient African populations which infer "ghost African" populations contributing to ancient Near Easterners.

2) The patterns in question present themselves with shotgun data as well.

Rob said...

@ Simon S.

In essence, an IUP spread from western Asia toward Europe toward Siberia then Asia(with exact details still to be sorted). It is looking like this a northern route from Europe to Siberia then East Asia.
Then a second radiation followed, bringing in the Zagros-Caucasus axis, radiating to Europe and also to Siberia.
All subsequent European Paleo history was internal re-arrangements.



@ Archi

''Bacho Kiro is not direct ancestral to GoyetQ116-1, even they on Admixture graph showed that.''

As stated above, BK is the ancestor of the most archaic component of Goyet-Q116. But G-Q116 has a dual (at least) ancestry, with the younger component making the majority (~80%), as per the qpGraph


Rob said...

“ All subsequent European Paleo history was internal re-arrangements. ”

Aside from the ANE -related arrival ~ 20 kybp to far Eastern Europe

Archi said...

@Rob

"As stated above, BK is the ancestor of the most archaic component of Goyet-Q116. But G-Q116 has a dual (at least) ancestry, with the younger component making the majority (~80%), as per the qpGraph"

IUP Bacho Kiro is Ust-Ishim, which is almost the same thing. IUP Bacho Kiro is an ancestor to GoyetQ116-1 as much as Ust-Ishim and Oase 1. Actually, they did not write anything new, GoyetQ116-1 has a share of Ust-Ishim, but Ust-Ishim is not an ancestor to GoyetQ116-1. Therefore, we cannot say that GoyetQ116-1 descended from IUP Bacho Kiro.

"In essence, an IUP spread from western Asia toward Europe toward Siberia then Asia(with exact details still to be sorted). It is looking like this a northern route from Europe to Siberia then East Asia."

There was no route from West Asia to Europe and from there on to Siberia and China. I pondered their suggestion again and realized that they did not write any such path, they simply described a distribution area from Europe to East Asia = from East Asia to Europe. It's not a path, it's just a description of the area.

There was a very different route from Africa through the straits to the north of South Asia and on to Siberia to Altai and from there to Europe and China.


Rob said...
This comment has been removed by the author.
mzp1 said...

I agree that both Basal Eurasian and ASI exist only in theory and imagination.

These things exist because this field attempts to explain things in terms of admixture when drift is a better explanation. For example, the relationship of modern humans and Neanderthals may not be of admixture but AMH drifting away from a common ancestor, hence varying levels of Neanderthal alleles remaining in modern humans.

Basal Eurasian is an explanation for an effect caused by the fact that West Asians (Iberomaurusians, Iran_N etc) are very early separated from the common Eurasian ancestor with North/East Eurasian, and NE Eurasian is geographically on the opposite side from the common dispersal point (likely SE ASIA). Hence NE Eurasian and SW Asians are the two most differentiated Eurasian populations, and Basal Eurasian as an actual population is just a bad explanation for this difference (imo).


The geneflow from between Africa and Eurasia may have two routes . One looks like a Africa to Eurasia movement up through East Africa into Gujarat, and another one in the opposite direction from West Asia into Africa. Hence Yoruba is much closer to Eurasians than Mbuti, because of geneflow from West Asia into Africa. Mbuti are connectect to Eurasians (hence greater Neanderthal affinity) but this is a more distant connection than that of Yoruba (who are more Northern in Africa) who are much closer to Eurasians.

AMH could have originated in a large region of SE Asia, South Asia, West Asia and Africa over many millions of years. It's difficult to pin it down to a sub-region of this area. I think the two points of contact between Eurasia and Africa is the key to understanding the massive drift and selection required to evolve AMH. Evolution is enhanced with movement and adaption, a circular movement allows for greatest drift, selection and adaptability and can continue without limit.

gamerz_J said...

@Seinundzeit

"Instead, it seems rather more parsimonious to allow ourselves a conception of "African" admixture in southern West Eurasia (from the Levant to northwestern India, with levels of this ancestry decreasing as distance from northern/northeastern Africa increases. I imagine that modern Southern Europeans and South Central Asians would also have this sort of ancestry, but much closer to "trace" levels)."

