January 21, 2013

Ancient DNA from Tianyuan Cave

Another new PNAS paper that hasn't yet appeared in the journal website. Still, from this description at ScienceNews this appears to be Very Important, as it pertains to a 40,000-year-old modern human, which, if I'm not mistaken is the oldest modern human tested so far:
Ancient DNA from cell nuclei and maternally inherited mitochondria indicates that this individual belonged to a population that eventually gave rise to many present-day Asians and Native Americans, says a team led by Qiaomei Fu and Svante Paabo, evolutionary geneticists at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. 
The partial skeleton, unearthed in Tianyuan Cave near Beijing in 2003, carries roughly the same small proportions of Neandertal and Denisovan genes as living Asians do (SN: 8/25/12, p. 22), the scientists report online January 21 in the Proceedings of the National Academy of Sciences.
The Max Planck press release adds some information:
The genetic profile reveals that this early modern human was related to the ancestors of many present-day Asians and Native Americans but had already diverged genetically from the ancestors of present-day Europeans.
This is an important finding because some published demographic models had Europeans and East Eurasians diverging as recently as ~20 thousand years ago. It now appears that they did so already at around the time of the Upper Paleolithic revolution, when unambiguous evidence of modern humans across Eurasia exists.

UPDATE I: While we wait for this paper to appear on the PNAS website, it might be useful to wonder whether the Tianyuan sample might fall on the East Asian/Amerindian group or the more general "Ancestral South Indian" (ASI)/East Eurasian group.

According to current dating, haplogroup M itself is ~50 thousand years old, and most of the subclades therein coalesce to younger than 40ky times. It's possible that the Tianyuan sample dates from a period where ASI/East Asian differentiation had only just begun or was just about to begin.

The press release makes clear that Tianyuan was already "Asian" rather than generalized Eurasian, proving that East/West Eurasian differentiation had begun by ~40kya. It will be interesting to see whether it can be placed on a more specific "East Eurasian" group rather than a generalized "Asian" one.

UPDATE II: The paper is now online.

UPDATE III: From the paper:
Thus, it is related to the mtDNA that was ancestral to present-day haplogroup B (Fig. 1), which has been estimated to be around 50,000 y old (18) (50. 7 ka BP; 95% CI: 38.1–68.3 ka BP). We note that the age of the Tianyuan individual is compatible with this date.
So, it appears to be within macro-haplogroup N, with haplogroup B being, I think, a fairly unambiguously East Asian/Native American clade of the phylogeny. It will certainly be interesting to see how the much more successful -and younger- M subclades ended up dominating East Eurasia.

UPDATE IV: The TreeMix analysis clearly places Tianyuan within the Asian group, but does not resolve whether Papuans are an outgroup to East Asians/Tianyuan:


I guess that is expected (see my UPDATE I), since Tianyuan dates from a period where within-Asia differentiation had only just begun or was about to begin.

UPDATE V: With respect to sharing of alleles with archaic Eurasian hominins, the Tianyuan sample is within the modern range of variation.



PNAS doi: 10.1073/pnas.1221359110

DNA analysis of an early modern human from Tianyuan Cave, China

Qiaomei Fu et al.

Hominins with morphology similar to present-day humans appear in the fossil record across Eurasia between 40,000 and 50,000 y ago. The genetic relationships between these early modern humans and present-day human populations have not been established. We have extracted DNA from a 40,000-y-old anatomically modern human from Tianyuan Cave outside Beijing, China. Using a highly scalable hybridization enrichment strategy, we determined the DNA sequences of the mitochondrial genome, the entire nonrepetitive portion of chromosome 21 (~30 Mbp), and over 3,000 polymorphic sites across the nuclear genome of this individual. The nuclear DNA sequences determined from this early modern human reveal that the Tianyuan individual derived from a population that was ancestral to many present-day Asians and Native Americans but postdated the divergence of Asians from Europeans. They also show that this individual carried proportions of DNA variants derived from archaic humans similar to present-day people in mainland Asia.

Link

35 comments:

genetiker said...

This is consistent with this post where a Caucasoid-Mongoloid split time of 40,000-80,000 years ago was estimated from the slower 1.25x10-8/bp/gen mutation rate.

genetiker said...

Given the results of the autosomal analyses of the Ainu, and the craniometric analyses of the Upper Cave skulls, it's possible that this individual was Ainoid and not Mongoloid in phenotype.

andrew said...

