The single origin of these regional phenotype traits has now been tracked to a single allele of gene EDAR (Wikipedia, SNPedia), specifically to its variant EDAR370A, typical of East Asians, using a genetically modified mouse model.
Yana G. Kamberov, Modeling Recent Human Evolution in Mice by Expression of a Selected EDAR Variant. Cell 2013. Pay per view → LINK [10.1016/j.cell.2013.01.016]
Summary
An adaptive variant of the human Ectodysplasin receptor, EDARV370A, is one of the strongest candidates of recent positive selection from genome-wide scans. We have modeled EDAR370A in mice and characterized its phenotype and evolutionary origins in humans. Our computational analysis suggests the allele arose in central China approximately 30,000 years ago. Although EDAR370A has been associated with increased scalp hair thickness and changed tooth morphology in humans, its direct biological significance and potential adaptive role remain unclear. We generated a knockin mouse model and find that, as in humans, hair thickness is increased in EDAR370A mice. We identify new biological targets affected by the mutation, including mammary and eccrine glands. Building on these results, we find that EDAR370A is associated with an increased number of active eccrine glands in the Han Chinese. This interdisciplinary approach yields unique insight into the generation of adaptive variation among modern humans.
As the summary mentions, it has also been suggested that the variant influences tooth morphology, although this could not be demonstrated in the mouse model.
From Science Daily:
The mutation was found in a gene for ectodysplasin receptor, or EDAR, part of a signaling pathway known to play a key role in the development of hair, sweat glands and other skin features. While human populations in Africa and Europe had one, ancestral, version of the gene, most East Asians had a derived variant, EDARV370A, which studies had linked to thicker scalp hair and an altered tooth shape in humans.
The ectodysplasin pathway is highly conserved across vertebrates -- the same genes do the same thing in humans and mice and zebrafish. For that reason, and because its effects on skin, hair and scales can be observed directly, it is widely studied.
This evolutionary conservation led Yana Kamberov, one of two first authors on the paper, to reason that EDARV370A would exert similar biological effects in an animal model as in humans. The HMS research fellow in genetics developed a mouse model with the exact mutation of EDARV370A -- a difference of one DNA letter from the original, or wild-type, population. That mouse manifested thicker hair, more densely branched mammary glands and an increased number of eccrine, or sweat, glands.
The authors argue for a selective pressure but I find hard to imagine which one could it be. Therefore I rather lean again towards a random founder effect.
Unless... notice that in the figure accompanying the summary also mammary gland density is shown as increased, what could result in improved nutrition for babies, what would indeed be selected for (assuming the rest of effects are neutral or quasi-neutral).
Anyhow, linking it obliquely to the previous entry (on Neanderthal admixture in East Asians), I have often wondered about the origins of straight hair in general, because it is obviously not ancestral in our species (all "Aboriginal African" populations have thinly curled hair, as do some non-Africans, especially remote Tropical populations like Papuans or Andamanese) and my main hypothesis is that it could be a Neanderthal genetic influence, possibly selected because of less intense need of head ventilation (for which thinly curled "African" hair seems best) and more intense need instead of protection against cold and rain. The thick hair East Asian variant might in this case be just the extreme variant of a more widespread Neanderthal introgression. Just food for thought.
Update (Feb 20): Razib Khan also finds the suggested "selection sweep" dubious, and his discussion of this paper is very interesting to read.
An interesting detail that I could not gather clearly from the abstract is that the allele causes (at least in mice) smaller breasts and not just denser mammal glands.
But surely the most interesting item is this map of the frequencies of the EDARV370A allele through the World, being very much and quite strictly limited to the peoples displaying what is classically described as the Mongoloid phenotype:
An intriguing issue is also the lack of it in Europe, even among peoples of well-known Siberian/East Asian low-level admixture (Northern Russians). I'm tentatively imagining some sort of "racist selection" against the trait among those peoples, as it expresses an exotic, possibly not favored (to put it mildly), very apparent phenotype. Food for thought.
Update (March 22 2014): Kevin Brook mentions this database, which shows the "East Asian" C allele present in some other populations than the ones reflected in the map above, notably some South Asians (rather high frequencies in some cases) but also (at very low frequencies) in some Africans, West Asians and Eastern Europeans. At the very least it is an interesting counter-point.
Update (Feb 20): Razib Khan also finds the suggested "selection sweep" dubious, and his discussion of this paper is very interesting to read.
An interesting detail that I could not gather clearly from the abstract is that the allele causes (at least in mice) smaller breasts and not just denser mammal glands.
But surely the most interesting item is this map of the frequencies of the EDARV370A allele through the World, being very much and quite strictly limited to the peoples displaying what is classically described as the Mongoloid phenotype:
An intriguing issue is also the lack of it in Europe, even among peoples of well-known Siberian/East Asian low-level admixture (Northern Russians). I'm tentatively imagining some sort of "racist selection" against the trait among those peoples, as it expresses an exotic, possibly not favored (to put it mildly), very apparent phenotype. Food for thought.
Update (March 22 2014): Kevin Brook mentions this database, which shows the "East Asian" C allele present in some other populations than the ones reflected in the map above, notably some South Asians (rather high frequencies in some cases) but also (at very low frequencies) in some Africans, West Asians and Eastern Europeans. At the very least it is an interesting counter-point.
Interesting idea on the neandertals.
ReplyDeleteAlso, did the native Americans attain this variant common in East Asia, or had the two already separated by 30 thousand ya?
Glad you liked it. It can of course be a very successful ex-novo mutation in H. sapiens but it looks a bit suspicious, right?
Delete"Also, did the native Americans attain this variant common in East Asia, or had the two already separated by 30 thousand ya?"
Not sure. I could not find enough data on this allele's distribution but, if Native Americans have it, they surely must have got it also already in NE Asia 30 or even more Ka ago.
princenuadha wrote,
Delete"Also, did the native Americans attain this variant common in East Asia, or had the two already separated by 30 thousand ya?"
Yes, the same EDAR variant in question is very common among Native Americans and populations that contain significant amounts of Native American admixture, such as Mexicans:
Frequency of ancestral rs3827760 allele (encoding EDAR370V)
1.0 Yoruba in Ibadan, Nigeria
1.0 Utah residents with Northern and Western European Ancestry
0.98 Gujarati Indians in Houston, Texas
0.59 Mexican ancestry in Los Angeles, California
0.20 Japanese in Tokyo
0.08 Chinese in Denver, Colorado
0.06 Han Chinese in Beijing
Frequency of derived rs3827760 allele (encoding EDAR370A)
0.94 Han Chinese in Beijing
0.92 Chinese in Denver, Colorado
0.80 Japanese in Tokyo
0.41 Mexican ancestry in Los Angeles, California
0.02 Gujarati Indians in Houston, Texas
0.0 Utah residents with Northern and Western European Ancestry
0.0 Yoruba in Ibadan, Nigeria
This sample of Mexican Americans from Los Angeles is known to be an approximately 50%/50% admixture of Europeans/Africans and Native Americans in terms of autosomal genetic ancestry, so I would guess that the frequency of the derived EDAR variant in the Native American ancestors of these Mexicans in Los Angeles was probably about the same as the frequency of the derived EDAR variant in Japanese from Tokyo.
That's very interesting, thanks.
DeleteEbizur: I was thinking if there's any data on this variant among SE Asians. I ask because one of the traits supposedly regulated is dental morphology (sinodonty I presume) and SE Asians are mostly sundandont. Also it's plausible under various population history models that they might have it in lesser frequency than NE Asians.
DeleteI believe that the paper has data on Southeast Asians. The pie charts on a global map seem to show this. It is a pretty stark divide between SE Asia and S Asia. The Munda in S Asia also have the derived allelle at an elevated frequency relative to other South Asians.
DeleteAlso, note that the derived EDAR variant is not found with significant frequency in Gujarati Americans. The very low-level presence of the derived allele in this population easily could be due to historically recorded invasions of the Subcontinent or even gene flow from populations of the northern fringes of the Subcontinent (Tharus, Lahaulis, Burushos, etc.), which are quite admixed with East Asians.
ReplyDeleteGiven as this variant became regionally dominant in a time when people were quite capable of making headgear that would overcome any advantages of different hair types, this is obviously a trait that spread through sexual selection. Any teenaged european girl in a european dominant social environment can tell you that any hair texture with more than a very relaxed wave lowers social status.
ReplyDeleteThat's not real in my experience. But I live in a country where "more than a very relaxed wave" is very common, even dominant maybe. It seems to reflect some kind of racist(?) prejudice in your particular modern environment and nothing else.
DeletePS: The children born to Vietnamese mothers fathered by (admixed African) "black" US soldiers suffered a dire social fate in Vietnam. Their pariah status was determined by hair texture rather than skin tone, which in many cases was not outside the range of SE Asian norm.
ReplyDeleteIn some East Asian populations curly hair is relatively common and does not seem to affect social status at all. I have spotted it among Tibetans and among island Koreans. Obviously social status in small communities where everyone knows everyone else is not derived from such minor elements, at least mostly not, but from each one's personality and maybe (where castes are already developed, not in the Paleolithic in any case) birthright (inheritance).
DeleteIf straight hair is a Neandertal trait that introgressed into human populations, then why are there populations with curly/kinky/woolly hair that do have Neandertal DNA (e.g., Papuans)?
ReplyDeleteNeanderthal admixture is in the best case very small in non-Africans, so unless selected for some reason (or lots of luck) it should have not become dominant. That's why I use the term "introgression" which implies a selection for that trait after incorporation by minor admixture.
DeleteWhatever the case, many Eurasians (including myself) have curly hair, just not the "African" thinly curled type (which is much rarer, although not so rare in beards, never mind pubic hair). This may well be an intermediate type. It's too common, not just in Europe and West Asia but also in India, among Australian Aborigenes, etc. to be considered a secondary introgression from Africa in "recent" times, so straight hair was not fully selected for (or randomly succeeded) in many areas but only somewhat.
We may not notice in modern contexts because men often have hair cut very short (more comfortable) and women often alter their hair texture artificially (for whatever aesthetic reasons) but the intermediate phenotype is still very common and even dominant in Southern Europe, West Asia, some parts of India and Australian Aborigines (also often found in SE Asian "Negritos").
Of course, as already mentioned, some non-African populations (Melanesians, Andamanese) still retain the ancestral "African" thinly curled type. Why? Random unknown reasons and/or tropical adaption.
"but the intermediate phenotype is still very common and even dominant in Southern Europe"
ReplyDeleteQuite common in Wales, Ireland, SW England also - dunno about Brittany.
It seems likely to me that African-style hair probably serves a particular tropical purpose.
Populations that moved out of the tropics might simply not have the same selective pressure so random variation breaks it down.
If there was more to it than that then wouldn't the percentage of curly hair have to decline as you moved further from the tropics?
It could of course be a founder effect. If there's no selective pressure nor randomly selecting bottlenecks involved, why don't most retain the thinly curled "Tropical" hair? Once a variant is fixated, as it seems it happened with thinly curled hair (judging on Africans and other tropical peoples), if no pressure of any sort is applied, it should remain there. Also, as the variant is already fixed, the novel allele must come from somewhere (either ex novo mutation or admixture).