What do you mean by trace levels? And if that's the case then all Europeans would have that ancestry because I suspect that EEF would have it right?
Furthermore, I am very skeptical of this African ancestry given that there are no African (maybe a few idk) mtDNA lineages in ancient Near Easterners and Natufians didn't really show affinities to Sub-Saharan populations (unless you mean African differently)

"It does seem that East Eurasians are heavily IUP, but with substantial "West Eurasian" admixture. "

Also have to disagree here, East Eurasians show very little uniparental evidence of West Eurasian ancestry unless you mean older than 30-40kya. I would also add that on the autosomal side, we simply don't have enough ancient DNA to tell but Ive seen many qpGraphs of Neolithic East Asians which don't need western input.


In both cases, did you consider a) an Basal Eurasian/North African population of some sorts also contributing to Africans and b) shared IUP ancestry West/East or even some East Asian ancestry in West Eurasians such as ANE for example was modeled as having?

gamerz_J said...

@Davidski

Is there no possibility these people expanded first into Siberia and then split East and West?

Seinundzeit said...

@gamerz_J

"And if that's the case then all Europeans would have that ancestry because I suspect that EEF would have it right?"

The hunter-gatherer sample from Pınarbaşı Gölü doesn't seem to have any, so it's likely that EEF don't possess much in the way of this sort of admixture. (It's probably there, but just not much).

"Furthermore, I am very skeptical of this African ancestry given that there are no African (maybe a few idk) mtDNA lineages in ancient Near Easterners and Natufians didn't really show affinities to Sub-Saharan populations (unless you mean African differently)"

Few things to note:

1) There are a few.... but that's not what I had in mind. Rather, I was thinking of their y-dna haplogroups (like E1b1b).

2) They do show affinities to Africans in individual stats.

3) "Africans" are not a clade, at all; there are multiple deeply-diverged lineages. The conceptual landscape here is exceddingly rich and complex.

4) Basal Eurasians were real, and I already have one in my topologies: Oase1. And I'd imagine that Zlatý kůň would fall in the same category.

"Ive seen many qpGraphs of Neolithic East Asians which don't need western input."

So... you're saying that you've seen many qpGraphs that were either very poorly fitted or missing crucial populations (or both)?

"In both cases, did you consider a) an Basal Eurasian/North African population of some sorts also contributing to Africans and b) shared IUP ancestry West/East or even some East Asian ancestry in West Eurasians such as ANE for example was modeled as having?"

It's not really a matter of my consideration; if it was, it wouldn't be "automated" graph exploration. My whole goal has been to do things unsupervised, without forcing any of my own assumptions on the data.

And once I deal with the output of automated graph exploration, I manually create trees that match the literature (in order to compare the fits, and in order to derive a sense of what's more likely in terms of the broader ontology).

Before working on topologies myself, my general assumptions were exactly what you've described:

ANE and WHG have ENA admixture (with ENA being unadmixed in relation to West Eurasians), and "Basal Eurasian" is a plausible (albeit mysterious) construct.

I've spent much time trying to add evidential flesh to those assumptions... but they haven't panned out.

mary said...

For those who are curious, there is an older article with a reconstruction of the Czech skull.

https://journals.plos.org/plosone/article/figure?id=10.1371/journal.pone.0201431.g007

mary said...

@Seinundzeit

But both Oase-1 and Zlatý kůň are closer to modern Asians than to modern Europeans and the explanation for this (at least the explanation given in the papers) is precisely that modern Europeans have Basal ancestry. Therefore, these two individuals cannot be Basal-Eurasians.

Rob said...

@ Sein


''it seems rather more parsimonious to allow ourselves a conception of "African" admixture in southern West Eurasia (from the Levant to northwestern India''

Yes, hg E seems like a key index fossil of this connection.
To which period or industry do you connect this to ?
Archaeological data suggests that populations were moving into northern Africa during the LGM (~ 25 - 19,000 BP)

Rob said...