@gentiker

"it's possible that this individual was Ainoid and not Mongoloid in phenotype."

Not if the autosomal and mtDNA results are basically consistent with the current population of China.

Also, one of the important findings of the study which shows up in the Science News headline but doesn't really get discussed very fruitfully in the body text of that story, is that the Denisovan admixture level in Northern China is basically zero. So, we have a complete absence of non-Neanderthal archaic admixture and no elevated levels of Neanderthal admixture as of 40,000 years ago in Northern China.

This means either that Denisovan admixture didn't happen at all, anywhere in the population ancestral to East Asians at levels sufficient to be apparent in its population genetics even 40,000 years ago prior to an immense amount of population growth (and given small founder populations this is very close to zero instances of admixture in a smallish population rapidly expanding into virgin territory), or a first wave that did admix with and exile/dilute out of existence/cause the extinction of the archaic hominins of mainland Asia had pretty much entirely been wiped out by subsequent waves as of 40,000 years ago, which pushes the first wave to points in time earlier than any really solid and consistent evidence of anatomically modern humans and their archaeological cultures in the region at all.

Dan Kalina said...

Haplogroup M split from N 60,000+ thousand years ago. Even Different haplotypes in Europe such as X an H have a greater than 50000 year disparity.

dix13 said...

Tianyuan Cave is about 50 km SW of Beijing. The ancient Chinese Zhongshan kingdom formed by the Di people was at Boading which is about another 50 km SW of Tianyuan Cave.

The Jilin study,

http://dienekes.blogspot.com/2013/01/ancient-y-chromosomes-from-china.html

attributed the high freq yHg Q to the Di people there. Native Americans are mostly yHg Q.

40000 bp is closed to the time when yHg's N,O,Q,R split off from yHg K. All these yHg groups are noted in the Jilin study.

Ainu are mainly from yHg D which is significant in Japan and Tibet. Tianyuan Cave being in the middle of these two places might have yHg D in the past but from 5000 bp there was little sign of them left at this place.

The mtHgs of Ainu are also present in a wide range of East Eurasian and native Americans.

eurologist said...

Article: Humans with morphology similar to present-day humans appear in the fossil record across Eurasia between 40,000 and 50,000 years ago.

Absence of evidence is not evidence of absence.

AMHs could have been present ~125,000 to at least 80,000 ya in Asia.

hamarfox said...

The genetic profile reveals that this early modern human was related to the ancestors of many present-day Asians and Native Americans but had already diverged genetically from the ancestors of present-day Europeans.

This is interesting indeed. It could mean a few different things: that this ancient genome differs from that which contributed to the majority of modern European genetics (i.e. the ~60-75% ancient West Eurasian in modern West Eurasians), or it could mean that this ancient East Eurasian genome isn't ancestral at all in any significant degree to Europeans (barring populations with relatively recent admixture).

I don't like trying to infer too much from abstracts or press releases, so I hope the paper is accessible soon.

Matt said...

it might be useful to wonder whether the Tianyuan sample might fall on the East Asian/Amerindian group or the more general "Ancestral South Indian" (ASI)/East Eurasian group

Yes, that's an interesting possibility.

I'd be interested to know how this sample fits into the existing models we have of East Eurasian / Oceanian diversification.

On the one hand we have the Rasmussen et al model where we have this first Asian southern wave model, which then admixes with a second Asian southern wave which follows the same route and undergoes constant admixture, but where much more of the first wave survives in Australia and mainland Asia is basically entirely second wave (with contributions from West Eurasians in India and Central Asia).

(There is a variation of this model whereby the second wave basically swooped around through central Asia, but I don't think this is currently supported).

On the other hand, the Rasmussen model is basically dependent (and this is stated in the paper) on further admixture between East Eurasian and West Eurasian populations not happening. And the latest Lipson et al paper shows contribution to all West Eurasian populations
http://dienekes.blogspot.co.uk/2012/12/efficient-moment-based-inference-of.html (I don't think the earlier papers showing contributions specifically to Europeans would've been strong enough to undermine the conclusions of Rasmussen et al).

So it would be interesting to see how this sample (and the Ainu sample, if possible) would slot into those models.

hamarfox said...

