search this blog

Friday, May 8, 2015

Ancient DNA from an Upper Paleolithic European with recent Neanderthal ancestry


Arguably the biggest news from this week's Biology of Genomes conference at the Cold Spring Harbor Laboratory is this:

Team Characterizing DNA from Ancient Human with Recent Neanderthal Ancestry

The article is free, but you might need to register at GenomeWeb to access it. This is basically what it says:

- the male sample, Oase 1, is 37-42K years old and comes from the Pestera cu Oase site in southwestern Romania

- it's estimated to harbor 5-11% of genome-wide Neanderthal ancestry, with as much as 50% on chromosome 12

- the admixture is in relatively long stretches, which suggests that the mixture took place four to six generations before Oase 1 was alive

- it's similar to present-day Eurasians, and apparently from a population that was on the way to becoming European, but probably did not contribute in any significant way to the genetic makeup of present-day Europeans

- thus far, the capture method used by the scientists has managed to snag around 78,000 genome-wide sites (presumably SNPs), but the sequencing efforts continue, and we might see a Y-haplogroup result in the near future

It sounds to me as if Oase 1 is similar to Kostenki14 (see here), so I'd say there's a good chance its Y-chromosome belongs to haplogroup C-M130.

Update 23/6/205: Oase 1: An early modern human from Romania with a recent Neanderthal ancestor

82 comments:

Krefter said...

Interesting. But I'm kind of tired of European-only ancient DNA.

He's "37,000 to 42,000 years ago", so a few thousand years older than K14 and a few thousand years younger than Ush_Ist.

Sounds like he was Crown Eurasian-ish, and part of or related to the "West Eurasian" branch.

Here are some quotes from the article I thought were interesting.

"Based on the SNP patterns detected in the sample, the researchers estimated that the individual had a Neanderthal ancestor within the past four to six generations"

"Meanwhile, comparisons between genetic variants in Oase 1 and those in present-day populations or previously sequenced ancient samples suggested that the ancient individual from Romania belonged to a population that was becoming somewhat European."

"But while this group resembled both European and Asian populations, Fu noted, it appears to have been far removed from agricultural populations in Europe and does not appear to have contributed much genetically to present-day human populations."

"The team is continuing to tease apart patterns from genetic profiles in the sample, including genotyping analyses of the Oase 1 individual's Y chromosome, she said."

Krefter said...

Oase 1 will probably come out to be some type of HIJK or C.

It'd be neat if he came out IJ(which is possible), because then we'd have R*, C1*, and IJ* from Upper Palaeolithic North Eurasians who were autosomally ancestral to R1b/a, I, and C1a2 Mesolithic Europeans.

Mike Thomas said...

!!!! Excellent
I bet he's either C or IJ
But my money is on C

Maju said...

Very interesting info, David, thank you. Some anomaly in Oase 1 was already suspected based on anthropometric data but it is indeed fascinating that he has been confirmed to be so close to a Neanderthal admixture event. Let's not forget that Oase 1 is the main MP-UP transitional remain, as burials did not became common, it seems, until Gravettian.

Krefter said...

@Maju,

Oase 1 wasn't Gravettian right? The article seems to state he's related to non-Neolithic ancient Europeans. So, related to K14.

Do you have any ideas why that is?

Mike Thomas said...

Oase was from c. 40 ky BP. So definitely not Gravettian. From what I remember, he was found without 'context', but must be related to the pre-Aurignacian / bacho Kirian , or Aurignacian proper. Either way, he must be one of the first "Europeans"

K14 was found from an Aurignacian layer (Kostenki 14 - III) so he's relation to Oase is only to be highly suspected.

Maju said...

@Krefter: What Mike Thomas said: from the chronology it should be Aurignacian or otherwise "Aurignacoid", however the remains are not associated to any specific cultural materials.

I didn't notice any specifics re. what exact kind of association did Pestera cu Oase had with other Paleolithic Europeans. I would expect the individual to be partly "ancestral" (senso lato) to later Paleolithic Europeans and therefore to us as well, but not sure exactly how. Also the late Neanderthal admixture may well be something very specific of that group and I would not necessarily extend the findings to other early UP Europeans (AFAIK there was no such thing in Uluzzian for example, nor I recall any such thing when dealing with Kostenki either).

Carlos said...

Why necessarily IJK or C etc.? Maybe it's finally the chance to get Neanderthal YDNA? :)

Mike Thomas said...

@ Maju

I wasn;t aware they had any Ulizian genomes, mere mtDNA RFLP, or otherwise.

Yes, PcO was likely ancestral to UP Europeans, but i think the article implies modern Europeans are much 'descended' from him. Ie you have 35k years of demographic impacts from "Gravettians" (if they were in fact a distinct population), Ice Age, neolithic exapnsion, and the LN/EBA events.

@ Carlos

PcO was homo sapiens, so his Y DNA would surely fall within modern human range.

Carlos said...

@ Mike Thomas

He was phenotipically homo sapiens, all right, but does it exclude the possibility that his neanderthal ancestor 6 generations earlier (as the article claims) was his direct male ancestor?