@ Archi

''IUP Bacho Kiro is Ust-Ishim, which is almost the same thing. IUP Bacho Kiro is an ancestor to GoyetQ116-1 as much as Ust-Ishim and Oase 1. Actually, they did not write anything new, GoyetQ116-1 has a share of Ust-Ishim, but Ust-Ishim is not an ancestor to GoyetQ116-1. Therefore, we cannot say that GoyetQ116-1 descended from IUP Bacho Kiro.


i know what they have written & support their scenario. There are subtle differences in archaic homo introgression between BK & U-I, which their qpGraph depicts.
The archaeological evidence shows the IUP expansion from from Danube across central Europe & their qpGrap (which I have complete confidence in)
There is simply no need to imagine undocumented migrations from western Siberia to Europe during the IUP when evidence is lacking. It would need some strong evidence brought forth by credible scholars. At present it is thrown around by amateurs and neo-creationist pseudoscholars in fora like Anthrogenica



''There was a very different route from Africa through the straits to the north of South Asia and on to Siberia to Altai and from there to Europe and China.''


This is hard to believe. The northern route is the what binds Siberia, Mongolia , Europe. These are a long blade Industries of large game hunters. South Asia is different hunter people adapted to savanah & rainforests using microlithic armature. They were simply unequipped & unlikely to have moved the northern latitudes
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0222606


As Davidski said- SEA was a sink, so was India.




''There was no route from West Asia to Europe and from there on to Siberia and China''

There are very similar IUP assemblages in eastern Turkey as in Europe (# 26 on map Ucagilzi & Kanal cave). Its just missing at present in western Turkey, due to poor research (the focus on Turkey has been Neolithic, and later periods). But at present, this seems to be more parsiminous


https://imgur.com/uByznp6

Genos Historia said...

@Sein,

Y DNA E is Eurasian.

Its is derived along the human Y DNA tree, compared to indiegous AFrican Y DNA.

Plus, Y DNA D, its brother, is Asian.

Genos Historia said...

@Sein,

People used to think of Y DNA E1b1b as indigenous in East Africa.

Ancient DNA shows it arrived there only 4,000 years ago from Southwest Asia.

gamerz_J said...

@Seinundzeit

" 2) They do show affinities to Africans in individual stats."

Hm, interesting, could you post or link me any? I am assuming then that Iran_Neolithic and CHG would also have this African input?

"Rather, I was thinking of their y-dna haplogroups (like E1b1b)"
E is possible though I recall Lazaridis had a model where it came from a "North African" population somewhere intermediate between Eurasia and Africa, whatever that means.

" 4) Basal Eurasians were real, and I already have one in my topologies: Oase1. And I'd imagine that Zlatý kůň would fall in the same category."

I would imagine these were Crown Eurasians not Basal in the sense the concept has been developed in the literature.

" So... you're saying that you've seen many qpGraphs that were either very poorly fitted or missing crucial populations (or both)?"

Well, I haven't seen your qpGraphs so I don't know, but if West Eurasian ancestry made its way into East Asia when did it do so? I doubt this happened post Mongolia_N and judging by the uniparentals only subclades of Q actually could have brought it there.

Archi said...

@Rob

"i know what they have written & support their scenario."

They wrote nothing new. Genos Historia wrote nonsense that completely contradicts both the article and all the data.

https://i.ibb.co/4WPsYNm/UP-PCA.png

"Its just missing at present in western Turkey, due to poor research"

No, Turkey is well researched. But Russia and especially Siberia are very bad, worse than ever.

"There is simply no need to imagine undocumented migrations from western Siberia to Europe during the IUP when evidence is lacking."

This is the way https://i.ibb.co/JpFdy4K/Initial-Upper-Palaeolithic-path.png.

It proves to be true both the genetic data and archeological. The haplogroup F in Bacho Kiro IUP is most likely the same pre-NO* as in Ust-Ishim and Oasa 1.


Archi said...


In Siberia in those days there were Mammoth Steppes with a great climate, much better than in Europe and Anatolia. Europe was covered by glaciers; there was no glacier in Western Siberia. The amount of food in the Mammoth Steppes was probably greater than in the savannahs. Things change in the Holocene, Siberia is covered by taiga and a harsh climate is established there, which caused people to disappear from there, and Europe establishes a good climate. The problem is that there are few places to live, and the archaeological excavations there are essentially only for construction projects. As far as I can remember, there are dozens of times fewer archaeological excavations a year in giant Russia than in little Germany, and in Siberia it's a paltry thing altogether. Although even under current conditions this way is already quite visible.

Cy Tolliver said...