The information in table 1 is difficult to interpret. Not all East Asian populations are closer to Tianyuan than the two European populations. Even the Han are barely closer to Tianyuan than are Sardinians. I'm not sure I'm interpreting it correctly, though, since it's not a simple proportionate quantification.

Dienekes said...

The information in table 1 is difficult to interpret. Not all East Asian populations are closer to Tianyuan than the two European populations. Even the Han are barely closer to Tianyuan than are Sardinians. I'm not sure I'm interpreting it correctly, though, since it's not a simple proportionate quantification.

Tianyuan is from a period where East Asian differentiation had just begun. A European and an East Asian at the time were very little differentiated (perhaps by a few thousand years). So, Tianyuan is only a little closer to East Asians, because 40ky separate it from them as opposed to 40+x ky from Europeans, where x is a few thousand years.

hamarfox said...

Tianyuan is from a period where East Asian differentiation had just begun. A European and an East Asian at the time were very little differentiated (perhaps by a few thousand years). So, Tianyuan is only a little closer to East Asians, because 40ky separate it from them as opposed to 40+x ky from Europeans, where x is a few thousand years.

Aha! Thanks. I almost put that explanation as a possible counterinterpretation. I imagine Papuans overtook Europeans in genetic differentiation owing to drift and Denisova admixture. Amerindians may be the closest population, despite the expected drift, because Tianyuan is a particularly proximate ancestor of theirs (meaning the other populations may have more input than Amerindians do from a different ancestral source)?

terryt said...

"I'd be interested to know how this sample fits into the existing models we have of East Eurasian / Oceanian diversification".

Now we know: mt-DNA B.

"Thus, it is related to the mtDNA that was ancestral to present-day haplogroup B"

I'm actually surprised B had got so far north so early. I'm reasonably certain that Y-DNA R11'B originated on the South China/Vietnam/Philippines region.

"It will certainly be interesting to see how the much more successful -and younger- M subclades ended up dominating East Eurasia".

I wouldn't be so sure that M is younger than B in the region. However I agree it is probably younger than N in the region.

"The TreeMix analysis clearly places Tianyuan within the Asian group, but does not resolve whether Papuans are an outgroup to East Asians/Tianyuan"

I suspect the main, or only, connection between Papuans and Central Chinese is the presence in Melanesia of B, in the form mainly of B4a1a1a. That haplogroup was almost certainly introduced by Austronesian-speaking people.

genetiker said...

My hypothetical timeline of extra-African human evolution:

73 ka - Toba erupts, killing all humans in Western Asia, and any in South Asia.

70-67 ka - CT and L3 Negritoids leave East Africa heading east, and continue all the way to the Philippines.

70-55 ka - Negritoids in North and Central India evolve into Veddoids. A key innovation is the evolution of straight hair from tufted, woolly Capoid hair. During this period, humans in varying stages of Veddoid evolution leave India. Some head west and then northwest. Others head east and then northeast.

65-47 ka - Veddoids in West Asia and Europe interbreed with a western race of heidelbergensis (Neanderthals).

65-47 ka - Veddoids in East Asia interbreed with an eastern race of heidelbergensis (Dali-like). Sinodonty is introgressed from these archaics in the Veddoids furthest to the north.

55-40 ka - Veddoids in West Asia and Europe evolve into proto-Caucasoid Cro-Magnons.

55-40 ka - Veddoids in East Asia evolve into proto-Mongoloid Ainoids.

40-25 ka - Cro-Magnons evolve into modern Caucasoids.

40-25 ka - Ainoids in North Asia evolve into modern Mongoloids. Stiff hair evolves during this period. The epicanthic fold characteristic of Mongoloids may also have independently evolved during this period.

eurologist said...

73 ka - Toba erupts, killing all humans in Western Asia, and any in South Asia.

I have often stated that Toba likely had a significant impact in SE Asia - but I would not go that far - neither with AMHs or ancient humans.

The large percentage of Denisovan DNA in some SE Asian populations makes it difficult to ascertain their relation to the very first AMHs in the region. To me, the most straightforward explanation is that there were several waves of AMHs between 120,000 and 60,000 ya, and part of the SE Asian difference is due to that and due to the fact that non-AMHs were eradicated and/ or assimilated during that time frame or just after.

The Tianyuan individual is what we expect of a typical person there and then. We need to go back in time further than this to better see what was going on.

terryt said...