Mike Thomas said...

Carlos

I see your point. Maybe. Can't wait to find out though.

Maju said...

@Mike: the Uluzzian aDNA is a recent news: http://anthropology.net/2015/04/23/were-protoaurignacians-modern-humans-or-neandertals/

Pay-per-view paper: http://www.sciencemag.org/content/early/2015/04/22/science.aaa2773

From the abstract: The lower deciduous incisor from Riparo Bombrini is modern human, based on its morphology. The upper deciduous incisor from Grotta di Fumane contains ancient mitochondrial DNA of a modern human type.

Unsure what haplogroup it may be however.

Maju said...

PS- Sorry, it was not Uluzzian (South Italy) but Proto-Aurignacian (North Italy). My bad. Previous studies of Uluzzian teeth have also supported H. sapiens morphology but there does not seem to be any genetic sequence yet.

Vincent said...

"- equally close to pre-agriculture europeans and Asians, further away from modern european human / post-farming"

http://alistairboettiger.info/wordpress/

What's a better description of Y-DNA E's position in the Y-haplotree?

Davidski said...

What? You think Y-DNA E was present in pre-agriculture Europeans?

No chance.

Vincent said...

As I have been saying for ages, I think E* was the first to enter Europe (through the Balkan route from the Levant) but was soon overwhelmed by the arrival of C* and I* guys, so that immediately before the LGM it became extinct due to a serious bottleneck. Morphological differences between pre-LGM and post-LGM Paleolithic Europeans are notable and well-established. Also, KNM-LH1 specimen from Kenyan Upper Paleolithic (23000 years ago) shows it is similar to Oase 1 and dissimilar from modern-day Negroids. Same story with Hofmeyr and Manot 1. All of these guys were once Basal Eurasians from Arabia/Levant IMO. Proto-Aurignacian/Bacho-Kirian has its origins in the southern Levant anyway, so E is a serious candidate. I don't exclude DE* or even D for Oase 1 though

Davidski said...

E is not a serious candidate.

It's a Neolithic and post-Neolithic marker in Europe.

Vincent said...

In fact it was re-introduced

Davidski said...

Bullshit.

Vincent said...

Explain why Kostenki 14 had considerable Basal Eurasian admixture while his mate in Central Asia didn't have any, so we'll see if this is bullshit or not. I only explain this as a result of increased contact between the Basal Eurasians that inhabited Europe and Central Asian UP populations

Mike Thomas said...

Well anything is possible I guess, but the exact "basalness" in K14 requires elaboration. He just looks like a basal west eurasian to me, excluding southwest ( Levantine-NTh African) groups.

Also the entry of E* as the earliest Y marker lack parasimony. It requires a rather contorted migration history

Vincent said...

E* as earliest doesn't lack parsimony at all. E* had to be in West Eurasia at one point, because it clearly back-migrated to Africa. During this back-migration it may have bifurcated and may have reached Europe. But even rejecting the back-migration hypothesis, who do you think inhabited the Levant 50000 years ago if not E*? The answer is: nobody else.

Davidski said...

Kostenki14 doesn't have the same Basal Eurasian admixture as early European farmers. It's just a really ancient genome.

Oase 1 doesn't have any Basal Eurasian admixture either. Did you miss this quote from the GenomeWeb article?

"It appears to have been far removed from agricultural populations in Europe".

Forget it. There's nothing here suggesting Y-DNA E.

Vincent said...

The quote from Alistair Boettiger says that Oase 1 was also distant from pre-agricultural Europeans and Asians alike, while until now Paleolithic genomes have always shown a tendency towards either West or East Eurasians.

Vincent said...

"Kostenki14 doesn't have the same Basal Eurasian admixture as early European farmers. It's just a really ancient genome"

It's only you who is saying this

Davidski said...

The article clearly says that Oase 1 is similar to pre-agricultural Europeans (you know, like hunter-gatherers) and Asians, but distant from early European farmers.

And Kostenki14 certainly doesn't have any of the Basal Eurasian that early European farmers carried, not only because I have the sample and so I can see this, but because this was shown in Haak et al. See Figure S8.3.

There was no Basal Eurasian admixture in pre-Neolithic Europe, and no E either.

jackson_montgomery_devoni said...

When it comes to Y-DNA this ancient individual is likely to be C, I or IJ and when it comes to mtDNA it is likely that he will belong to either haplogroup R or U.

Vincent said...

I trust more Sikora et al. (2014) and especially Alistair Boettiger, who has more info on Oase 1 than the article you posted

Vincent said...

Also, the article says that Oase 1 didn't contribute to modern human populations [from Eurasia, presumably], instead K14 contributed significantly

Davidski said...

Sikora didn't test their model formally.

The claim that Kostenki14 has Basal Eurasian admixture of the kind found in early Europeans farmers is bizarre and wrong.

And not even Sikora claimed that Kostenki14 contributed significantly to modern Europeans, so I have no idea where you pulled that out of. Probably a hat, or maybe your ass.