DeleteSo I conclude that, unless I'm missing something, (1) there straight hair allele is either a new variant or one taken from Neanderthals and (2) there was some sort of notable "pressure", either by random bottleneck (founder effect) or by some sort of selection in favor of straight hair. I speculate that it may have to do with a trade-off between ventilation (curly) and warming (straight) but I may be totally wrong.
"I speculate that it may have to do with a trade-off between ventilation (curly) and warming (straight) but I may be totally wrong."
DeleteThat seems the most plausible.
If the west coast of the UK and Ireland is an exception to the rule (and it may just be a perception on my part) it could simply be a coastal effect outweighing a northern latitude effect.
Ireland was not (or almost not) populated until "a few" thousand years ago (10,000 years?) and the effect does not seem too extreme anyhow. First is the availability of the allele (i.e. does it exist in the population?, how common is it to begin with?) and then the weight of the effect, which in this case I suspect quite weak.
DeleteMaybe it's about founder effects (and sexual selection?) Who knows.
EDAR was not a random founder effect. It is the strongest signal of a selective sweep in the entire East Asian genome and would be in the top ten for the entire human genome in the last 100,000 years.
ReplyDeleteI agree that explaining it as a result of temperature and humidity adaptation or arbitrary preferences leading to sexual selection are implausible. Neither is a strong enough selective pressure to produce the nearly complete fixation of this gene in all of East Asia and the Americas.
My personal suspicion is that EDAR conferred protection against bubonic plague outbreaks in China to some extent (I outline my reasoning a post at my blog today and comments at Gene Expression), but while that specific hypothesis is speculative, the notion that this was a powerful fitness based selection is not.
I am skeptic about "selective swaps" when the benefit is not obvious and the allele is concentrated in one specific population. This can also be explained, maybe better, by reason of founder effect, at least partly.
DeleteThis article discusses this one and, at the bottom, another related paper. The second one did find an allele that looks very much as being selected for health reasons (TLRS5) but this one is a gene known to be related to immunology. But EDAR is not apparently immunologic and the best reason for a selective swap (if real) that I can imagine is an increase in breastfeeding capacity (however I never read that East Asian women excel in that so...)
The selective sweep classification follows from the statistical profile. The ancestral allelle is almost completely absent, except as a low frequency recessive carrier gene, in the Han Chinese. It is near 100% from SE Asia to East Asia to the Americas prior to known historic era admixtures (and hence must have reached fixation ca. 15,000 kya). It is absent from populations with Denisovan admixture (except to the extent of their Austronesian admixture). It is almost completely absent (except in Munda populations that arrived in the last few thousands years) in South Asia.
DeleteA mutation that didn't exist 45,000 years ago has reached fixation ca. 15,000 years ago. Unless the effective size of the founding population is in the single digits, random founder effects involving selective neutral traits don't bring about that kind of total replacement in that time frame without a selective driver, particularly in the Upper Paleolithic where there is no large scale political organization. If the effective size of the entire Asian founding population is even in the high hundreds or low thousands the numbers just don't work for a random founder effect on an autosomal trait in a growing population that is settling virgin territory for the first few thousand years at least.
The possibility of an immunological effect is quite real. EDAR governs, among other things, sweat gland density. Sweat glands each hold a lymph node. Lymph nodes are your body's warehouses of white blood cells and microphages. The possibility that more warehouses mean mean immune capacity is quite plausible.
But, whatever the selective factor involved is, and we don't know for sure, to get its present distribution, the derived allelle of EDAR needed to be providing a 10% fitness advantage per generation for people who have it relative to people who have the ancestral version of this one gene, which is huge!
But what is the benefit of the allele and why it has not spread to similar levels by means of introgression to other populations in contact with them? Why is it restricted to a geographically well defined population (NE Asians plus the Native American offshoot) and not widespread along a climate range or something like that across population barriers? Europeans for example have been in contact with this population since long ago (at least the Epipaleolithic judging from recent aDNA data) and live in a similar climatic zone, yet the allele has not spread in any significant manner. Elsewhere you mention that its presence is weak in SE Asia, another contact population (and surely ancestral to NE Asians): if the allele would be so desirable for non-climatic reasons, it would also have spread to them and it has not (not significantly).
DeleteI understand that most of the ancestry of NE Asians and Native Americans comes from the same founder population, which migrated northwards from SE Asia, if there was a selective swap it seems more likely to me that it was mild (for the reasons mentioned above) and it affected the founder population rather than larger more recent ones, so it surely originated before the end of the LGM.
As a side not, if you or someone has the paper, I'd appreciate if I could be sent a copy, if nothing else to peek on the graphs and figures in more detail (lialdalmiz[at]gmail[dot]com). Thanks in advance.
"I am skeptic about 'selective swaps' when the benefit is not obvious and the allele is concentrated in one specific population".
ReplyDeleteIt is obvious that what you really mean is that you refuse to accept selective sweeps if accepting them would conflict with what you wish to believe. Just because you (or anyone else) cannot see the reason for the sweep doesn't mean it didn't happen.
"It is a pretty stark divide between SE Asia and S Asia".
That divide is of long standing, yet Maju can still insist that the route into SE Asia via India was a doddle for early humans of all species.
"The Munda in S Asia also have the derived allelle at an elevated frequency relative to other South Asians".
Which makes complete sense. The language and the bulk of their haplogroups come from the east. In fact the boundary between South and Southeast asia has been crossed far more often from east to west than in the other direction.
It's obvious that what I'm saying is just that: that the benefit of the "selection" is not obvious and the allele is concentrated in one specific populations, which would have the same "adaptive needs" as many other populations in which there is no such statistical signal at all. There is at least another very reasonable explanation for that and its name is founder effect.
DeleteYou dislike the concept of founder effects and there's where real problem lays, IMO.
Update published based on the article and discussion at Razib's blog (includes map of frequencies).
ReplyDelete"Update published based on the article and discussion at Razib's blog (includes map of frequencies)".
ReplyDeleteThanks for drawing attention to that post. Extremely revealing, especially the maps.
"An intriguing issue is also the lack of it in Europe, even among peoples of well-known Siberian/East Asian low-level admixture (Northern Russians)".
But the map shows just one North European group so we cannot draw valid conclusions concerning that region.
"I'm tentatively imagining some sort of 'racist selection' against the trait among those peoples"
I don't think so. Look at the other map Razib posted. The gene is centred on where the Hwang Ho(Yellow River) emerges from the hills. Exactly the region where I have suggested that Y-DNA O spread from, and Y-DNA C. And presumably N.
"It's obvious that what I'm saying is just that: that the benefit of the 'selection' is not obvious and the allele is concentrated in one specific populations, which would have the same 'adaptive needs' as many other populations in which there is no such statistical signal at all. There is at least another very reasonable explanation for that and its name is founder effect".
That 'benefit of the selection' (whatever it was) occurred while the population was confined to its source region though, not during its expansion. The allele's expansion was not through 'adaptive needs' or introgression into neighbouring people, but through the huge expansion of the population containing the allele. That is something I have been trying to convince you of for years.
"Elsewhere you mention that its presence is weak in SE Asia, another contact population (and surely ancestral to NE Asians)"
In SE Asia the allele has been diluted by the presence of a pre-Mongoloid population. And it is extremely doubtful that the SE Asian contribution to NE Asia was very great. Just your belief that it is so.
"I understand that most of the ancestry of NE Asians and Native Americans comes from the same founder population, which migrated northwards from SE Asia"
But the allele expanded from around the middle reaches of the Yellow River, not from SE Asia.
"But what is the benefit of the allele and why it has not spread to similar levels by means of introgression to other populations in contact with them? Why is it restricted to a geographically well defined population (NE Asians plus the Native American offshoot) and not widespread along a climate range or something like that across population barriers?"
You should now know the answer to those questions. Of course I don't expect for a moment that you will agree. The obvious explanation doesn't fit your belief.
"The thick hair East Asian variant might in this case be just the extreme variant of a more widespread Neanderthal introgression. Just food for thought".
Unlikely because of its geographic distribution. However it could be a variant introduced to modern humans from an archaic population from somewhere within the Yellow River catchment.
"But the map shows just one North European group so we cannot draw valid conclusions concerning that region".
DeleteIt's obviously North Russians, very likely the very used sample of Vologda, who are known to have levels of Siberian/East Asian ancestry comparable to those of Finns and other Uralic peoples of Europe. True that it'd be best if Finnic peoples and also West Siberian peoples were detailed, but nope.
"The gene is centred on where the Hwang Ho(Yellow River) emerges from the hills. Exactly the region where I have suggested that Y-DNA O spread from, and Y-DNA C. And presumably N".
That's not diversity from the allele but surely: an allele is generally defined by a single SNP, which can have no diversity value. It is surely some other sort of diversity (maybe based on the adjacent genetic segment?) and the label actually reads "Probablity Diversity".
Also I thought you imagine all those lineages expanding from Mongolia or whatever.
I withheld my opinion on this until I can read the paper and see what is it about exactly and in detail.
"That 'benefit of the selection' (whatever it was) occurred while the population was confined to its source region though, not during its expansion".
Strange to say the least that the benefit of selection stopped being important as soon as the population began expanding. I would call that a founder effect. Assuming your explanation as correct, if the initial fixation was caused by selection or random drift in a small population is extremely hard to discern and mostly irrelevant. If there was a selection sweep why would it lose its effectiveness or even, if ecologically determined, be selected against in other environments? That you say is not what they mean when they talk of selection sweep, clearly not.
"In SE Asia the allele has been diluted by the presence of a pre-Mongoloid population".
Then why is it as common as in China, Siberia, Japan (excepting in Cambodia it seems). In South China, where your alleged pre-Mongoloids would also have been dominant once they have very high levels, even fixation, so do in Burma.
The frequency maps are very compatible with a South to North expansion model in which founder effects (and maybe mild selective pressure of some sort?) would have increased the frequency.
"You should now know the answer to those questions".
I do not. Otherwise why would I ask them? I repeat:
But what is the benefit of the allele and why it has not spread to similar levels by means of introgression to other populations in contact with them? Why is it restricted to a geographically well defined population (NE Asians plus the Native American offshoot) and not widespread along a climate range or something like that across population barriers?
"Unlikely because of its geographic distribution".
Hmmm... Let's take another look at the map: the dominance of EDAR alleles follows continental/"racial" patterns. The South-West Eurasian region is dominated by the dark blue allele, while Africa is dominated by one or two light blue ones (not sure because I don't have access to the original map), while Melanesia is dominated by a middle blue type. However only the red allele is novel out of Africa, the others existing at lower frequencies in Africa.
This pattern seems to me a product of founder effects at the beginnings of each population's expansion. The East Asian variant may be novel mutation or Neanderthal introgression but actually all the others look original from Africa, i.e. ancestral Homo sapiens.
Some comments Razib made that I certainly find supportive of what I wrote above:
ReplyDelete"One thing to note is that the derived East Asian form of EDAR is found in Amerindian populations which certainly diverged from East Asians > 10,000 years before the present (more likely 15-20,000 years before the present). The two populations in West Eurasia where you find the derived East Asian EDAR variant are Hazaras and Uyghurs, both likely the products of recent admixture between East and West Eurasian populations. In Melanesia the EDAR frequency is correlated with Austronesian admixture. Not on the map, but also known, is that the Munda (Austro-Asiatic) tribal populations of South Asia also have low, but non-trivial, frequencies of East Asian EDAR. In this they are exceptional among South Asian groups without recent East Asian admixture. This lends credence to the idea that the Munda are descendants in part of Austro-Asiatic peoples intrusive from Southeast Asia, where most Austro-Asiatic languages are present".