"To me, the most straightforward explanation is that there were several waves of AMHs between 120,000 and 60,000 ya, and part of the SE Asian difference is due to that and due to the fact that non-AMHs were eradicated and/ or assimilated during that time frame or just after".

I agree totally. The 'several waves of AMHs' were almost certainly not all from Africa of course.

andrew said...

"73 ka - Toba erupts, killing all humans in Western Asia, and any in South Asia."

Almost certainly not. There are South Asian modern human relics in continuity on both sides of the Toba ash in Southern India. While Toba's impact would have been great in SE Asia, it would not have had nearly so intense an effect in Western Asia (e.g. Persia, Arabia, Turkey), where the effects would have been milder than in South Asia where we know that modern humans survived.

There does seems to be a gap in the Levantine archaeological record with modern humans present from about 100 kya to 75kya then absent auntil about 50 kya. Modern humans are then present in the Levant at all times after that, and this gap could be Toba driven climate change related. But, the Neanderthal admixture dates (not all of which depends upon disputed mutation rate methodologies) points to a time frame of 75 kya for an admixture time, suggesting that modern Eurasians are in continuity with first wave Out of Africa and into SW Asia populations. The times of divergence between Eurasian and African mtDNA and Y-DNA lineages also tends to support continuity. New Arabian finds seem to show continuity in refugia in post-Toba the interior of SW Asia.

Finally, there is no solid archaeological evidence for any pre-Toba modern humans in SE Asia or East Asia at all. There are a couple of alleged older modern human finds to about 100 kya in China, but there is reason to doubt both the dating and the catagorization of the finds that are that old as modern human.

A more plausible scenario given what we know is that Toba was an intense short term blow to vegetation, megafauna and archaic hominins in SE Asia, opening the door for modern humans to migrate to SE Asia and beyond from South Asia, which they seized.

eurologist said...

I'm reasonably certain that Y-DNA R11'B originated on the South China/Vietnam/Philippines region.

Terry,

The B-tree in the paper seems to indicate otherwise. Except for the Nicobarese, the majority of the first populations to branch away are northerly extant populations (and Native Americans). With Tianyuan sitting all the way on top, my guess is that B evolved from R at a more central/ northern latitude, and then some B-people moved south along the coast, hitting Taiwan (the staging ground for Melanesia and Polynesia) and Thailand. A separate, earlier group eventually reached Nicobar, again likely along the coast and by boat.

If you propose a southern origin, you would need something like 5 unrelated migrations northward at very different time frames.

Va_Highlander said...

@dix13: Tianyuan Cave is about 50 km SW of Beijing. The ancient Chinese Zhongshan kingdom formed by the Di people was at Boading which is about another 50 km SW of Tianyuan Cave.

The most obvious difficulties are that "Di" -- 狄 -- was used to denote multiple ethnic groups and the "White Di" -- 白狄 -- the specific ethnic group that founded the Zhongshan state, were from Shaanxi Province, not Hebei.

It is likely misleading to speak of the "Di people" as if they were a homogeneous population, ethnically, culturally, or genetically. The ancient Chinese did not seem to view them this way and neither should we.

Distance Yourself said...

@genetiker

55-40 ka - Veddoids
in West Asia and Europe
evolve into proto-Caucasoid
Cro-Magnons

As Dienekes wrote in this post:

http://dienekes.blogspot.de
/2010/12/world-craniometric-analysis-with-mclust.html

"Grimaldi, described by some
as Negroid is actually
assigned to the Australoid
cluster and so is Markina Gora."

This fits perfectly to your timeline.

genetiker said...

This fits perfectly to your timeline.

Indeed. I remember reading that page, but had forgotten about it. Thank you for connecting the dots and pointing that out.

Ebizur said...

Eurologist wrote,

"The B-tree in the paper seems to indicate otherwise. Except for the Nicobarese, the majority of the first populations to branch away are northerly extant populations (and Native Americans)."

The tree in this paper does not really indicate a "Northern" origin of mtDNA haplogroup B. It is clear that a relative of proto-B4'5 (i.e. a relative of the ancestor of most extant haplogroup B) had reached North China by approximately 40,000 YBP, but it does not bear much significance in regard to the place of origin and initial diversification of the now-common subclades B4 and B5.