Davidski said...

Here's what Boettiger wrote on his blog:

"equally close to pre-agriculture europeans and Asians, further away from modern european human / post-farming"

That's basically what the news article says:

"But while this group resembled both European and Asian populations, Fu noted, it appears to have been far removed from agricultural populations in Europe and does not appear to have contributed much genetically to present-day human populations."

In other words, similar to European hunter-gatherers but very different from European farmers.

What differentiators European hunter-gatherers from farmers? Basal Eurasian admixture, yes or no?

Vincent said...

"The claim that Kostenki14 has Basal Eurasian admixture of the kind found in early Europeans farmers is bizarre and wrong"

Only because you and people similar to you (Haak, for example) don't like it. Delusional.

"And not even Sikora claimed that Kostenki14 contributed significantly to modern Europeans, so I have no idea where you pulled that out of. Probably a hat, or maybe your ass"

You are both rude and seemingly blind. Read the abstract from the paper and tell me if they don't say it. They say that K14 "shares a close ancestry with... European Mesolithic hunter-gatherers... and many Europeans".

Vincent said...

"But while this group resembled both European and Asian populations, Fu noted, it appears to have been far removed from agricultural populations in Europe and does not appear to have contributed much genetically to present-day human populations."

"In other words, similar to European hunter-gatherers but very different from European farmers."

Nothing about Oase 1's similarity with pre-agricultural Europeans is named.
Boettiger says that Oase 1 was equally close to pre-agriculture Europeans and Asians, but he then says "*further away* from modern Europeans". This means that Oase 1 was not close to Europeans and Asians in an absolute way, but only relatively.

Maju said...

I'm with Davidski in this one, at least in the essentials: the lineages arriving to West Eurasia (including Europe) with the early UP must belong to the Eurasian branches, what in practical terms means IJ and P1 (G and T also in West Asia) for Y-DNA and U and HV (JT, N1, X, W and M1 also in West Asia/NE Africa) for mtDNA. And that's what we see in the few pre-LGM data we have (add the C1 oddball of La Braña and maybe others of the kind but still Eurasian).

Said that, it's possible that E1b-M81 arrived to West Iberia in the LGM, but that's not "the essentials", rather a localized and relatively late exception.

The bulk of E (E1b only) arrived to Europe with the Neolithic.

"... immediately before the LGM it became extinct due to a serious bottleneck".

How convenient. Fairy stories to me: nothing to back it up, neither modern nor ancient, neither genetic nor archaeological, just wild speculation without any base.

Also your links don't work for me, Vincent.

Vincent said...

"How convenient. Fairy stories to me: nothing to back it up, neither modern nor ancient, neither genetic nor archaeological, just wild speculation without any base."

Plenty of evidence for the replacement of EUP populations.
http://onlinelibrary.wiley.com/doi/10.1002/ajpa.20950/abstract
http://horizon-magazine.eu/article/ice-age-europeans-roamed-small-bands-fewer-30-brink-extinction_en.html
See also the previous links about the similarity between UP individuals from the Levant/Kenya and the earliest UP Europeans

I'm still waiting for an answer to my question. Anyone is welcome to tell me what haplogroup was in the southern Levant 50000 years ago if not E*. E* is clearly Proto-Aurignacian, there's no evidence for any other lineage.

Maju said...

"They say that K14 "shares a close ancestry with... European Mesolithic hunter-gatherers... and many Europeans"".

Yes, K14 was West Eurasian and necessarily he or his relatives left a legacy among later Europeans and West Asians, leading eventually to us. How does this anything to do with Y-DNA E*?

"Nothing about Oase 1's similarity with pre-agricultural Europeans is named". And then: "Boettiger says that Oase 1 was equally close to pre-agriculture Europeans and Asians"..

Do you realize your contradictions?

"but he then says "*further away* from modern Europeans". This means that Oase 1 was not close to Europeans and Asians in an absolute way, but only relatively".

No. It obviously means that Oase 1 was closer to Paleolithic Europeans than to modern Europeans, who have important West Asian admixture (Neolithic). So it means that Oase 1 was significantly closer to (Epi-)Paleolithic Europeans than to Paleolithic (and Neolithic) West Asians. Simple, straightforward and also making good sense, because, other than the issue of Gravettian (surely a secondary wave from West Asia at the origin), there is continuity between Aurignacian and Epi-Paleolithic in Europe. In fact Magdalenian is like "Aurignacian 2.0", a refined revival of Aurignacian traditions.

Vincent said...

"So it means that Oase 1 was significantly closer to (Epi-)Paleolithic Europeans than to Paleolithic (and Neolithic) West Asians."

Where did you find that "West" in West Asians? You made it up, simple, because in these cases "Asians" alone means East Asians.

Maju said...

The Pinhasi paper is only talking of Central Europe ("the genomes sequenced from hunter-gatherers from Hungary and Switzerland"), where a near-extinction of the population has long been held by archaeological evidence. It's not about Europe at large but just Central Europe: no Franco-Cantabrian, nor South Iberian, nor Italian nor East European ancient genomes were considered.