"I am perplexed why such an ancient allele, which is found in Amerindians, or Munda populations, is absent in Europeans who have putative East Eurasian admixture".
The prortion of East Asian genetic mixture in Europe may be far lower than that in the other regions listed.
"In addition to EDAR I’m thinking of the pigmentation loci, such as SLC24A5. I wouldn’t be surprised if there was something similar for the epicanthic fold".
And possibly those genes too are from an archaic East Asian population.
"If it is visible, and defines between populations differences, it is generally not genomically trivial. There’s usually a story underneath that difference".
Not through founder effect or drift.
"A climate adaptation is always a good bet. The problem I have with this hypothesis is that modern day gradients in the distribution of this allele are exactly the reverse of what one might expect in terms of adaptation to heat and humidity".
Which supports the expansion of a population not introgression of the gene. Razib then launches into the very unlikely explanation of sexual selection.
"The prortion of East Asian genetic mixture in Europe may be far lower than that in the other regions listed".
DeleteIt is important enough (many different ADMIXTURE/STRUCTURE studies show it) for the nearly fixated "red allele" to show up at some 10% levels or at least 5% among Northern Russians. But not even a 1% line!
The rest I agree that may not have any relevant East Asian admixture, contrary to what some recent MixMapper-based studies claimed.
"Which supports the expansion of a population"...
It does not. If there was a climatic adaption, then the allele would have been selected against in other ecosystems and also introgressed into populations living in similar ones to the favorable case.
We see nothing of that.
What we see fits best with a quasi-neutral allele favored only by the population dynamics of that continental region.
Maju, the frequency of this East Asian allele is not zero in Northern Russians as you alleged multiple times in your comments on this page. I do see a tiny red sliver on their pie chart in the study you cited - look carefully slightly to the right of the 12 o'clock position for red shading. This is confirmed by small frequencies of this allele in tests of Russians in "Positive Selection in East Asians for an EDAR Allele that Enhances NF-kB Activation" by Jaroslaw Bryk, et al., and "The ALlele FREquency Database's Graphical display of Allele Frequencies for Val370Ala - Locus Ectodysplasin A receptor" by Kenneth K. Kidd.
DeleteKidd also found this allele in a couple of other European populations like Hungarians, Greeks, Finns, and Ashkenazi Jews.
Sorry but nope: I downloaded the image and zoomed in all I could and it's all blue, no red, no purple, nothing.
Delete"This is confirmed by small frequencies of this allele in tests of Russians" [in other studies].
Cool that it was found. However are the frequencies (<1% I presume) the ones we should expect? That it is not the case in North Russians, what is as good as saying Finno-Permians. Maybe that arrival of EDAR was from a secondary source such as Turkics and/or specific Uralic subpopulations but was pruned in the main Uralic migration arrow somehow.
Data from Kidd's database showing the frequency of the East Asian allele 1540C in ethnic Russians:
Delete2% frequency in Russians from unspecified regions, sampled by the Human Genome Diversity Project and the Foundation Jean Dausset. Sample code SA001510H.
2.2% frequency in Northern Russians from the Vologda Administrative Region who were sampled by E. Grigorenko. Sample code SA000019K.
3.1% frequency in Northern Russians from "the southwest border of the Arkhangel'sk region". Sample code SA001530J.
The East Asian allele wasn't found in any of Kidd's samples of Italians and Basques.
The entire table can be viewed at http://alfred.med.yale.edu/alfred/SiteTable1A_working.asp?siteuid=SI663326A
I'm noticing a subtle difference between the map and your data, which is that the map seems to represent haplotypes rather than a single-locus SNP as the Kidd's data does.
DeleteHowever the "red" haplotype appears to have a single locus specific marker for that site (marked in red in the smaller graph) so in the end it should be about the same. Yet Kidd's data does not only include C-allele >0% for Russians and others but also for Africans and South Asians, something that is not at all apparent in the Kamberov data (India is blank but Africa is not and the Yoruba are definitely sampled and yet no trace of that 1.3% reported by Kidd).
Personally I'm not too sure on how to deal with the contradictions implicit in both datasets. However I do think that Kidd's dataset is probably a useful counter-reference, so I'll mention it in an update.
"What we see fits best with a quasi-neutral allele favored only by the population dynamics of that continental region".
ReplyDeleteCertainly so. The allele was carried from its region of origin in the form of a population expansion mildly, if at all, advantaged by the possession of the genetic variation. But the variation itself had almost certainly originally come to dominance in its region of origin as an advantageous gentic adaptaion, whether it arrived through founder effect or as introgression from an archaic Homo species.
"Strange to say the least that the benefit of selection stopped being important as soon as the population began expanding".
Surely it is obvious, even to you, that the climatic adaptation involved had almost nothing at all to do with the expansion of the gene. The gene was not 'responsible' for the expansion. The gene was carried into regions that were obviously different from the conditions for which it evolved.
"It does not. If there was a climatic adaption, then the allele would have been selected against in other ecosystems and also introgressed into populations living in similar ones to the favorable case".
The fact it wasn't selected against in dissimilar habitats, and wasn't selected for in similar habitats, surely further supports the fact that the expansion was not assisted in any meaningful way by the mutation.
"If there was a selection sweep why would it lose its effectiveness or even, if ecologically determined, be selected against in other environments?"
Again you're letting your beliefs blind you to the obvious.
"I repeat: But what is the benefit of the allele and why it has not spread to similar levels by means of introgression to other populations in contact with them? Why is it restricted to a geographically well defined population (NE Asians plus the Native American offshoot) and not widespread along a climate range or something like that across population barriers?"
I repeat: because its greater expansion had very little to do with the mutation itself.
"Then why is it as common as in China, Siberia, Japan (excepting in Cambodia it seems)".
What has Cambodia to do with Japan? I'm prepared to accept its expansion through those northern regions was aided by the genetic adaptation to the climate.
"That's not diversity from the allele but surely"
No, it represents the spread of the allele from its point of origin.
"the label actually reads 'Probablity Diversity'".
Yes, of the heterozygosity. The dark red is 100% homozygous for the gene in question. The lighter the red the more the other mutation in the population as it moves towards 0% homozygous for the alternative gene.
"Also I thought you imagine all those lineages expanding from Mongolia or whatever".
Ahh. Maju. Check your atlas. 'Where the Hwang Ho(Yellow River) emerges from the hills' is within spitting distance of the border with Inner Mongolia.
"In South China, where your alleged pre-Mongoloids would also have been dominant once they have very high levels, even fixation, so do in Burma".
The Y-DNA O haplogroup is dominant through the whole region. We would surely expect a high level of the gene, although considerably lower than the level further north.
"The frequency maps are very compatible with a South to North expansion model in which founder effects (and maybe mild selective pressure of some sort?) would have increased the frequency".
You really have to use your imagination considerably though. And why postualte 'founder effects' just to make the evidence fit what you want it to?
"I repeat: because its greater expansion had very little to do with the mutation itself".
ReplyDeleteFor that very reason we cannot talk of "selective sweep". The correct term is "founder effect". We agree in the process but you confuse the correct term to use.
http://en.wikipedia.org/wiki/Selective_sweep
Whether prior to that founder effect, there was a very localized selective sweep or the quasi-fixation was caused by random drift is not important (nor can probably be discerned easily).
"Yes, of the heterozygosity. The dark red is 100% homozygous for the gene in question. The lighter the red the more the other mutation in the population as it moves towards 0% homozygous for the alternative gene".
It cannot be that way. Just compare your claims with the allele frequency map.
"The Y-DNA O haplogroup is dominant through the whole region. We would surely expect a high level of the gene, although considerably lower than the level further north".
Look at diversity, even better: to phylogenetically hierarchical structured diversity (or basal diversity to simplify). Very often regions of great frequency are not the most basally diverse, indicating a founder effect after migration from the most basally diverse region. For example R1b in West Europe is derived from R1b in West Asia, where it is less common but more basally diverse. Change R1b with O, West Europe with NE Asia and West Asia with SE Asia and you have almost the same scenario translocated to the Far East.
"For that very reason we cannot talk of 'selective sweep'. The correct term is 'founder effect'. We agree in the process but you confuse the correct term to use".
ReplyDeleteWe have a 'selective sweep' during the establishment of the gene in the founder population. We then have a population spread that carried the gene into regions beyond that founder population. Perhaps we could refer to that later expansion as spreading the gene through 'founder effect' but there were a large number of them. They outnumbered (or at least came to) any original population in the regions into which they expanded except, of course America and presumably northern Siberia. Can we call such an expansion a 'founder effect'?
"Whether prior to that founder effect, there was a very localized selective sweep or the quasi-fixation was caused by random drift is not important (nor can probably be discerned easily)".
'Quasi-fixation'? One hundred percent fixation in the central population. Random drift is unlikely to explain that level. Selective sweep in relation to climate adaptation is a far more reasonable assumption.
"It cannot be that way. Just compare your claims with the allele frequency map".
The two maps deal with separate issues. The second is an attempt to calculate where the gene first entered the modern human population. Surely you're not claiming that the gene reached its level in the American population through a selective sweep. It presumably made up a reasonable proportion of genes in the first people into America. The element of the population with it have their ultimate origin within the region shown in Razib's second map.
"Look at diversity, even better: to phylogenetically hierarchical structured diversity (or basal diversity to simplify)".
I have. And there is no way O is an original SE Asian haplogroup.
"For example R1b in West Europe is derived from R1b in West Asia, where it is less common but more basally diverse".
Exactly.
"We have a 'selective sweep' during the establishment of the gene in the founder population".
DeleteNo, we do not. You hypothesize that such is the case but you fail to demonstrate it in any way. And, in any case, that is not what the authors mean.
"One hundred percent fixation in the central population".
I have no reason to believe that such population is the origin. You arbitrarily just decided so but that's probably not the case.
" The second is an attempt to calculate where the gene first entered the modern human population".
Have you read the paper? Neither you nor I know what that map means, you are just making things up as you go.
"Surely you're not claiming that the gene reached its level in the American population through a selective sweep".
Not me: the authors of this paper are saying that.
"Change R1b with O, West Europe with NE Asia and West Asia with SE Asia and you have almost the same scenario translocated to the Far East".
ReplyDeletePossibly. So let's examine the distribution of O using your belief in basal diversity mutation tails but ignoring molecular-clockology.
In some unknown region, or while wandering, NO accumulated a string of at least 6 mutations before O-M175 wents its own way. O then developed a string of at least 4 mutations before members of it formed O1-MSY2.2, O2-P31 and O3-M122. Presumably the three haplogroups each developed in separate regions within O-M175's geographic range. O1 seems to have quickly formed O1a-M119, and then O1a1-P203 and O1a2-M110 formed within O1a-M119's geographic region. We can deduce a fair bit about these two haplogroups' expansion over the last few thousand years. The two were almost certainly involved in the Tai-Kradai/Austronesian language expansion. That expansion is not Paleolithic. A recent paper on Chinese Neolithic haplogroups from the three Gorges region placed O1 at that time at the mouth of the Yangtze.