Branch 1: Tianyuan (40,000 BCE North China) vs. all others

Branch 2: 4 Nicobarese (indigenous peoples of some small islands in Southeast Asia between the Andaman Islands and Sumatra
who speak several languages that are generally considered to constitute one "Nicobaric" branch of the diverse Austroasiatic language family)
+ 1 Negidaltsy (North Tungusic-speaking East Siberian aborigine) vs. all others

(Perhaps this Nicobarese+Negidal branch represents haplogroup B5? Since the Nicobarese are of ancient residence in their islands deep in Southeast Asia and the Negidals are likewise relatively indigenous inhabitants of the area of the lower Amur in coastal East Siberia, this data set is equivocal regarding the region of origin of this branch.)

Branch 3: 1 Thai + {1 Uzbek + 1 Thai} vs. all others

(This branch contains an Uzbek in addition to two Thais, but the Uzbek is contained within an interior branch, which makes this subset of haplogroup B (probably an early branch of B4) more "Thai" than "Uzbek," or in other words more Southeast Asian than Central Asian, according to this study's data.)

Branch 4: 1 Buriat + {1 Tubalar + 2 Native Americans} vs. all others

(This seems to be a branch of B4, probably B4b'd'e or B4b, that moved northward into southern Siberia and eventually all the way into America, where it produced B2, the subclade of haplogroup B4b that is ubiquitous in America.)

Branch 5: 2 Taiwan Aborigines vs. all others

Branch 6: {Korea + Khirgiz + South Siberian Turkic peoples (+ Negidal?)} vs. {several Taiwan Aborigine branches, including the Oceanian subset}

(Branches 5 and 6 may reflect a later expansion of B4a or at least proto-B4a in East Asia, probably from around Korea, Taiwan, or some place in between (East China? Japan?). It is notable that this branch also has reached the Altai region in addition to its well-known expansion into Oceania.)

eurologist said...

Ebizur,

I am not convinced. To me, the evidence is evolution of B at approximately Beijing - Seoul latitude and location.

The top branches are the northerly Bejing, Negidal, Uzbek, Buriat, Tubalar, native American - with the only exceptions being Nicobarese (but we know from mtDNA B that's not a problem) and Thai (again, mtDNA B shows a similar secondary pattern).

Ed said...

Even if mtDNA B did originate in North Asia in some ancestral state when East and West Eurasians recently separated, it seems that there was an extensive incubation period in (or around) Southern China that connects it with present day East Asians.

Out of curiosity, what is the likelihood that the population ancestral to both East and West Eurasians was mtDNA B?


Ebizur said...

eurologist wrote,

"I am not convinced. To me, the evidence is evolution of B at approximately Beijing - Seoul latitude and location."

I made a comment that the data provided in this study do not logically support a hypothesis that you have presented claiming an origin of mtDNA haplogroup B in "central/ northern (East Asia)." Now, you have responded that you are "not convinced." Frankly, since I have not presented any hypothesis, that does not make much sense.

eurologist wrote,

"The top branches are the northerly Bejing, Negidal, Uzbek, Buriat, Tubalar, native American - with the only exceptions being Nicobarese (but we know from mtDNA B that's not a problem) and Thai (again, mtDNA B shows a similar secondary pattern)."

These "branches" of which you speak are not located at phylogenetically equivalent levels in the tree. The Buriat, Tubalar, and Native American instances all together count as only one branch in this exercise, for example. I have already mentioned that the Uzbek instance is contained within an internal branch of a clade that is otherwise represented by Thais.

Since you seem to lack a basic understanding of cladistics, I do not feel like spending any more of my time discussing this matter with you. Sorry.

eurologist said...

"These "branches" of which you speak are not located at phylogenetically equivalent levels in the tree."

Ebizur,

They don't have to be (and usually aren't) exactly on the same level to demonstrate such arguments - people move around, sometimes a level pretty close to top might be (and actually is) in a vastly different location today - we have only a small number of branches to work with (in this study). Again, what are the top levels?

- Bejing
- Negidal & Nicobar
- group:
. - Uzbek & Thai
. - group:
.. - Buriat, Tubalar & Native American
.. - Taiwan / Korea / Kirgiz /Tofalar/ Tuvan cluster

So, the top level supports Bejing. The next level is mixed northerly and southern on both branches. The second of these is again mixed; however, one is a major northern cluster, and the other again is mixed.