Judging by the excellent analysis of Bocquet-Appel 2005, Central Europe was almost emptied in the LGM, with maybe as few as 300 people (actual pop., not effective pop.) surviving in Moravia (no other sites are known in that period) or up to 4500 (largest CI). By comparison in the Franco-Cantabrian region there were between 1500 (min.) and 24,000 (max.)

The article is imprecise about the actual size of the overall population anyhow. It talks of "small bands" of 20-30 people, what is absolutely normal for hunter-gatherers: it's the operative group figure, the band. Bands are not endogamous: bands belong to much larger "nations" or "tribes" and are exogamous within these (and more rarely outside them). In fact bands are kind of volatile, according to anthropological evidence: they form and reform once and again, much depending on the whim of their members.

For example, if we take the Franco-Cantabrian region in most periods, we should probably consider the whole area (or at least large sections of it) as a "tribe" or "nation" and the local bands exploiting each district as those 30-100 people bands which constituted its operative units. The size of the area exploited by one such band in Magdalenian Biscay would be roughly that of a large "comarca" (district), including coast and mountain, and sometimes crossing the mountain line to the south (surely only in summer, else too cold) for specific resources like flint stone.

Interactions between district bands were common: whale bone spear points have been found in Isturitze, many kilometers inland. When people went to Treviño to get flint in summer, they surely arrived from several districts (bands). We can't know for sure but most likely they held yearly festivals in order to keep their relationship strong and also find mates from different groups. Pretending that small bands were isolated and endogamous is absurd and contradictory with all we know. Just that Central Europe and possibly also other areas did suffer a major bottleneck (near-extinction event) in the LGM. But not the Southwest, which effectively acted as refuge.

Maju said...

"Where did you find that "West" in West Asians? You made it up, simple, because in these cases "Asians" alone means East Asians."

In Neolithic and Modern Europeans. IF Oase 1 was closer to Epipaleolithic than Neolithic/Modern Europeans, then he was closer to Paleo-Europeans than to West Asians, because that's the main difference between a Magdalenian and me: that I am more West Asian.

spagetiMeatball said...

Why do they always use post-farming europeans in these studies.

What's so hard about comparing these ancient samples to, you know, living near easterners, who are even more basal eurasian than ancient european farmers.

Vincent said...

Still, physical/morphological differences between pre-LGM and post-LGM Paleolithic Europeans are unexplained if a replacement is not assumed.

Romulus said...

It would be interesting and give some context to this to know the estimated Neanderthal ancestry in the Mesolithic European samples.

Mike Thomas said...

@ Vincent

You're making wild generalisations. Eg

"Levant 50000 years ago if not E*. E* is clearly Proto-Aurignacian, there's no evidence for any other lineage."

There's no evidence for *any* lineage yet. I don't understand your logic.

"physical/morphological differences between pre-LGM and post-LGM Paleolithic Europeans are unexplained if a replacement is not assumed."

Where in Europe? Iberia , Central ? Eastern ? Balkans ?
Morphological change could be due to many things apart from replacement - like drift, selection and gradual admixture.
In fact, the Krause website suggests that pre and post-LGM europe was very similar genetically. Ie depopulated North was recolonized by Southern Europeans and Eastern Europeans, and not an exodus from northern Africa.


Chad Rohlfsen said...

Eurasia K10. Doesn't look too bad. I may tweak it some more.

https://docs.google.com/spreadsheets/d/1qswdj-Mys4EUD-LMJ53TSWhHTqkVRk32Jj9UzcubP38/edit?usp=sharing

Vincent said...

"There's no evidence for *any* lineage yet. I don't understand your logic"

Here's a description of Manot 1 (55 kya, Israel):

"It is *similar* in shape to *recent African skulls* as well as to *European skulls from the Upper Palaeolithic period*, but different from most other early anatomically modern humans [Skhul, Qafzeh] in the Levant."

To understand my logic, you have to understand the phylogeny of Y-DNA haplogroups. Y-DNA CF originated in or around South Asia at least 65000 years ago, as evidenced by several cases of F* and C* paragroups in the region, and also by aDNA (Mal'ta, K14 and Ust'-Ishim both show a lot of South Asian admixture). Y-DNA DE* certainly originated in Eurasia, and there is countless evidence for this, notably the complete lack of African admixture in D-bearing Asians. It probably originated in West or Southwest Asia. Therefore DE*/pre-E* was closer, MUCH closer to the Levant and is by consequence a much more serious candidate for Levantine Ahmarian/proto-Aurignacian, even if you don't accept the fact that CF originated in South Asia. To reinforce the case for E* proper, the highest frequency of the paragroup is found in North-Central Arabia (Abu-Amero et al., 2009).

Mike Thomas said...

Vincent I fully agree that DE * originated in he levant, and then E* migrated back to africa, but its a bit of logical leap that it was a, or the aurignacian marker.