Unlike O1 O2-P31 accumulated at least 3 mutations before O2a-PK4 and O2b -M176 formed. Again O2a and O2b presumably developed within the margin of O2-P31' geographic range. O2a look likely to have been involved in the expansion of the Hmong-Mien and the Austro-Asiatic language expansion, but in several regions the local inhabitants later adopted either Tai-Kradai or Austronesian languages. O2a1a carried Munda into India and even reached the Pashtun. Presumably it is responsible for introducing the EDARV370A allele into India. O2b is the northern version of O2. And let's not forget that its inclusion shifts O2's centroid fairly well north. In fact a recent paper made a convincing case for O2a to have spread south from the Three Gorges region of the Yangtze. That region is where the Yangtze emerges from the hills. A similar environment to where the EDARV370A allele emerged a little to the north.
So O3-M122. It looks to have remained in the same place for a considerable time, perhaps where it had first coalesced from O-M175. Perhaps even where O-M175 itself had first coalesced. After at least 7 mutations O3a1-L127, O3a2-P201, O3a3-M300 and O3a4-M333 formed. The above mentioned paper presented a case for O3 to have developed upstream from the Three Gorges. That makes sense. Members of the haplogroup seem to have been involved with the (perhaps periodic) expansion of the Sino-Tibetan languages and the Chinese Neolithic.
The last two of the four O3 haplogroups have just a single mutation listed. I have no idea where O3a4-M333 has been found. Although O3a3 is associated with Sino-Tibetan speakers it has been found in Japan. O3a1 developed a tail of at least 3 mutations before members of the haplogroup formed O3a1a-M121, O3a1b-M164 and O3a1c--002611. The last was the only haplogroup to have anything that could realistically be called 'an expansion'. It is spread fairly evenly through China, making up 20% of East asian Y-DNA. Its expansion certainly looks Neolithic.
To O3a2-P201. It is the Y-DNA haplogroup with the widest expansion, with members of O3a2c probably associated with the Austronesian/Polynesian expansion into the Pacific. But th ehaplogroup is most closely associated with populations speaking Sino-Tibetan languages and makes up the majority of Chinese Y-DNA. O3a2b is scattered through China although possibly concentrated in South China. That is hardly convincing evidence for the whole O Y-DNA haplogroup having originated in South China let alone SE Asia.
You cannot count SNPs at ISOGG and pretend to measure something with such a count. The Y chromosome is very large and AFAIK no haplogroup has ever been counted in all its length (it will happen, soon maybe, but not yet). That's a method that so far can only be used with mtDNA (and with caution in the case of less well-studied clades).
DeleteThe only thing you can count is STR diversity and haplogroup diversity (the latter being more important but not always sufficiently understood).
O3 and O1 originate in SE Asia (O1 is obvious and for O3 there is Hong Shi 2005), O2 is ambiguous, N originates in South China... Hence NO originated in SE Asia and so did O and N (South China). That's all there is to this issue.
"No, we do not. You hypothesize that such is the case but you fail to demonstrate it in any way. And, in any case, that is not what the authors mean".
ReplyDeleteOf course you don't know what the authors mean. To accept what they suggest conflicts with your belief. The authors do use the term 'selective pressure'.
"I have no reason to believe that such population is the origin".
Possibly not. But it makes sense. Again the only reason you refuse to accept it is that such an acceptance conflicts with your belief. Arguing with you is certainly like arguing with a fundamentalist preacher.
"Have you read the paper? Neither you nor I know what that map means, you are just making things up as you go".
Neither of us has read the paper but Razib seems to have access to more than we do. He obtained the map from somewhere and he says it is an attempt to find the region from which the EDAR mutation expanded from in the last 30,000 years. Why didn't you read what he wrote?
"Not me: the authors of this paper are saying that".
No they're not. Read what Razib wrote.
It does not make any sense that the only sample with the novel EDAR variant is fully fixated is the origin... unless you presuppose that a massive population replacement must have happened. You're biting your tail once again, Terry.
Deleteterryt wrote,
ReplyDelete"Unlike O1 O2-P31 accumulated at least 3 mutations before O2a-PK4 and O2b -M176 formed. Again O2a and O2b presumably developed within the margin of O2-P31' geographic range. O2a look likely to have been involved in the expansion of the Hmong-Mien and the Austro-Asiatic language expansion, but in several regions the local inhabitants later adopted either Tai-Kradai or Austronesian languages. O2a1a carried Munda into India and even reached the Pashtun. Presumably it is responsible for introducing the EDARV370A allele
into India. O2b is the northern version of O2. And let's not forget that its inclusion shifts O2's centroid fairly well north. In fact a recent paper made a convincing case for O2a to have spread south from the Three Gorges region of the Yangtze. That region is where the Yangtze emerges from the hills. A similar environment to where the EDARV370A allele emerged a little to the north."
Since the publication of the paper by Shi Yan, Chuan-Chao Wang, Hui Li et al., "An updated tree of Y-chromosome Haplogroup O and revised phylogenetic positions of mutations P164 and PK4" (European Journal of Human Genetics (2011) 19, 1013–1015), the PK4 mutation has been regarded as a marker of a newly defined Haplogroup O2a ("pre-M95"), with the former O2a-M95 being demoted to O2a1-M95. The PK4-carrying Y-chromosomes in the Tharus, Pashtuns, and some other populations of South Asia were published before the aforementioned update to the phylogenetic tree, and they belong to
O2a1a-M88. Munda Y-DNA is predominantly O2a1*-M95(xM88), with the remainder being comprised of typically South Asian clades (R-M17, H-M82, etc.); no published sample of Munda Y-DNA has included any representatives of O2a1a-M88.
O2a1a-M88 is rather a modern Southeast Asian clade, found especially in Vietnamese and populations of southern China (including Hans). I do not have a good explanation for its marginal presence in Indo-Iranian-speaking populations of South Asia, but until M88 is identified in a sample of Munda speakers, a hypothesis linking it to a Munda intrusion into South Asia is untenable.
"Munda Y-DNA is predominantly O2a1*-M95(xM88) ... no published sample of Munda Y-DNA has included any representatives of O2a1a-M88".
ReplyDeleteThanks for that clarification. I assumed that just a single clade of O2a had made it into and beyond India.
"until M88 is identified in a sample of Munda speakers, a hypothesis linking it to a Munda intrusion into South Asia is untenable".
That is very interesting.
"You cannot count SNPs at ISOGG and pretend to measure something with such a count. The Y chromosome is very large and AFAIK no haplogroup has ever been counted in all its length"
I did say 'at least' so many mutations.
"Have you read the paper? Neither you nor I know what that map means, you are just making things up as you go".
This is what Razib wrote regarding the second map:
"The second figure shows the inferred region from which the East Asian EDAR haplotype expanded over the past 30,000 years. The authors utilized millions of forward simulations with a host of parameters to model the expansion of EDAR, so that it fit the distribution pattern that is realized (see the supplements here for the parmeters). To make a long story short they infer that there was one mutation on the order of ~30,000 years before the present, and that it swept up in frequency driven by selection coefficients on the order of ~0.10 (10% increase relative fitness, which is incredibly powerful!). This is on the extreme end of selective sweeps"
I hope that helps you understand the situation in East Asia, but I'm sure it won't.
"O3 and O1 originate in SE Asia"
Only in your opinion. There is no evidence that they did so. In fact it's difficult to make such a case with the updated phylogeny so quoting a 2005 paper cannot be used as 'evidence'. In fact I'm sure you have blogged on that paper and I tried to point out the problems with it at the time. If you insist I will go through it once more for your benefit.
"N originates in South China..."
How on earth do you come to that conclusion? I presume you're basing it on the presence of some downstream N clade that has a very minor presence in SE Asia.
"That's all there is to this issue'.
If you're determined to ignore the evidence I suppose that statement is correct.
[Terry citing Razib:] "To make a long story short they infer that there was one mutation on the order of ~30,000 years before the present, and that it swept up in frequency driven by selection coefficients on the order of ~0.10 (10% increase relative fitness, which is incredibly powerful!)".
DeletePrecisely. That has nothing to do with what you say. The authors claim that allele expanded because it gave an extremely useful survival kit, 'an offer no one could reject', so to say. That is a selective sweep!
But of course the authors fail to convince (neither Razib nor me at least) very especially because none of the traits produced seems that kind life-or-death useful at all. We are not in the survival kit department of biology here but in the accessories one, which may make people look different but gives no or pretty trivial survival edges instead.
"I hope that helps you understand the situation in East Asia, but I'm sure it won't".
I understand it reasonably well. I would hope instead that some of our conversations would help you to understand that science demands humility and self-criticism but, sadly, I'm pretty sure by now that you won't ever incorporate these traits into your personality much less your analytical thought.
"O1 is obvious and for O3 there is Hong Shi 2005)"
ReplyDeleteOK. Let's look at the paper. An early statement:
"In East Asian populations, there are three regionally distributed (East Asian–specific) Y-chromosome haplogroups under the M175 lineage (fig. 1)—O3-M122, O2-M95, and O1-M119—together accounting for 57% of the Y chromosomes in East Asian populations (table 1)".
So their O2 excludes O2b, the northern version of O2. Note Ebizur's comments above. And the paper shows O3a branching into multiple separate basal clades, but we now know that O3 forms just four basal clades. Their O3a1, O3a2 and O3a4 form one of those clades as do their O3a3, O3a4 and O3a5. O3a6 and O3a7 are the only basal clades remaining from the phylogeny in the paper. So that takes care of a large proportion of the 'southern diversity' they claim for a start. Another thing to notice from their map is that the majority of groups they studied were from southern China. And they admit as much:
"Most of the populations sampled were from southern and southwestern China, where ~80% of the Chinese ethnic populations live"
So they ignore the majority of the Chinese population. Well done. Surely that would bias their data. Or had they decided in advance what they wanted to find?
"The Yangtze River was used as the geographic border to separate the SEAS and NEAS".
The Yangtze River is hardly a natural boundary, especially for any Neolithic flows. We actually know that it was a region of early presence of the Chinese Neolithic and source of expansion in fact.
"In the SEAS, there are 14 Tibeto-Burman–speaking populations with a recorded history of migration from northern China ~3,000 years ago (Wang 1994). The three Altai-speaking populations in Yunnan (southwestern China) were recent immigrants from northern China, <1,000 years ago"
They actually eliminate these populations from their calculations, thus adding further bias to their conclusions:
"To remove the influence of relatively recent population admixture, we constructed the STR network excluding the Tibeto-Burman, Altaic, Hmong-Mien, and southern Han populations".
And they say:
"In general, the distribution of the O3-M122 haplotypes did not show distinctive divergence between southern and northern populations, with all the major subhaplogroups shared between them—except for O3-M7, which was observed only in the southern populations and therefore indicates a recent common ancestry of the O3-M122 lineage in East Asia".
Hmmm ... I doubt you would agree with that very last bit! And then:
"Using the STR data, we calculated the gene diversities; no significant differences were observed between SEAS and NEAS or among different language groups (data not shown). The AMOVA analysis did not show significant between-group divergence either"
'No significant difference'? And:
"However, the MDS analysis showed that the NEAS are closely related by clustering together, whereas the SEAS showed relatively loose connections with larger variance, indicating that SEAS are genetically more polymorphic than are NEAS"
Isn't that what we would expect if the whole O3 haplogroup had originated in a small region somewhat to the north and then expanded sparodically into the southern hills?