@ Sein

Regarding Y-DNA E and Natufians, yes that's clearly indicative of African movement into the Levant, but one thing that's always been perplexing is that Iran_N shows stronger affinity to Africans in D-Stats than Natufians do. Iran_N is more closely related to Pinarbasi and Anatolia_N and even to WHG than they are Natufians. So it always seemed like the "African" affinity of Iran_N was something strange, because it sure doesn't seem like it's due to any Natufian > Iran related flow.

Arza said...

Off-topic

Medieval Y-DNA from what appears to be a 16th century site in Ellwangen, Germany (based on: https://www.nature.com/articles/s41467-019-12154-0/figures/1 ; sequences from ENA PRJEB44124):

Sample_name Hg Hg_marker Total_reads Valid_markers QC-score QC-1 QC-2 QC-3
ELW025 J2b2~ J-Z575*(xYP84,Z8257,Y21763) 456122 2345 1.0 1.0 1.0 1.0
ELW035 R1a1a1b1a2 R-S204*(xVK04,YP5525,YP4464,YP4480,CTS11505,YP5308,FGC19249,Y17488,FGC13708,Y10895,A7018,FGC39336,YP6405,FGC21215,FGC19279,YP315,FGC74376,Y92440,Y29593,Y81519,Y10807) 437430 2238 1.0 1.0 1.0 1.0
ELW031 R1b1a1b1a1a2c1a5c1a R-L720 247354 1179 1.0 1.0 1.0 1.0
ELW003 R1b1a1b1a1a1c R-S263*(xS265,FGC30618,FGC15439,Z8180,CTS11549,S383,A2426) 446863 2148 1.0 1.0 1.0 1.0
ELW036 E1b1b1b1a1 E-M5172/etc*(xBY9780,Z36168,MZ129,MZ6,MZ87,Y21135,MZ54) 634129 3239 1.0 1.0 1.0 1.0
ELW033 R1b1a1b1b R-CTS1078 570531 2862 1.0 1.0 1.0 1.0
ELW030 I2a1b2a I-FGC29577*(xY17132,Y129082,BY13995) 185628 926 1.0 1.0 1.0 1.0
ELW028 R1b1a1b1b3a1a1 R-CTS9219 787042 3978 1.0 1.0 1.0 1.0
ELW013 J2b1 J-Y3160/etc*(xY22081) 480498 2366 1.0 1.0 1.0 1.0
ELW018 E1b1b1a1b1a E-PF2211*(xZ16660,B410,M8166,BY4890,BY5249,Y16726,A9487,BY5124,Z38899,BY5809,BY5840,L540,Z42775,BY4224,FGC5638,Z25461,Z37882) 667178 3311 1.0 1.0 1.0 1.0
ELW008 R1b1a1 R-PF6524*(xY13872,L265) 891355 4527 1.0 1.0 1.0 1.0
ELW017 R1b1a1b R-CTS11468*(xFGC36478,S12993,FGC22500,CTS2501,Z17624,FGC18033,A16,S7898,Z16539,A94,A11720,Y31466) 136261 642 1.0 1.0 1.0 1.0
ELW029 E1b1b1b1b1a2~ E-FGC18986*(xPF2440^) 558666 2787 1.0 1.0 1.0 1.0
ELW020 R1b1a1b R-L1351/etc*(xFGC36479,PF7598,Z264,L356,FGC5541,A94,CTS12966,Y18209,A8041,S20902,A11718,GG729,FGC31939) 724755 3600 1.0 1.0 1.0 1.0
ELW040 R1b1a1b R-PF6436/etc*(xFGC20733,FGC13326,S9257,S18461,A96,S10024,A9452,FGC19756,L200,A7220,FGC419,FGC18015,S26140,Z262,L356,A228,A8062,FGC37114,M12132,PH1348) 354282 1755 1.0 1.0 1.0 1.0
ELW012 R1b1a1b1a1a R-PF6543*(xFGC36479,M228.1,S11481,A8064,FGC37108) 788786 3950 1.0 1.0 1.0 1.0
ELW027 I2a1b1~ I-CTS3111/etc*(xPF3966,CTS9139,Y55873,S10269,BY14030,BY14066,BY25365,BY14002,Y106155,Y19666,S9168) 403558 1993 1.0 1.0 1.0 1.0


ELW037 appears to be https://www.yfull.com/tree/C-F1756/ C2a1a1b1

mzp1 said...