And I don't have much faith in skull shapes

Vincent said...

Ok, tell me why E* couldn't have been the/a proto-Aurignacian marker, instead of simply saying that it couldn't

Mike Thomas said...

There is no a priori reason why, Vince. As I stated from the outset it is possible. And we're all just speculating on scraps of a magazine abstract. But I think its less likely than C or IJ because
: we know C was presebt in aurignacian EE (K14), which came from Central Europe.
: C was still present in mesolithic and early neolithic
: I was dominant in mesolithic central -western Europe; meaning it must have been in Europe since at least the UP, because the UP in anatolia is poorly represented, and it certainly didn't fly over by plane
: Hg E is a southwest eurasian marker, dominant in NE Africa. I'd entertain a mesolithic arrival of E in Europe, but not UP.

Davidski said...

Kostenki14 doesn't have any Basal Eurasian or Near Eastern admixture, and neither does Oase 1.

They're both more closely related to Loschbour and La Brana-1 than to anyone else.

Kostenki14 is C, and so is La Brana-1. Oase 1 will also be C. You won't find any E or Near Eastern admixture in pre-Neolithic Europe.

Vincent said...

"we know C was presebt in aurignacian EE (K14)"

K14 was not Aurignacian. The site of Kostenki 14 is generally regarded as Undetermined Upper Paleolithic by archeologists

Vincent said...

"Kostenki14 doesn't have any Basal Eurasian or Near Eastern admixture, and neither does Oase 1."

False.

"They're both more closely related to Loschbour and La Brana-1 than to anyone else."

False. They are equally similar to pre-agriculture Europeans and Asians.

"Kostenki14 is C, and so is La Brana-1. Oase 1 will also be C. You won't find any E or Near Eastern admixture in pre-Neolithic Europe."

You jump too quickly to conclusions.
If Oase 1 comes out not having E*, but having significant Basal Eurasian/Near Eastern (at least more so than K14) it would be a confirmation of E*'s presence among the earliest UP Europeans.

Mike Thomas said...

No. It has clear Aurignacian elements, albeit with local variations thereof

Davidski said...

Vincent you fruitcake,

Kostenki14 is even more similar to modern northeast Europeans than to Asians. Red = more similar, see?

https://drive.google.com/file/d/0B9o3EYTdM8lQWHZHRGZ3MTNDOTg/view?usp=sharing

But Kostenki14 is even more similar to Loschbour and La Brana-1 than to northeastern Europeans. This is very clearly seen in Fig. S16 from Sikora et al.

Loschbour and La Brana-1 are pre-agricultural Europeans, because they're hunter-gatherers.

So what does E have to do with Northeast Europeans, pre-agricultural Europeans and East Asians? How did you manage to make this remarkable link?

Krefter said...

What's causing East Asians to be more distant from K14 than to WHG/ANE is probably not the "Basal Eurasian" in EEF.

This is because K14 shows the same trend of relation to modern Eurasians as WHG/ANE, excluding recent drift(Ma1 relation in native Americans, Loschbour relation in EEF and modern Euros).

K14 doesn't have any special relation to EEF and Middle Easterns. If he had the same distant basal Eurasian ancestor, plus "West Eurasian", you'd expect him to.

Maju said...

@Vincent:

1. Morphology: I believe you mean the issue with Crô-Magnon 1 being a particularly robust man. It must be said that other "cromagnons" (from Iberia, NW Africa) are much more gracile. I believe you also mean the issue of claims of Late UP Europeans (Magdalenians) being smaller. They are all females incidentally, although there may be some residual truth to the claim. Anyhow we are talking of processes that took tens of thousands of years, so per se morphology (always slippery terrain) does not need new immigrant waves, waves that are not documented after Gravettian anyhow.

2. Re. age(CF). CF MUST be c. 100 Ka old, not 65 Ka. By that time K2 had already coalesced and scattered. Your reliance on the "molecular clock" miscalibrated speculations is confusing your thought.

3. Manot 1: is indeed a modern West Eurasian, just like Crô-Magnon 1, etc. And unlike the early Palestinian skulls of c. 125 Ka ago. These in some cases may show signs of admixture with Neanderthal, which, if confirmed, would illustrate the main Neanderthal admixture episode at the beginnings of the Out-of-Africa migration (Arabian-Palestinian Pluvial stage, c. 125-100 Ka BP).

4. DE: the presence of DE* in several locations in Africa evidences that, most likely, DE coalesced in Africa, with pre-D joining the CF-dominated OoA population at a late stage and getting later a founder effect in East Asia. It's much easier to see it via mtDNA though because there was indeed some process in which E became absolutely hegemonic in NE Africa (in the mtDNA L2-L3 dominated populations) but not in Arabia (CF with mt L3 and L0 derivates, no traces of early E* in Arabia either).

5. Kostenki: "The site of Kostenki 14 is generally regarded as Undetermined Upper Paleolithic by archeologists". That's not quite true: pre-Aurignacian early UP in Kostenki is "undetermined" (other "Aurignacoid" arriving via the Caucasus, Mezhmaiskaya related it seems) but after the Campanian-Ignimbrite catastrophe (HE4), Aurignacian proper became the only remaining culture in all Europe (barring some late Neanderthal pockets), including the Kostenki area.