"the O3-M122 lineages were probably dominant in the population involved in the initial northward migration; therefore, no obvious bottleneck occurred for the O3-M122 lineage, in contrast with the skewed distribution of the O2-M95 and O1-M119 lineages"
Now that should surely have put them on their guard against making too many rapid conclusions. Their claim doesn't actually make sense. I would suggest that it was the southward movement of O3 that was not subject to bottlenecks.
"Isn't that what we would expect if the whole O3 haplogroup had originated in a small region somewhat to the north and then expanded sparodically into the southern hills?"
DeleteNo. It indicates specific founder effects in the North.
"No. It indicates specific founder effects in the North".
ReplyDeleteTo me the evidence indicates founder effects from the north. You appear to have conveniently forgotten this paper:
http://comonca.org.cn/lh/doc/A37.pdf
Remember though that their O2 is just O2a. And O3* is O3(xO3a2b,O2a2c1), O3d is O3a2b, and O3e is O3a2c.
"It does not make any sense that the only sample with the novel EDAR variant is fully fixated is the origin..."
Surely that is the most likely scenario. Why would it not be, and how could it not be? The only other option is that it introgressed from another population where it may have formed somewhere within that population's geographic range. But the implies breeding with an Archaic population. The gene has to be fixated somewhere before it can expand, especially if it is recessive.
"unless you presuppose that a massive population replacement must have happened. You're biting your tail once again, Terry".
And this paper shows that such replacement is extremely likely to have occurred. The real problem arises for you because you assume a huge Paleolithic population in South China, but I believe you are very much mistaken in that assumption.
"The authors claim that allele expanded because it gave an extremely useful survival kit, 'an offer no one could reject', so to say".
But we both know very well that the later stages of the expansion did not involve any advantage at all from the gene itself. So when they say, 'there was one mutation on the order of ~30,000 years before the present, and that it swept up in frequency driven by selection coefficients on the order of ~0.10 (10% increase relative fitness, which is incredibly powerful!)' they're almost certainly talking about the original selective sweep within the region shown on the second map. If that was not the authors' intention why would they bother with the second map at all. They must have compiled it to demonstrate something, surely. But of course for you a consideration of that map would require a re-evaluation of your belief, and we can't possibly have that, so you are forced to ignore it.
"the authors fail to convince (neither Razib nor me at least) very especially because none of the traits produced seems that kind life-or-death useful at all".
Because you misunderstand what they're saying. You assume the gene spread through a static population, but that is not the case. Much of its expansion was by migration, and you are well aware of that fact but it doesn't fit your belief when it comes to South China and SE Asia. How could a 'selective sweep' have occurred in America when there was no-one there until the population that already had the gene arrived? I'm sure the authors are aware of that fact but you are determined to ignore it.
"I understand it reasonably well".
You've just proved you don't understand it at all.
Why don't you write a blog post at your blog with all you "know" about the origins of O? Seriously: it is infinitely more efficient than once and again here and there writing everything out from the blue. Just a link and a lot of work you don't have to do anymore.
DeleteI personally find your comments so long and combative (most often about the wrong things and with the wrong evidence) that I'm losing interest. You don't even seem to read and then reply anymore but you begin replying (confronting) as soon as you read the first sentence or almost.
The result are extremely and unnecessarily lengthy comments. Half of them empty personal accusations.
Hui Li 2007 is not too relevant here by the way, we see a lot of O3, still more diverse towards the South and we know nothing about other locations, notably Indochina, Sundaland... but also within modern China the blanks are huge. It does suggest (awaiting more data) that the coastal populations of around Shanghai have been largely replaced from people from maybe Hunan (SE Asian -i.e. "not NE Asian"- province of China distinguished for being the origin of the Neolithic and birthplace of Mao) or elsewhere. This might have also been the case with the people of the Yellow River(?? - they were not 02 in any case) but it says nothing at all about any migration southwards and it specifically seems to deny a Yellow River origin for any outstanding migration.
DeleteAs for the rest, more than enough. You should have very clear my viewpoint by now.
"But of course the authors fail to convince (neither Razib nor me at least) very especially"
ReplyDeleteThe difference between you and Razib is that he says:
"There is a weak point in one or more of my assumptions".
Now you would never admit that. I have never noticed any adherence to the 'science demands humility and self-criticism' on your part. But presumably his weak point is similar to your own so hopefully you will soon be able to explain things to him. He does say:
"If it is visible, and defines between populations differences, it is generally not genomically trivial. There’s usually a story underneath that difference".
You refuse to even consider the possibility that there is a story underneath the regional human difference, just random drift, bottlenecks and founder effects.
"The two populations in West Eurasia where you find the derived East Asian EDAR variant are Hazaras and Uyghurs, both likely the products of recent admixture between East and West Eurasian populations. In Melanesia the EDAR frequency is correlated with Austronesian admixture".
So again, unlike you, he accepts that some of the gene's distribution is a result of population expansion rather than of a selective sweep.
"And yet one thing that jumps out at me is that there is no East Asian EDAR in European populations, even in Russians. I am a bit confused by this result, because of the possibility of Siberian-affiliated population admixture with Europeans within the last 10,000 years, as adduced by several researchers"
This is not as much of a problem as either you or he makes of it. Of course any admixed population will not carry the same representative sample of genes from both parent populations.
"Much of the elegance and power of the paper applies to the discussion section as well, but to be frank this is where things start falling apart for me".
So, unlike you, he accepts the body of the research. It is just some of the conclusions the authors propose that he disagrees with.
"A climate adaptation is always a good bet. The problem I have with this hypothesis is that modern day gradients in the distribution of this allele are exactly the reverse of what one might expect in terms of adaptation to heat and humidity".
I agree 100% with Razib here. The authors themselves are off target with their explanation of how the gene provided a selective advantage.
"the derived EDAR mutant is found in high frequencies went through the incredible cold of the Last Glacial Maximum, and groups like the Yakuts are known to have cold adaptations today. Not only that, but the Amerindians from the arctic to the tropics all exhibit a cold adapted body morphology"
The evidence actually suggests that adaptation to seasonal cold is a much more likely explanation for the selective advantage. The 'sexual selection' hypotheses tell us more about the people proposing the hypothesis than about any selective sweep. As Razib says:
"The main question on my mind is how seriously prominent population geneticists such as Joshua Akey actually take sexual selection to be as a force driving variation and selection in human populations. It seems that quite often sexual selection is presented as a deus ex machina. A phenomenon which can rescue our confusion as to the origins of a particular suite of traits".
"This is not as much of a problem as either you or he makes of it".
DeleteI borrowed the idea from David "Polako" (a comment at GNXP and later also at his blog Eurogenes). Razib also sees it as a problem. Both believed the results produced by MixMapper (but I did not).
"Of course any admixed population will not carry the same representative sample of genes from both parent populations".
It should carry them in approximately equal apportions. Something is wrong here: either West Siberians lack the allele or there was some sort of "racist selection", there's no other reasonable explanation. Just stating something, as you do here (and do too often) does not make it true.
Erratum:
ReplyDeleteI have reviewed the original sources, and it appears that the four Tharus from Morang District of southeastern Nepal and the single adivasi from Andhra Pradesh reported to be O-PK4 by Fornarino et al. 2009 have been assumed to belong to O-M88 in accordance with the version of the Y-DNA phylogenetic tree that had been accepted at that time. These five individuals were found to be positive for M95, so they certainly do belong to O2a1-M95. However, their status regarding O2a1a-M88/M111 is undetermined.
On the other hand, the four Pathan individuals carrying PK4 belong to a subtribe (AusoKhel) of one of the major Pathan tribes (Yousafzai) from the Dir area in the North West Frontier Province (NWFP) of Pakistan, and these four individuals have been found to be positive for M95, M88, and M111 in addition to PK4. This is the source of the original misplacement of PK4 downstream of M88/M111.
Thus, it is certain that O2a1a-M88/M111 is found in at least one tribe of Pashtuns in the northwestern extremity of South Asia, but this subclade has not been confirmed in any sample that I have seen from the eastern parts of the Subcontinent. Kumar et al. (2007) did not find this subclade even in their samples of Khasis and Garos from Meghalaya in Northeast India.
O2a1a-M88/M111
ReplyDelete20/29 = 69.0% Xinhmul (Laos/Khmuic?)
15/34 = 44.1% Hani (Yunnan/Loloish)
12/39 = 30.8% Suy (Laos/West Katuic)
23/76 = 30.3% Kinh (Hanoi, Vietnam/Vietic)
9/31 = 29.0% Lowland Yao (NE Guangxi/Mienic)
10/37 = 27.0% Kataang (Laos/Ta'Oi–Kriang branch of Katuic)
5/23 = 21.7% Zhuang (Yongbei, C Guangxi/Tai branch of Kradai)
8/37 = 21.6% Zaomin (N Guangdong/Mienic)
6/35 = 17.1% Buyi (Guizhou-NW Guangxi/Tai branch of Kradai)
5/30 = 16.7% Alak (Laos/Central Bahnaric)
6/38 = 15.8% Aheu (Laos/Vietic)
5/36 = 13.9% Bunu (NW Guangxi/Hmongic)
4/31 = 12.9% Southern Mien (SW Guangxi/Mienic)
6/50 = 12.0% So (Laos/West Katuic)
3/25 = 12.0% Lao (Luang Prabang, Laos/Tai branch of Kradai)
4/34 = 11.8% Han (Chengdu, Sichuan/Sinitic)
2/19 = 10.5% Zhuang (Yongnan, C Guangxi/Tai branch of Kradai)
1/11 = 9.1% Top Board Mien
1/11 = 9.1% Thin Board Mien
5/59 = 8.5% Cham (Binh Thuan, Vietnam/Malayo-Polynesian branch of Austronesian)
4/50 = 8.0% Mal (Laos/Khmuic)
3/39 = 7.7% Zhuang (Hongshuihe, C Guangxi/Tai branch of Kradai)
3/41 = 7.3% Lowland Kimmun (NE Guangxi/Mienic)
12/166 = 7.2% Zhuang total (Guangxi & vicinity/Tai branch of Kradai)
1/15 = 6.7% Zhuang (Shangsi, SW Guangxi/Tai branch of Kradai)
2/33 = 6.1% Northern Mien (NW Guangxi/Mienic)
3/51 = 5.9% Hmong Daw (Laos/Hmongic)
3/51 = 5.9% Khmu (Laos/Khmuic)
1/17 = 5.9% Thai (Northern Thailand/Tai branch of Kradai)
1/18 = 5.6% Cambodia & Laos (Underhill et al. 2000)
1/19 = 5.3% Flower-head Mien (SW Guangxi/Mienic)
2/39 = 5.1% Taiwan (aborigines) (Taiwan/Austronesian)
5/100 = 5.0% Miao (Hunan/Hmongic)
1/20 = 5.0% Mountain Straggler Mien (Hunan/Mienic)
1/20 = 5.0% China (Underhill et al. 2000)
1/21 = 4.8% Zhuang (Guibei, NE Guangxi/Tai branch of Kradai)
1/22 = 4.5% Banjarmasin (Borneo, Indonesia/Malayo-Polynesian branch of Austronesian)
1/28 = 3.6% Bo (Laos/Vietic)
1/32 = 3.1% Brau (Laos/West Bahnaric)
2/65 = 3.1% Kota Kinabalu (Borneo, Malaysia/Malayo-Polynesian branch of Austronesian)
1/33 = 3.0% Qiang (Sichuan/Qiangic)
1/34 = 2.9% Li (Hainan/Hlai branch of Kradai)
1/35 = 2.9% Yao (Liannan, Guangdong/Mienic)
1/35 = 2.9% Talieng (Laos/Central Bahnaric)
1/41 = 2.4% Native Mien (NE Guangxi/Mienic)
1/49 = 2.0% Miao (Guizhou/Hmongic)
1/50 = 2.0% Bai (Dali, Yunnan/Tibeto-Burman? Sinitic?)