As Yoruba is much closer to Eurasians than Mbuti, a useful calc to look at human population structure can be

d (W=Mbuti, X=Yoruba) (Y=Test, X=Test), for a group of samples Test.


For 21 members in Test the matrix is quite large at 21x21 so we can just take the mean average for each test=Y.

1 Morocco_Iberomaurusian -0.0160
2 Gujarati Muslim -0.00370
3 PUN.SG -0.00246
4 Jatt -0.00158
5 Saudi -0.00154
6 Greek -0.000975
7 Ukrainian -0.000345
8 GujaratiB -0.000345
9 English -0.000175
10 Iranian 0.0000700
11 Kalash 0.000255
12 Balochi 0.000345
13 Pathan 0.00051
14 Han 0.0016
15 Mongol 0.00174
16 ONG.SG 0.00179
17 GujaratiA 0.00181
18 Iran_GanjDareh_N 0.00196
19 Russia_Ust_Ishim.DG 0.00432
20 Belgium_UP_GoyetQ116_1_published_all 0.00594
21 Russia_Kostenki14.SG 0.00676

The top ones with negative score are closest to Yoruba vs Mbuti. Ancient Eurasians score closest to Mbuti, meaning there is more recent DNA sharing between Yoruba and Eurasians from South and West Eurasia.

Slumbery said...

@Seinundzeit + Rob

As it was pointed out by Genos Historia, the assumption that YDNA E represents a migration from Africa leads us to the old problem of E vs. D. There were debates about this a decade ago and I cannot say we know much more about the issue that back then. YDNA D is not only Eurasian exclusive, but have a very Eastern distribution since ancient times (it is definitely a deep Paleolithic lineage in East Asia). Do you think that the split happened in Africa, but their distribution became disconnected due to later movements?

Slumbery said...

@mzp1

"For example, the relationship of modern humans and Neanderthals may not be of admixture but AMH drifting away from a common ancestor, hence varying levels of Neanderthal alleles remaining in modern humans."

There are ancient modern humans including the ones from this study that contain Neandertal DNA in blocks. There is linkage disequilibrium. This is definitely admixture.

mzp1 said...

In my earlier post Morocco Iberomaurusian is much closer to Yoruba vs Mbuti compared with Iran_GanjDareh_N. The single d calc shows

d (Mbuti, Yoruba) (Iran_GanjDareh_N, Morocco_Iberomaurusian) = 0.0172

and only Ancient North Eurasians MA1, Ust_Ishim, Kostenki and Goyet score higher.

So I think the best explanation is a Southward migration of North Eurasian ancestry due to the LGM. This can affect all of South Eurasia and can be the cause of the large differences for South Asians on the above calc.

Archi said...


Haplogroup E can in no way be from Africa, because there was a very ancient reverse migration to Africa from Eurasia, and it could only be people with haplogroup E and probably with D relics, no one else.

Rob said...

@ archi

“ In Siberia in those days there were Mammoth Steppes with a great climate, much better than in Europe and Anatolia. Europe was covered by glaciers; there was no glacier in Western Siberia. The amount of food in the Mammoth Steppes was probably greater than in the savannahs.”

That’s true; but that’s only relevant for a much later period - LGM 26-19,000 BP. This is why & when the ANE migration to Europe occurred

But the IUP happened during a relative warm conditions, and the timing of 47,000+ in the west is earlier than anything in Mongolia or China. Siberia & Altai is synchronous with Europe . South Asia is irrelevant ; and the inland parts of Central Asia and Iran are too late
Anything older than 45,000 in altai is attributed to Denisovans
We remain open minded; but at present the route sketches from Europe to Siberia. The other way lacks likelihood; we have no way of getting AMH there unless we believe AMH speciated in Siberia or America.

Archi said...

@rob

"but that’s only relevant for a much later period - LGM 26-19,000 BP. This is why & when the ANE migration to Europe occurred
But the IUP happened during a relative warm conditions, and the timing of 47,000+ in the west is earlier than anything in Mongolia or China."

That is absolutely not true. It is a mistake. Гt is relevant to the entire Ice Age.
https://upload.wikimedia.org/wikipedia/commons/b/b8/Vostok_Petit_data.svg

What does Mongolia and China have to do with it? You are making a big mistake. We are talking about the territory of Russia, in the Altai IUP dating is even older than in the Middle East and certainly older than in Europe.