Vincent said...

"4. DE: the presence of DE* in several locations in Africa evidences that, most likely, DE coalesced in Africa, with pre-D joining the CF-dominated OoA population at a late stage and getting later a founder effect in East Asia."

Tell me, where's SSA in D-dominated areas of Asia? If DE is African we have to see its effects in Tibet, the Andaman Islands and Japan, but frankly I don't see them at all. Instead I see that all E-dominated Sub-Saharan Africans carry Eurasian admixture.

Matt said...

Ust Ishim is a little older than Kostenki and has much flatter relatedness to world populations, despite being at the Altai.

PcO1 (the mandible sample) could be between K and UI in terms of how much affinity it has to world populations, as AFAIK its temporally older than K14 and not as old as UI. Especially with that higher level of Neanderthal admix (which East Asians have a little more of, but not much more of).

So it could be evenly or almost evenly related to East Asians and pre-agricultural Europeans (or even more related to present day East Asians than present day Europeans).

It's hard to know if we're reading too much into the lines from the article which are vague on affinity except that they're elevated to pre-ag Europeans and present day East Asians. His Y-dna group could be pretty much anything though, provided the mutational date for that clade is early enough.

Tianyuan man, although the sample quality is not as good, unfortunately, also has this quality of low specific relatedness to close by people from the present.

Vincent said...

Maju, D was probably the earliest in Asia too. I don't care if you consider SNP-dating or not, but D-L136, so far only found in the Philippines but probably representing the majority or totality of D* cases in mainland Asia, is an estimated ~51 ky old, much more ancient than any other lineage in SE Asia.

So, 1) quit saying that D *joined* CF guys (all places with high D seem to have only that lineage as indigenous: see Tibetans, Japanese and especially Onge) and that 2) it did so *later*.

And CF IS NOT 100 ky old. This is absolutely absurd and impossible. The same can be said for all your estimates.

Mike Thomas said...

But Tinyuan is clealry related if not directly amcestral to East asians isn't he ?

It's also interesting to not e that of the basal components in Ust-Ishm and K14; the "red sea" in K15s is missing as is "west asian". Are these later developments ?

Matt said...

Mike, it's complicated.

It's true that when they put him in a dendrogram Tianyuan man fell basal to the clade of other ENA (Han, Dai, Karitiana, Papuan) so is less related to them than they are to one another.

http://2.bp.blogspot.com/-o4Zid3z0beg/UQAzDIVkN0I/AAAAAAAABjM/DtQpe0CiSCU/s1600/TianyuanAutosomal.png

And he had a y-dna group which linked him to ENA.

At the same time, there are caveats.

Tianyuan is not reliably closer to many ENA populations when you look at nucleotide differences, e.g. ND to French 23,168 while to Dai is 23,339, Sardinian 22,906 and 22,802 Han and 21,944 Karitiana.
http://4.bp.blogspot.com/-Ygeyjg0KKjM/UQA-xfvF89I/AAAAAAAABjo/RY19NiL3ETg/s1600/TianyuanPairwise.png

So the tree position may be mostly a consequence of other factors than an especially closer relationship to ENA by Tianyuan, about how the other populations relate to one another, e.g. essentially how ENA are bound together and ENA further from Africans than West Eurasians are, stuff like that. The tree position of Tianyuan is essentially a compromise that depends on a lot more factors than just who it is closest to.

Also, crucially, the Tianyuan studies were before we had pre-agricultural European and Siberian genetics, so Tianyuan has not been tested wrt to Ust Ishim, MA-1, pre-agricultural Europeans like Loschbour and MA1. And the Tianyuan work was before our model for Native Americans (who seem to be marginally closest to Tianyuan) was ANE+ENA.

Ust Ishim in theory is equally closer to pre-agricultural Europeans, MA-1 and East Asians, depressed in its affinity to present day West Eurasians by lack of Basal Eurasian, as per Reich Lab (that's the working model atm, although K14 and UI's relationships probably complicate this so whether it ultimately works out once we have more adna.) Tianyuan could well be the same - equally related to all Crown Eurasians, less to present day Europeans he's basically the same age, and population turnovers are not necessarily less likely where he lived.

Sequence quality may not be good enough to test Tianyuan against other ancient dna, unfortunately.
I wouldn't get too hung up on that specific example, Tianyuan is the worst adna we have and its hard to test a lot of hypotheses about it - just the special relationship with Tianyuan and East Asians was somewhat complicated at the time, and has got much more so with more adna, especially Fu's later work on Ust Ishim as a sequence of comparable age. There are still lots of questions about population turnover in East Asia.

Tobus said...

They also only sampled a single chromosome of Tianyuan - chromosome 22, one of the smallest ones. So the comparisons we have are only on 1.5-2% of the genome and may not be representative of the whole.