1/50 = 2.0% Laven (Laos/West Bahnaric)
1/50 = 2.0% Oy (Laos/West Bahnaric)
2/160 = 1.3% Han (Wuhan, Hubei/Sinitic)
O2a1a-M88/M111 appears to be distributed geographically rather than ethnically, with highest concentration in a contiguous area centered on Guangxi Province of southern China. Frequencies of this haplogroup are high in most of Guangxi, in southern Yunnan (judging from the Hani), and in populations of the neighboring countries of Vietnam and Laos, pretty much regardless of ethnolinguistic affiliation. The frequency gradually fades out into Thailand, northern Yunnan, Sichuan, Guizhou, Hunan, Hubei, Guangdong, Hainan, Borneo, Taiwan, etc. Burma and Bangladesh are large blanks in my knowledge of haplogroup distributions, but the apparent total lack of O2a1a-M88/M111 in Nepal, Tibet, and (at least most parts of) Northeast India makes me suspect that this haplogroup probably does not spread much to the west of Shan-inhabited areas of Burma.
This is all very interesting, generally speaking, Ebizur. But, honestly, I'm not even sure anymore (or even to begin with) why O2a1a is relevant to this debate. Could you explain?
DeleteIt is only relevant in a very tangential manner (if at all), which was my original point. If you review my comments in this thread, you should see that I started posting in response to the following comment by terryt:
Delete"Unlike O1 O2-P31 accumulated at least 3 mutations before O2a-PK4 and O2b -M176 formed. Again O2a and O2b presumably developed within the margin of O2-P31' geographic range. O2a look likely to have been involved in the expansion of the Hmong-Mien and the Austro-Asiatic language expansion, but in several regions the local inhabitants later adopted either Tai-Kradai or Austronesian languages. O2a1a carried Munda into India and even reached the Pashtun. Presumably it is responsible for introducing the EDARV370A allele into India. O2b is the northern version of O2. And let's not forget that its inclusion shifts O2's centroid fairly well north. In fact a recent paper made a convincing case for O2a to have spread south from the Three Gorges region of the Yangtze. That region is where the Yangtze emerges from the hills. A similar environment to where the EDARV370A allele emerged a little to the north."
I was refuting terryt's hypothesis of a connection among Y-DNA haplogroup O2a1a-M88, Munda languages, and the presence of the derived EDAR allele in mainstream populations of South Asia (e.g. Gujaratis) at very low frequencies. It is still possible that the people who have introduced the ancestral form(s) of the Munda languages to India may have been responsible for introducing some (or all) of the derived EDAR allele that is marginally found there, but terryt's attempt to additionally link Y-DNA haplogroup O2a1a-M88 to this hypothetical Proto-Munda expansion into India is unfounded, since O2a1a-M88 has not been
found in speakers of Munda languages nor in any other population of India as far as I know. In the greater South Asian region, O2a1a-M88 seems to have been confirmed only in a certain tribe of Pashtuns who presently inhabit an area in the northwest of Pakistan, with Chitral to the north and Kunar Province of Afghanistan to the west.
Ah, alright. Thanks for the explanation.
Delete"these four individuals have been found to be positive for M95, M88, and M111 in addition to PK4".
ReplyDeleteThaks for that information. I note the four Tharus from Morang District of southeastern Nepal and the single adivasi from Andhra Pradesh may belong to that haplogroup but I presume the Munda O2a definitely does not. Is that still so?
"It is still possible that the people who have introduced the ancestral form(s) of the Munda languages to India may have been responsible for introducing some (or all) of the derived EDAR allele that is marginally found there, but terryt's attempt to additionally link Y-DNA haplogroup O2a1a-M88 to this hypothetical Proto-Munda expansion into India is unfounded"
However O2a may have introduced both the haplogroup and the language.
"O2a1a-M88/M111 appears to be distributed geographically rather than ethnically, with highest concentration in a contiguous area centered on Guangxi Province of southern China".
But we know there has been considerable language shuffling in the region. Tai-Kradai and the Sino-Tibetan languages presumably overly a substrate of Austro-Asiatic and Hmong-Mien languages. Hmong-Mien languages are discontinuously distributed suggesting they are remnants of a once more widely-spoken language group:
http://en.wikipedia.org/wiki/Hmong%E2%80%93Mien_languages
And Wiki even says:
"Early linguistic classifications placed the Hmong–Mien languages in the Sino-Tibetan family, where they remain in many Chinese classifications, but the current consensus among Western linguists is that they constitute a family of their own. The family is believed to have had its origins in central-southern China. The current area of greatest agreement is that the languages appeared in the region between the Yangtze and Mekong rivers, but there is reason to believe that speakers migrated there from further north with the expansion of the Han Chinese".
And Austro-Asiatic/Mon-Kmer languages also have a discontinuous distribution:
http://en.wikipedia.org/wiki/Austroasiatic_languages
I notice that Pakanic looks to be found in Guangxi Province of southern China. It seems not to be a root Mon-Kmoer languge although there does seem to be disagreement as to its position.
http://en.wikipedia.org/wiki/Khasi%E2%80%93Khmuic_languages
"However, Sidwell (2009, 2011) notes that Khmuic and Pakanic (or at least Mangic) appear to be independent branches, and that the only valid connection is Khasi–Palaungic.
The extent of the recently described Pakanic branch is contested, with some including the Mangic languages and others placing those in Palaungic".
"I personally find your comments so long and combative"
ReplyDeleteMy comments are combative because you're absolutely committed to a particular belief, so much so that you're thoroughly inconsistent as to how you interpret data. And when the data conflicts with your belief you simply dismiss it. An example of the latter approach:
"Both believed the results produced by MixMapper (but I did not)".
Yes. You refuse to believe it because to do so would require you to seriously alter your belief: that haplogroups have not shifted since the Upper Paleolithic. An example of your altering your interpretation to fit your belief:
"It indicates specific founder effects in the North".
Now. You have consistently claimed that offspring haplogroups cannot replace parent haplogroups while confined to their region of origin because parent haplogroups will dominate numerically and cause offspring haplogroups to drift to extinction. It is generally accepted that novel haplogroups usually develop along the advancing front of a haplogroup's expansion. But here you are claiming the opposite.
"It should carry them in approximately equal apportions. Something is wrong here: either West Siberians lack the allele or there was some sort of "racist selection", there's no other reasonable explanation. Just stating something, as you do here (and do too often) does not make it true".
Any farmer will tell you that although the first cross of any two breeds (F1) tend to all look the same (dominant genes from each breed are expressed) once you interbreed the hybrids (F2) almost any combination can be thrown up. There is no reason at all why two populations formed from even the same two original populations will have the same genes. And the difference will be even more marked if the admixture consists of other than a mixture of other than 50% from each parent population.
"It does suggest (awaiting more data) that the coastal populations of around Shanghai have been largely replaced from people from maybe Hunan"
Yes. And those 'original' coastal populations look to have been mainly O1. That haplogroup expanded south into Taiwan and beyond.
"This might have also been the case with the people of the Yellow River(?? - they were not 02 in any case)"
At present we have no idea of what haplogroups were present in the Hwang Ho before the Neolithic.
"You should have very clear my viewpoint by now".
I am. And I realise no amount of evidence is going to change it.
I have the right to be "wrong" (i.e. in disagreement with you or whoever) and you are not persuading me at all, as you know. Surely because you are too combative and wasteful of words. You have no excuse in any case to be excessive lengthy and unfocused, always bringing too many matters to any single discussion and wasting too much time accusing me of your imagination's "sins" and (sometimes at least intentionally) misunderstanding half of what I say, so you have even more imaginary space to yell at. For example above you waste four paragraphs (two quotes and two of your own) just yelling at me and accusing me of "heresy" or something.
DeleteEn fin I do not expect you to amend at your advanced age but, please, surprise me.
"... once you interbreed the hybrids (F2) almost any combination can be thrown up".
And that supports "racist selection". The farmer acts with intention of picking these or those traits and so does "racist selection", focused on the visible phenotype, some of which are preferred (for example blonds with blue eyes, or, in this case, people who look less East Asian). With no selection against some traits, there should be nothing being "thrown up" at all.
"And those 'original' coastal populations look to have been mainly O1. That haplogroup expanded south into Taiwan and beyond".
Or maybe came from there (from Taiwan and beyond) via the coast. Or are you telling me that you have genetic reasons of diversity or obvious ancestry (in the case of aDNA → modern DNA) that backs a Shanghai area origin for Austronesians? Just because you wish so it does not make it true: look at the two sides of the issue before picking your "truth".
"At present we have no idea of what haplogroups were present in the Hwang Ho before the Neolithic".
Actually the link you insist on, Hui Li 2007, includes a Yellow River area sample from Taosi (Shanxi, Longshan culture), which 3/4 O3* and 1/4 O3e (the only such finding of today's most common "Sino-Tibetan" lineage). You go half-blind on all this, even in what you could use (in your usual one-sided and sloppy manner) to support your models.
I was going to leave this topic but then you made this idiotic comment:
ReplyDelete"And that supports 'racist selection'. The farmer acts with intention of picking these or those traits and so does 'racist selection'"
One minute you're stressing the variation within 'racial types' and now you're claiming racial selection has made each type uniform. The F2 human hybrids between groups have maintained considerable diversity. Minimal selection.
I'm not talking about generalities here but the particular case of Northern Russia, which is anomalous considering what we know about Siberian/East Asian admixture in that area: it is very strange that with a 10% or more such admixture not even a thin red line appears in the map. That requires an explanation and the best I could come up with was "racist selection" in favor of European looking phenotypes.
DeleteDo you have better ideas on the matter, smartass? No? Well, the don't be so cocky and insulting.
"Do you have better ideas on the matter, smartass?"
ReplyDeleteYes, I do. 10% East Asian admixture is far less than that of America, Northeast Siberia or Southeast Asia. The anomaly is the Munda. Perhaps they have not admixed with their neighbours much at all since the arrived.
Which are the admixture levels in Near Melanesia? We do see a 5% or so of this EDAR variant there. But in North Russia: 0%. That is most strange.
DeleteSeems the Munda are not so anomalous after all:
ReplyDeletehttp://blogs.discovermagazine.com/gnxp/2010/10/sons-of-the-conquerers-the-story-of-india/#.USwtM_KJmuI
"Indian AA speakers have high frequencies of Y chromosome haplogroup O2a"
In that blog we find the information that:
"The G allele exhibits co-dominance"
So its influence is in direct prortion to its presence. Even this paper, which claims the Munda are the 'original' South Asian people (an unlikely hypothesis in fact):
http://www.ispub.com/journal/the-internet-journal-of-biological-anthropology/volume-4-number-2/munda-speakers-are-the-oldest-population-in-india.html#sthash.718qY3ts.dpbs
Says:
"The Munda y-chromosome is O2a (M95)".