This is why to me, the best explanation is an origin at central latitudes - neither in the extreme north (there are no documented AMHs during that time, anyway) nor in the south (this would require ~5 different, unrelated northward expansions; conversely, from a central location, you only need 3 southern migrations if the two Taiwan ones happened contemporaneously - which makes sense).

Also, I didn't expect blatant insults from you, quite disappointing, actually. The cladistics are straightforward - their interpretation in space and time not.

eurologist said...

Ebizur:

"Now, you have responded that you are "not convinced." Frankly, since I have not presented any hypothesis, that does not make much sense."

How is that difficult to understand? I am not convinced by your arguments. Your logic is flawed.

"Branch 1"

Fact is: Tianyuan sits on top of your branches 2 & 3 - not at the same level.

"This branch contains an Uzbek in addition to two Thais, but the Uzbek is contained within an interior branch, which makes this subset of haplogroup B (probably an early branch of B4) more "Thai" than "Uzbek," :

Logical fallacy: given this is the only occurence of a Thai destination in the entire data set, the most parsimoneous explanation is that there was a single migration towards Thailand when both subgroups existed contemporaneously. So, this in fact strongly supports a central location.

Finally, since this data set (at face value) except for the top level argument is somewhat inconclusive, another approach is to count the number of required migrations - which is what I did.

terryt said...

"There does seems to be a gap in the Levantine archaeological record with modern humans present from about 100 kya to 75kya then absent auntil about 50 kya".

It is my understanding that Neanderthals were present in the Levant during the intermediate period. I'm fairly sure that the intermediate period coincided with a time of climate cooling, which suggests that biology rather than culture remained a major element in human life.

"a time frame of 75 kya for an admixture time, suggesting that modern Eurasians are in continuity with first wave Out of Africa and into SW Asia populations".

Completely agree. It also suggests that humans had moved beyond the Levant by the time Neanderthals replaced them there.

"The B-tree in the paper seems to indicate otherwise. Except for the Nicobarese, the majority of the first populations to branch away are northerly extant populations (and Native Americans)".

To expand on Ebizur's commnet here:

"It is clear that a relative of proto-B4'5 (i.e. a relative of the ancestor of most extant haplogroup B) had reached North China by approximately 40,000 YBP, but it does not bear much significance in regard to the place of origin and initial diversification of the now-common subclades B4 and B5".

Yes. In fact 'B' is just one part of haplogroup R24'B, actually three haplogroups: R24 in the Philippines, R11'B6 both branches being South Chinese, and B4'5 also both branches mainly present in South China (B5 almost exclusively so). Native Americans are a very downstream branch of B4b and perhaps should be called 'B4b2'. B4b1 is especially common in Hainan.

"The top branches are the northerly Bejing, Negidal, Uzbek, Buriat, Tubalar, native American - with the only exceptions being Nicobarese"

I think you are making the false assumption that the diagram in the paper includes all B haplogroups.

"we have only a small number of branches to work with (in this study)".

Exactly. We need to look at the haplogroup as a whole.

"So, the top level supports Bejing. The next level is mixed northerly and southern on both branches".

It would have had more releavce if the authors had noted the actual current nomenclature. As it is the diagram is virtually meaningless.

"nor in the south (this would require ~5 different, unrelated northward expansions"

On the contrary, if we look at the B haplogroup as a whole we see that B4'b'd'e was actually the only branch with representatives outside South China and regions very nearby. One migration north. What surprised me originally was that memebers of B4b'd'e must have been present in Northern China by 40 kya. It's possible that the diagram refers only to haplogroup B4b'd'e. That would then indicate a southward movement from Central China for elements of that haplogroup, probably Neolithic. Certainly it seems that the Nicobar Islands were settled no earlier than that period.

terryt said...

"Out of curiosity, what is the likelihood that the population ancestral to both East and West Eurasians was mtDNA B?"

Extremely unlikely.

eurologist said...

Terry,

I don't know how the authors selected the 311 modern human mtDNAs for comparison - they simply claim that 36 of them belong to "B", and all of those are below Tianyuan's ancestry. And given the populations listed, it's clearly a pretty broad B spectrum.

We surely need more data, but taken at face value, this demonstrates that "B" was represented rather central/northerly quite early on.