Maju said...

@Vincent: it has nothing to do with *recent* African origins: D is as African as CF is or in general non-African humankind is, it's something to be dated to the OoA process c. 125-100 Ka ago.

"Maju, D was probably the earliest in Asia too".

Why? Why not just contemporary with C, which also seems centered in SE Asia?

"all places with high D seem to have only that lineage as indigenous"

Fixation after founder effect. If you look at associated mtDNA and not just at Y-DNA, you see it clearly. Why do people who only look at Y-DNA get everything so wrong? Because Y-DNA is misleading... unless you integrate it with the rest of the data.

"And CF IS NOT 100 ky old. This is absolutely absurd and impossible".

100-125 Ka old. It is the only logical date because that's the only archaeologically proven Out of Africa migration: c. 125 Ka BP in Arabia-Palestine, c. 100-95 Ka BP in South and East Asia, c. 80-60 Ka BP in Australia, c. 50-45 Ka BP in Europe and Siberia (and back to NE Africa), c. 17 Ka BP (or older) in America. Those are the archaeo-facts. Add to them the c. 74 Ka BP Toba event, which seems very clearly related to the Y-DNA K2 and mtDNA R expansion from SE Asia.

Those who take at face value molecular-clock-o-logical speculations end up in a total nonsense of scheme, that does not fit the real archaeological chronology in any way, nor the divergence age from Chimpanzee c. 10-13 Ma ago, nor anything that makes sense. And that is because molecular-clock-o-logical archaeo-references have remained stuck in the 1990s, while real archaeology and paleontology has advanced a lot in the 21st century, providing some very important new info that pushes back (a lot) the hominid evolution and definitely the Out-of-Africa. A few geneticists have gone this way (Langergraeber for example) but most just parrot totally dated references with no value whatsoever anymore.

Besides, it just seems logical that the only windows for out of Africa or back into Africa migrations are the pluvial periods and there is no pluvial of any sort between 90 and 50 Ka BP: crossing Arabia at that time would have not worked. The Abbassia Pluvial (125-90 Ka BP is the time of the Out-of-Africa), the Mousterian Pluvial (c. 50 Ka BP) is the time of the back from India (and SE Asia) to the West, including parts of Africa (again deserts to be crossed are a problem) and finally the Early Holocene semi-pluvial is the time of some extra bidirectional flows outside and inside Africa (as well as into Arabia, etc.) related in the Mesolithic and Neolithic.

Get real!

Maju said...

@Mike: in mtDNA at least Tianyuan is clearly within East Asian variability, although the sublineage as such does not exist anymore (apparently). Tianyuan suggest to me not a "modern East Asian" population but an East Asian population in any case. While they may not be directly related I'd compare with the Ainu, for example, who are also outliers but still very much East Asians. I would think that people of the type of Tianyuan were the East Asian ancestors of Native Americans (other ancestors are Western of the Ma1 type). Later, after the LGM, new waves altered the landscape, especially so far North (the permafrost reached to Beijing in that climatic extreme episode, so it was surely similar to what we see in Central Europe, regardless that some groups could survive in coastal areas or wherever).

Matt said...

Chad: Eurasia K10. Doesn't look too bad. I may tweak it some more.
https://docs.google.com/spreadsheets/d/1qswdj-Mys4EUD-LMJ53TSWhHTqkVRk32Jj9UzcubP38/edit?usp=sharing


Interesting ADMIXTURE. And again like another of your experiments it is capturing divergence between EN populations and Sardinians and gene flow between Sardinians from steppe, WHG and Near East that sets them apart from EN (and as shown by Haak).

And the intra European clines show up well, that can be noisy on other cluster analysis (e.g. the distinct clines through SW Europe to Basques and SE Europe to East Central Europe seemed a little less distinct on some of the other calculators).

It seems like you had to prune the choice of samples pretty harshly to make admixture pop out the those clusters though - even in West Eurasia no MN and no LNBA, other than HungaryGamba, proportionately few West Asians other than Turks and North Caucasians, no Hindu Kush, really pretty limited Levantine / Arabian peninsula.

Kristiina said...

Matt, you say that Tianyuan man "had a y-dna group which linked him to ENA". I am not aware of his yDNA, so could you update me on this.

Matt said...

Sorry, Kristiina, he was only mtdna tested, my bad.

Grey said...

Avoiding the specific argument over E and this sample (my guess is IJ) but it's always seemed to me generally likely that there was / were probably OoA crossings at Iberia and Sicily when the sea levels were at their lowest.

The thing about that is if the groups doing it were coastal adapted and at the time of lowest sea level then the evidence would mostly be underwater.

I doubt the numbers would ever have been great but I could see E being present all along the Atlantic coast of Europe in small numbers quite early.

(Although I imagine the bulk of it would be from Atlantic Megalith later.)

Kristiina said...

Matt, never mind! If it were true, it would have hit the headlines and it could not have passed unnoticed.