If the proportion of O2a is as much as 50% of the Y-DNA haplogroups that gives 25% East Asian genes. That is more than twice the level in Northern Russia.
I didn't realise the importance of this research until a few days after you had posted it. It reveals a huge amount abouth East Asian prehistory.
ReplyDelete"Which are the admixture levels in Near Melanesia? We do see a 5% or so of this EDAR variant there".
Razib explains the problem you see:
"In Melanesia the EDAR frequency is correlated with Austronesian admixture".
Of course Melanesians have minimal amounts of Y-DNA O. But male expansion is not the only vector for the EDAR370A's expansion. Female expansion can be another vector. I realise that you believe mt-DNA B is ancient in Melanesia. But most believe it arrived with the Austronesians. B4's deeper origin is most likely the South China coast. By the time B4a reached Taiwan, and certainly by the time it left, the EDAR370A variant was almost certainly well established through southern China. So Razib is correct: the Austronesians carried the variant into Melanesia.
This is probably a good time to respond to some of your doubts about Y-DNA's origin.
"Or maybe came from there (from Taiwan and beyond) via the coast. Or are you telling me that you have genetic reasons of diversity or obvious ancestry (in the case of aDNA → modern DNA) that backs a Shanghai area origin for Austronesians?"
Yes, I do have good reasons to doubt a movement north from Taiwan to the mouth of the Yangtze rather than in the other direction. O1 is quite rare on the mainland, including the Shanghai region, today. It would be very strange if a SE Asian/Taiwan haplogroup had moved to to make up three quarters of the population near Shanghai to later diminish to almost nothing again. I know Wikipedia insists on a southern origin:
http://en.wikipedia.org/wiki/Haplogroup_O-MSY2.2_(Y-DNA)
But that insistence leads to the contradictions in the article. Such as:
"Neolithic incursions made only a minor impact on the paternal gene pool, despite the large cultural impact of the Austronesian expansion".
We know that what we may prefer to call 'post-Neolithic incursions' must have had a huge impact on the Shanghai region gene pool then.
"Approximately 5000 BCE, Haplogroup O-MSY2.2 coalesced at Sundaland and migrated northwards to as far as Taiwan"
No. We know that it became very prominent on the mainland way to the north of Taiwan. It seems very likely that O1a has been replaced on the mainland to a large extent by the later expansion of O3 haplogroups. It became virtually confined to Taiwan and SE Asia from where it expanded further as part of the Austonesian movements.
"Actually the link you insist on, Hui Li 2007, includes a Yellow River area sample from Taosi (Shanxi, Longshan culture), which 3/4 O3* and 1/4 O3e (the only such finding of today's most common "Sino-Tibetan" lineage)".
Have you checked out the region where the Taosi culture developed?
http://en.wikipedia.org/wiki/Taosi
"Taosi (Chinese: 陶寺; pinyin: Táosì) is an archaeological site in Xiangfen County, Shanxi, China".
But there's more:
http://en.wikipedia.org/wiki/Shanxi
"The name Shanxi literally means 'mountain's west', which refers to the province's location west of the Taihang Mountains.[1] Shanxi borders Hebei to the east, Henan to the south, Shaanxi to the west, and Inner Mongolia to the north and is made up mainly of a plateau bounded partly by mountain ranges".
That is the exact same region that the secong map Razib posted is centred on. The region where the EDAR370A gene swept to prominence. It seems most reasonable to me to suppose O3-M122 coalesced in the region that O-M175 had coalesced in the first place. O1 and O2 had coalesced as O-M175 had expanded from that region. What we may be seeing here is evidence that O3 had remained behind in the region of O-M175's coalescence while O1 and O2 had pushed south.
If you look at autosomal data, the percentage of East Asian influence in Island Melanesians can't be much higher than among Finnic-ancestry peoples of Northern and NE Europe, including Northern Russians. Two examples from Dienekes (global K=5 shows almost 20% but global K=15 makes them a single distinct cluster with 0% apparent admixture, the results may vary somewhat depending on K-level and particular sample anyhow). The same that Melanesians have mtDNA as more outstanding haploid signal of that admixture, Finnic-ancestry peoples have especially Y-DNA N (c. 50% in Finland for example) as such signature. They are clearly comparable.
Delete"I realise that you believe mt-DNA B is ancient in Melanesia"...
Again you misunderstand and misrepresent my opinion: B is clearly of Austronesian origin - at least I think so.
"It would be very strange if a SE Asian/Taiwan haplogroup had moved to to make up three quarters of the population near Shanghai to later diminish to almost nothing again".
Regardless of the exact origin, that is exactly what the aDNA mentioned above suggests: that the Shanghai area population was dominated by O1a and now it is not. Whether that O1a came from Taiwan or from Mars is irrelevant to its demise/dilution.
"[Shanxi] is the exact same region that the secong map Razib posted is centred on".
It is not. The red region corresponds to Henan, south of Shanxi. As I mentioned before, without reading the paper, we cannot interpret its real value (it may well be a ghost, an statistical artifact).
It is irrelevant anyhow because the relevance of the Taosi sample is that it doesn't either appear as "modern" in any way. You are just over-reading in search for burning nails again.
"You are just over-reading in search for burning nails again".
ReplyDeleteSorry, your Majusty.
"If you look at autosomal data, the percentage of East Asian influence in Island Melanesians can't be much higher than among Finnic-ancestry peoples of Northern and NE Europe, including Northern Russians. Two examples from Dienekes (global K=5 shows almost 20% but global K=15 makes them a single distinct cluster with 0% apparent admixture"
How has that got anything to do with the matter under discussion here? As I understand things from the context the mutation is confined to Austronesian-speaking groups. The only populations Dienekes entry mentions are 'Papuans' and 'Non-Austronesian Melanesians'.
"The same that Melanesians have mtDNA as more outstanding haploid signal of that admixture, Finnic-ancestry peoples have especially Y-DNA N (c. 50% in Finland for example) as such signature. They are clearly comparable".
No they're not. Austronesian-speaking Melanesians have a considerable proportion of mt-DNA B. Although Nenets and Sami have considerable proportions of N it seems they were not sampled in this study. Do you have any information as to the proportion of N in the Vologda? But aside from that you have consistently maintained that mt-DNA is a better indication of aDNA than is Y-DNA. So it is you here who is clutching burning nails.
"Regardless of the exact origin, that is exactly what the aDNA mentioned above suggests: that the Shanghai area population was dominated by O1a and now it is not".
Yes. It has been replaced by O3. To claim that a haplogroup that expanded south from Taiwan a mere 4-5000 years ago had, in a completely unrelated expansion, reached the Shanghai area is unsopportable.
"Whether that O1a came from Taiwan or from Mars is irrelevant to its demise/dilution".
True. But the matter under consideration here is the expansion of the EDAR370A variant. And the sourse of O1 has everything to do with that topic. Unless you're proposing some alternative vector for the variant's expansion? Come on. Out with it.
" the relevance of the Taosi sample is that it doesn't either appear as 'modern' in any way".
What?! One third of the haplogroups there are O3a2c1, the most widely distributed modern Chinese Y-DNA. And we can only be sure that the two thirds shown as O3* is not O3a2c1-M134 or O3a2b-M7. It could be O3a2a-M159, O3a1-L127 or even O3a2c-P164(xO3a2c1-M134). Burning nails again your Majusty.
"The red region corresponds to Henan, south of Shanxi".
I don't think so! Check your Chinese geography. The most that could be said is that the red zone is near the Shanxi/Henan border, exactly where the Taosi culture developed. Burning nails yet again.
I just judging on the map, because I have no access to the paper, and the Melanesians with the EDAR "red" variant are Island Melanesians. Papuans do not have it. All Island Melanesians I know of have some apparent (even if variable and generally minor) Austronesian-East-Asian admixture and the red fraction is clearly correlated with admixture. Similarly all Finnic-ancestry Northern Europeans have some apparent (even if variable and generally minor) Siberian-East-Asian admixture but they don't appear to have the EDAR variant (according to what we can see in the maps at least).
Delete"No they're not. Austronesian-speaking Melanesians have a considerable proportion of mt-DNA B".
And Finnics have a considerable proportion of Y-DNA N. Both have similar autosomal admixture signals from East Asia.
"Although Nenets and Sami have considerable proportions of N".
I'm not talking of Nenets nor Sámi (not of Nenets in any case, who are autosomally much more East Asian than anything else). I'm talking of Finns, Karelians, Northern Russians, Volga Finnics, Estonians and Latvians - read at least Wikipedia before commenting on a matter you seem to know nearly nothing.
"But aside from that you have consistently maintained that mt-DNA is a better indication of aDNA than is Y-DNA".
But it does not matter here, because at least I (not sure about you) have seen many ADMIXTURE and STRUCTURE analysis that always confirm that those peoples have some 15% Siberian-East-Asian admixture (with whatever particular variations). Equally I have seen a number of such autosomal analysis for Island Melanesians, which give similar apportions.
Why to look at the part when you already have the whole?
" To claim that a haplogroup that expanded south from Taiwan a mere 4-5000 years ago had, in a completely unrelated expansion, reached the Shanghai area is unsopportable".
Why? You are being just knee jerk.
Anyhow I don't believe that O1a as a whole expanded from Taiwan: it did first from Indonesia probably, and the Taiwan centered secondary expansion southwards only had a minor impact (mayor in some isolated populations). But do not start a discussion on the origins of O1a, please. It does not matter at all. They may well be unrelated processes, why not? (although IMO the single S→N coastal expansion can well explain both the arrival to Taiwan and to Shanghai region).
"But the matter under consideration here is the expansion of the EDAR370A variant".
Exactly and it does not seem particularly related to O3, otherwise Native Americans would be drowning in O3, or also Mongolians, Siberians, Burmese, Koreans, Japanese, etc. We know it is not the case and that's why the expansion seems to be from a less specific source and in a less specific process.
"One third of the haplogroups there are O3a2c1"... [in Taosi]
1/4 or 1/5. But the other 3/4 are O3*, what is inconsistent (today's Sino-Tibetans only have 17% of that). So it is impossible that modern ST Y-DNA pool spawned from Taosi.
You just don't seem to be able to see the whole picture and cherry-pick your "evidence" without apparently being aware of any contradictions. Rebating that is for me a waste of time, sorry, a time that I do not have anymore.
"I don't think so!"
I do. I contrasted the maps and it is very clear. But anyhow I don't know yet why the red region would be meaningful: nobody explains that clearly.
"But anyhow I don't know yet why the red region would be meaningful: nobody explains that clearly".
ReplyDeleteAbsolute rubbish. Are you really so silly? The red shows where the authors calculated that the variation came to dominance. Have you not read Razib's post?
"the Melanesians with the EDAR 'red' variant are Island Melanesians".
You seem to have shifted what you're talking about. You said, 'The red region corresponds to Henan, south of Shanxi'. What has that to do with Melanesians of any kind? But to move to what you're talking about here: even though they are 'Island Melanesians' they are specifically 'Non-Austronesian Island Melanesians'. We would therefore expect them to lack the EDAR370A variation. Read the post. Don't simply make things up.