"Anything older than 45,000 in altai is attributed to Denisovans "
That is absolutely not true. It is a mistake.

"We remain open minded; but at present the route sketches from Europe to Siberia. The other way lacks likelihood; we have no way of getting AMH there unless we believe AMH speciated in Siberia or America."

Such a route does not exist at all and has never existed for any archaeologist or anyone, nor does it exist in this paper. You are prejudiced and you are the only one who made it up. Don't write we about yourself, you are the only one. The paper says exactly the area of distribution and nothing more.

Rob said...

@ Archie
Lol I’m prejudiced ? Nothing you ever proclaim ever pans out, and even when the data solidly debunks all your whacky theories ; you still don’t get it . Pipe down

Rob said...

@ Slumberry

Perhaps D & E split in Arabia; which was a putative pre-IUP homeland; or in other words, where the parental populations of Sukul & Qafzeh were ~ 100 kyBP.
Hg D might be associated with an earlier southern radiation, whilst E might have back-migrated to Africa ~ 50 kybp

But Hg E looks to have diversified and moved out of NE Africa during the late Pleistocene, perhaps as a series of founder effects after the LGM
Curiously, the pre-EV13 of Europe is closer to iberomaurisians rather than Natufian E-M35
Now I see there is a Capsian genome which is also on the ~ E-M35 line

Slumbery said...

@Rob

Fair enough. After all such a long time (with a lot of regional migrations) passed that E lineages could have multiple dispersions and local extinctions. The current distribution might be very misleading.
BTW this reminds me of Dienekes. His theory was that the final steps of formation of modern humans before global dispersal had happened at least partly in Arabia, including the start of CT diversification.

Archi said...

@rob

Exactly that you are prejudiced, you have always written about it. you use science as a tool for propaganda, you write about yourself we and call yourself a god, everything you write is refuted from first to last word.

Rob said...

@ Slumberry

I can't recall what Dienekes theory was exactly. I rather refer to the increasing interest around Arabia by archaeologists and its possible links between the African Nubian complex and emerging Emiran (proto-IUP) in the Near East, on the one hand, and the southern route dispersal, on the other. It's all still developing days.
But it is clear that after 50 kyBp, there was a rapid radiation of AMHs around the globe (the question of an earlier pre-Toba southern dispersal aside); but we need to understand what precipitated this. It's almost supernatural ? gift from gods



@ Archie
yes you're the champion of wokeness. Keep licking your wounds as the prejudiced look to evidence instead of imagining that everything is linked to R1a or comes from Siberia. Your 7D ''theories'' must take a whole 10 seconds to construe.

Chad said...

I don't have Zlaty Kun yet. If someone would be willing to send them to me, I'll join in. My thinking just from reading the paper and my own experience, I'll say that Zlaty Kun would be Basal Eurasian. Anything deeper is secondary and maybe a third movement from Africa.

Goyet shows a significant relationship to Ust Ishim relative to about anything. East Asians too. Goyet will likely be an Ust Ishim and Bacho Kiro mix, with the least to no Zlaty Kun. Later Europeans have additional Zlaty Kun and later some more Near Eastern admixture.

Bacho Kiro may be related to the SE branch that mixes into Oceania.

Zlaty Kun is likely in Dzudzuana, causing that Gravettian relationship. Pinarbasi probably has less than Dzudzuana due to admixture with an HG group of more Eastern affinities. I think Pinarbasi will be our oldest with a second African wave that is involved with the formation of Iberomaurusians/Oranian and Natufians.

Natufians may show a third African wave that is most closely related to Mota, which is also admixed into Iberomaurusians. This last may be the Y link.

Chad said...

Also seeing an Onge wave by the time of Ganj Dareh and Boncuklu.

Andrzejewski said...

Speaking of Y-dna E and its spread and origin, is it possible that it arrived in North Africa as a back migration from West Asian Anatolian (or Levantine) Neolithic farmers? I read here on a different blog entry that during the Neolithic North Africans had approximately 40% of their aDNA constituting Neolithic Anatolian. Thereafter, their rate might’ve only actually *increased*, owing to the Bell Beaker, Punic, Roman and Vandal migration. Since we know that current Maghreb pops are Arabized Berbers with a very small Arab veneer (the latter also have somewhat Levantine/Anatolian mix of course), how much ANF is currently extant within North Africans?