I agree with Grey, and IMO many haplogroups are possible in Oase 1’s case (C, IJ, E, F(xIJKLT)). To me the available ancient yDNA gives the impression that there has been big fluctuations of yDNA in the same areas over millennia. Moreover, it looks like modern expansive haplotypes are quite young and have replaced earlier haplotypes and we probably tend to reduce the geographic area of a haplogroup, including its extinct branches, to too small an area although old haplogroups such as E or P may have covered wide areas already 25 kya ago. In Haak’s paper we had two very different haplotypes of R1b in Iberia and Russia with completely different autosomal profiles. Moreover, we know that for example hg C covered whole Eurasia at a very early date.

Chris Davies said...

Y DNA haplogroups C and IJ and F and H are found in African pygmies (Mbuti and Biaka). Interesting how this phenomena seems to have been universally ignored.

Vincent said...

"Y DNA haplogroups C and IJ and F and H are found in African pygmies (Mbuti and Biaka). Interesting how this phenomena seems to have been universally ignored."

Could you please provide us with links?

Chris Davies said...

@ Vincent - It is from this paper:-

http://link.springer.com/article/10.1007%2Fs00438-014-0903-8

However, you will need to see the supplementary data. Do you have an email address which I can send this over to you? I have already previously emailed the data to Maju.

Vincent said...

Thanks. No need to email me the data: they are already freely available there. I see the haplogroup labels are only predicted, not SNP-tested, so such odd results as H, C or IJ are most likely mispredictions.

Chris Davies said...

They used Vadim Urasin's YPredictor v1.5.0 . A 70% prediction strength threshold was implemented. That still leaves: H*-M69(xM82) at 92% probability [Biaka pygmies]; R2a-M124 at 87% probability [Biaka]; IJ-M429 just shy of the threshold at 68% probability [Biaka]; J2a4-L26 at 98-99% probability [Mbuti]; the C and the F examples in the Mbuti are in the 62-65% probability range however. Also noteworthy are Q1a-MEH2 in the Tanzania Sandawe [best examples being 97%-100% probability]; DE in Sandawe [99%/100% probability];I2a in Sandawe [75%/90%/95% probability]. Also in Nigeria Yoruba: F-M89 at 77% probability; DE at 87% probability; and IJ-M429 at 78% probability.

andrew said...

Going from 50%+ Neanderthal to 5%-11% Neanderthal in 4-6 generations, means that the population as a whole had multiple separate instances of Neanderthal admixture, as an isolated instance would have gone to 3.1% in four generations, 1.56% in five generations, and 0.78% in six generations. Thus, you can infer the average community admixture level from this one individual with much more precision than simply treating this individual as a statistical sampling of one.

andrew said...

The timing also favors a place of admixture in SW Europe or West Asia, but not Arabia proper, assuming fairly moderate rate of migration from the Middle East towards Europe over 120-180 years or so by UP Hunter-Gatherers.

And, it favors the existence of secondary admixture, not just a single event, since the location is probably to West Eurasian and too late to be the source of East Eurasian admixture.

andrew said...

"you can infer the average community admixture level from this one individual with much more precision than simply treating this individual as a statistical sampling of one"

At five generations and 8% Neanderthal, the community-wide level of admixture would be a bit over 6%, almost triple current levels of West Eurasian admixture, implying significant post-Neanderthal dilution of West Eurasian gene pools with lower Neandertal admixture populations from ca. post-29kya until the present. As recently as copper age Otzi there were probably 8% Neanderthal admixture populations in Europe, with him being half at about 4%. Indo-European influx may have been a major factor in the dilution.

Maju said...

@Andrew: not if you marry your cousin, as probably happened now and then in such small forager populations. There can still be a single instance of Neanderthal admixture affecting various ancestors at the same time. In small communities you often get the same ancestor in various branches several generations ago.

jackson_montgomery_devoni said...

@Chris Davies,

There is a J2a4-L26 sample among the Mbuti people?

J2a4-L26 at 98-99% probability [Mbuti]''

capra internetensis said...

3/9 J1a1 Mbuti *and* 7/18 I2a1+Q1a Sandawe *and* 3/24 H1*+R2a Biaka all in one paper. None in any other study. 0/47 Mbuti with F in Wood 2005, 0/68 Sandawe with F in Tishkoff 2007, 0/20 Biaka in Cruciani 2002 (and there are more).

Either Xue et al just happened to pick unprecedentedly weird samples for 3 different groups, or the haplogroup predictor got some obscure African haplotypes wrong.

Nirjhar007 said...

Kristiina,Capra
I want you guys to have kind thought sharing here or some kind of debate cause its getting really boring;)-
http://new-indology.blogspot.in/2015/05/sumerian-and-indo-european-surprising.html

capra internetensis said...

@Nirjhar

That's an interesting discussion, thanks for the link.

Unfortunately I know pretty much nothing about Sumerian, so I can't really add anything. All I know is that Whittaker's proposal has not been generally accepted, but I can't evaluate it myself.

Nirjhar007 said...

Well just have a discussion there its not rocket science you know:) and the relation is very convincing imo.