Again, what we have to remember is that even if there were few AMHs in SE and E Asia before MIS4, there were still 25,000 years between 45,000 ya and LGM for people to move around in many directions. For some (sub-) haplogroups, that simply means south instead of north, or both (from a central location).

eurologist said...

"if we look at the B haplogroup as a whole we see..."

Terry, what paper or online tree are you referring to? I would like to take a look at it, to better understand the populations and branches in the tree of this paper.

Thanks.

terryt said...

"and all of those are below Tianyuan's ancestry. And given the populations listed, it's clearly a pretty broad B spectrum".

But it doesn't fit the phylogeny given at Phylotree. According to Maju the 'B' found at Tianyaun is B4'5. B5 is a Sumatra/Malay haplogroup, and may be the 'Nicobarese' in the diagram. So Tianyuan B may be ancestral B4. 'Native American' B is B2, a clade within B4b'd'e. The 'Taiwan/Pacific' haplogroup would be B4a1. The other 'Taiwan Aborigines' would be B4a. B4a is certainly not 'downstream of' B4b. I cannot fit the remainder of the diagram here to Phylotree.

"We surely need more data"

We actually have it at Phylotree.

"Terry, what paper or online tree are you referring to? I would like to take a look at it, to better understand the populations and branches in the tree of this paper".

Do you not know it? B is part of R:

http://www.phylotree.org/tree/subtree_R.htm

eurologist said...

http://www.phylotree.org/tree/subtree_R.htm

Terry,

I am talking about a (or several pertinent, recent) paper(s) or websites aligning the tree with extant populations. Of course I am aware of phylotree - but I am obviously not going to read the >150 papers cited at phylotree - many of them outdated.

I am not willing to judge that quickly that the tree published in this paper is simply wrong, nor that their data base is highly biased.

terryt said...

"I am not willing to judge that quickly that the tree published in this paper is simply wrong, nor that their data base is highly biased".

Phylotree is based on far more data than is the diagram in this paper. From what I can see is that the authors have based their tree around American B4b'd'e rather than using a deeper B haplogroup. The paper's main value is that it shows that B4b'd'e was so far north so early. It is basically the only B haplogroup to have made it so far to the north. The main outlier is Nicobarese. Evidently They have mt-DNA B5a1, found also in the Philippines and as far north as Japan.

http://www.andamanese.net/Origin%20of%20Andamanese-%20Thangraj%20et%20al.pdf

"Most Nicobarese mtDNA lineages belong to either of the two common haplogroups B and F (1), which are specific to East Asia. All four Nicobarese B5a sequences clustered
together in a branch defined by three coding region substitutions (at nucleotide positions 11881, 13145, and 13395) (fig. S1) (7). The low variation observed both in control and coding region sequences implies that these lineages among Nicobarese coalesce to their most recent common ancestor within the past couple of
thousand years".

Which fits the long accepted view of the Nicobar Islands. The authors next seem to have considered just one of B4b'd'e's 'brother' haplogroups, B4a. That gives a very simplified view of the phylogeny. The fact the diagram is simplified is demonstrated by the complete absence of any 'Chinese' B.

terryt said...

"I am not willing to judge that quickly that the tree published in this paper is simply wrong, nor that their data base is highly biased".

I've cracked it. The authors have looked at the extremities of B's expansion (America, Polynesia and the Nicobar Islands) and followed the haplogroups back. As a result they've confined their search to B5, B4b'd'e and B4a, and ignored the fact that B5 is found as far east as the Admiralty Islands and as far north as Japan as well as in the Nicobar Islands. But B5's spread probably goes no further back that the Hoabinhian's development in SE Asia, especially Vietnam. And B5's presence in the Nicobar Islands may not even go as far back as that. And B4a's spread to the Philippines is probably no earlier than the Hoabinhian either. B4a1a's spread from the Philippines to both Madagascar and the Central Pacific is an even later step in the same process: improved communication between the islands through improved boating technologies. We are left with just one B haplogroup that was eventually to move north beyong modern China: B4b'd'e.

F.W.NMaCP said...

I have no background in these things, but have been reading this discourse and surrounding information with considerable interest- it doesn't appear to have continued to develop, which is a shame- and only one insult in there too. So, unless I have missed some conclusive contribution, might I encourage a continuation of this discussion? It seemed to be able to go somewhere...