"I'm talking of Finns, Karelians, Northern Russians, Volga Finnics, Estonians and Latvians"
None of which appear to have been tested in the relevant paper. Read Razib, again if necessary.
"Why to look at the part when you already have the whole?"
Because when it comes to Melanesia the parts are each very different from each other, even though you consistently lump everything together as 'Australasian'. Read at least something on the subject before commenting on a matter you seem to know nearly nothing about.
"Anyhow I don't believe that O1a as a whole expanded from Taiwan: it did first from Indonesia probably"
And you believe that because ...? Any evidence?
"Exactly and it does not seem particularly related to O3"
On the contrary, through most of the southern regions where the variant is found it is closely associated with Y-DNA O.
"otherwise Native Americans would be drowning in O3, or also Mongolians, Siberians, Burmese, Koreans, Japanese, etc."
You are being deliberately obtuse here, your Majusty. Obviously the variant is associated with Y-DNA N and C3, and possibly C1, as well as with O. And you are very well aware of that fact because you keep bringing up the subject of Northern Eurasia. But many mt-DNA haplogroups also look to have expanded, or at least partially expanded, from the region where the variant originated.
"We know it is not the case and that's why the expansion seems to be from a less specific source and in a less specific process".
In your imagination. We can be almost certain that it is associated with a selective sweep within the region shown in the map at Razib's blog, and a subsequent population expansion from that region. To plead ignorance as to any process is simply a ploy to avoid facing the facts.
"But the other 3/4 are O3*, what is inconsistent (today's Sino-Tibetans only have 17% of that)"
Yet another ridiculous comment. Most Sino-Tibetans are O3 of some sort. You're closing your eyes to facts again. All we know is that the O3* shown in the map is not O3a2b or O3a2c1. That still leaves a number of possibilities other than the O3* you're insisting on. And surely the fact that O3a2c1 was found only in the Taosi sample yet is now the most widely distributed and common Y-DNA O in China indicates it expanded southwards, not northwards.
"You just don't seem to be able to see the whole picture"
Your latest comments on this topic show absolutely that you carefully avoid looking at the whole picture.
"Absolute rubbish. Are you really so silly?"
DeleteDoubly insulting comment. Nth warning!
"The red shows where the authors calculated that the variation came to dominance"...
While I can't read HOW they calculated it I do not care. I've been begging for a free copy of the paper since comment #1 or so.
Their calculation can always be wrong and I refuse to accept any conclusion just because someone says so. Data!
"You seem to have shifted what you're talking about. You said, 'The red region corresponds to Henan, south of Shanxi'. What has that to do with Melanesians of any kind?"
"Red" as in EDAR370A variant marker in the main map above. That is real data and that's what I'm discussing when we talk of admixture. You seem to understand nothing at all, and I'm robbing time from other activities to pay attention to your ignorance and confusion.
All you the rest say are one-liners devoid of meaning, words to argue words, without any content, without any single general sense, almost babbling. Seriously...
"Most Sino-Tibetans are O3 of some sort."
Generic claim devoid of any meaning: your accidental claim that Sino-Tibetans or whatever come from Taosi clashes with the known composition of that sample. O3* in this context is not "03 of some sort" but specifically excluding this and that and that other clade. Also there are other lineages. 1:3 O3e:O3* apportion creates after expansion 1:3 O3e:O3* apportion (plus whatever remnants of assimilated substrate pools), and what we see in modern ST is 2:1 O3e:O3*, what must have other more complex (or simpler but unknown, not Taosi) origins.
Seeing you have not actually read Razib's post I'll put some of his comments here:
ReplyDeleteMaju: "But anyhow I don't know yet why the red region would be meaningful: nobody explains that clearly".
Razib: "The second figure shows the inferred region from which the East Asian EDAR haplotype expanded over the past 30,000 years".
I'm sorry, but I can't explain it any more simply than that.
Maju: "the Melanesians with the EDAR 'red' variant are Island Melanesians. Papuans do not have it".
Razib: "In Melanesia the EDAR frequency is correlated with Austronesian admixture".
In spite of what you claim, not all island Melanesians have Austronesian-East-Asian admixture. Only the Austronesian-speaking populations and their very near neighbours.
Maju: "All Island Melanesians I know of have some apparent (even if variable and generally minor) Austronesian-East-Asian admixture"
Obviously incorrect.
That is not any explanation: it is a description of what the authors say. I don't want you to explain more simply but more in depth, detail, with enough data for me to comprehend HOW they reached that conclusion. Meanwhile for me it is trivial, pointless, meaningless.
Delete"Razib: "In Melanesia the EDAR frequency is correlated with Austronesian admixture"".
Exactly my point as well. But then among Finnic peoples, which have similar amounts of East Asian admixture, that does not happen and it is most strange.
We are not discussing Melanesians here but Finnic-ancestry Northeastern Europeans!
"Generic claim devoid of any meaning"
ReplyDeletehttp://forwhattheywereweare.blogspot.co.nz/2011/04/haplogroup-o3-downstream-structure.html
"your accidental claim that Sino-Tibetans or whatever come from Taosi clashes with the known composition of that sample. O3* in this context is not '03 of some sort' but specifically excluding this and that and that other clade".
Incorrect. Specifically excluding only two clades: O3a2b-M7 and O3a2c1-M134. So their O3* could be O3a2c*, O3a2a, O3a2* or O3a1, most likely O3a1c which makes up 16.9% of the modern Chinese Y-DNA and is widely, and evenly, spread through China (see above link).
"Also there are other lineages. 1:3 O3e:O3* apportion creates after expansion 1:3 O3e:O3* apportion (plus whatever remnants of assimilated substrate pools), and what we see in modern ST is 2:1 O3e:O3*, what must have other more complex (or simpler but unknown, not Taosi) origins".
Remember that their 'O3e' is actually O3a2c1-M134. That haplogroup now makes up around 30% of Chinese Y-DNA. Again widely and evenly spread through China (see above link). I agree its origin may be complex but that doesn't rule out a northern origin. Note that O3a2c1-M134 was not found in any of the other Neolithic regions tested. It is confined in this study to the Taosi, although that tells us nothing about its earlier history. We can be reasonably sure of some level of continuity though. O3a2b-M7 (O3d in the study) is still concentrated in the Daxi and Hmong Mien. So O3a2c1's concentration in the region from where the later Chinese Han expansion originated may be significant.
"Their calculation can always be wrong and I refuse to accept any conclusion just because someone says so. Data!".
At present we just have to accept their map. I note Razib seems to accept it at face value. But you are obviously not even prepared to consider the possibility that their conclusion may be extremely sound. Why is that?
"But then among Finnic peoples, which have similar amounts of East Asian admixture, that does not happen and it is most strange".
You find it 'strange' only because you are confusing two sets of data. You are comparing apples with oranges. I repeat: Dienekes post considers only Papuans and non-Austronesian Melanesians. Both the Papuans and the Melenesians have a level of East Asian admixture, with the Melanesians having more than Papuans. That makes complete sense as it results from introgression from Austronesian-speaking peoople. The level in these non-Austronesian Melanesians is comparable to the level in the North Russians. Neither group has the EDAR variation. Now, have another look at the map you posted above. You will see two population sets from the New Guinea/Melanesia region. One has no EDAR370A even though, presumably, it has some level of East Asian ancestry. The other, as specifically stated by Razib, is 'Austonesian Melanesians'. These do have the variation and presumably have a higher level of East Asian ancestry. Higher than the North Russian population shown in the diagram. Why do you have such difficulty seeing that? Or is it that:
"You seem to understand nothing at all"
It seems that I can't speak English anymore or that you just don't care at all about what I say. The Hiu Li paper already has a comparison where we can see that:
Delete1. Taosi (Longshan): 75% O3* - 25% O3e - (0% O3d - 0% O2a - 0% O1a)
2. Modern Sino-Tib.: 17% O3* - 33% O3e - (1% O3d - 8% O2a - 2% O1a)
The signatures do not match. From 75% to 17% is a HUGE difference!
I'm not "comparing apples and oranges": the two populations are perfectly comparable on their East Asian admixture levels! I don't care what you (mis-)understood about what Dienekes might have said: there are two obviously different pie charts in that map: one for Papuans and another for Island Melanesians (from maybe New Brunswick?) And those Island Melanesians are who are perfectly comparable in this aspect of minor but noticeable East Asian admixture with NE Europe's Finnic-ancestry peoples.
DeleteBy the Monkey King Sun Wukong, Equal to Heaven!
"The signatures do not match. From 75% to 17% is a HUGE difference!"
ReplyDeleteAre you claiming haplogroup proportions in a particular region never change? Come on. Anyway, even here you are comparing apples with oranges. Taosi is just one small sample within a whole Sino-Tibetan-speaking region. The figures do not tell us the O3(xO3d,o3e) figure in the immediate vicinity of Taosi today.
"there are two obviously different pie charts in that map: one for Papuans and another for Island Melanesians"
Maju. I repeat: non-Austronesian Melanesians.
"I'm not 'comparing apples and oranges': the two populations are perfectly comparable on their East Asian admixture levels!"
Where on earth did you get that idea from? You have seen the first of these papers before. The last is an extract from a Bellwood book:
http://mbe.oxfordjournals.org/content/25/7/1362.long
"Thus, the Austronesian language replacement on the Admiralties (and elsewhere in Island Melanesia and coastal New Guinea) was accompanied by an incomplete genetic replacement that is more associated with mtDNA than with NRY diversity".
Leaving non-Austronesian Melanesians with less of the EDAR variation.
http://www.ncbi.nlm.nih.gov/pubmed/15009807
Again showing a distinction between Austronesian and non-Austronesian Melanesians.
http://www.ncbi.nlm.nih.gov/pubmed/18208337
Quote:
"A small 'Austronesian' genetic signature (always <20%) was detected in less than half the Melanesian groups that speak Austronesian languages, and is entirely lacking in Papuan-speaking groups".
http://books.google.co.nz/books?id=Ovm_HRqOJH0C&pg=PA110&lpg=PA110&dq=austronesian+melanesian+genes&source=bl&ots=UP3DRrGHi9&sig=0L0dRm8piywsinNjiiRqTBSSwf0&hl=en&sa=X&ei=8BkxUZvbG-qSiQfSsoCQBg&ved=0CGEQ6AEwBw#v=onepage&q=austronesian%20melanesian%20genes&f=false
"Are you claiming haplogroup proportions in a particular region never change?"
DeleteEverything eventually changes but, when it does, it is for some reason. So if YOU are claiming such a change, which is very extreme, up to the point that the two genetic pools don't look the same AT ALL, you first need to offer a very good explanation.
"Where on earth did you get that idea from?"
If you would have read what I said (with informative links) previously about the autosomal genetics of Island Melanesians, you'd knew by now. I'm not willing to repeat myself once and again because you write faster than read.
"A small 'Austronesian' genetic signature (always <20%) was detected in less than half the Melanesian groups that speak Austronesian languages"...
Exactly my point.
A small East Asian genetic signature <20% is also present in all Finnic-ancestry peoples of NE Europe.
"Modern Sino-Tib.: 17% O3* - 33% O3e - (1% O3d - 8% O2a - 2% O1a)"
ReplyDeleteHere is the paper the data Liu used in his table comes from:
http://www.ncbi.nlm.nih.gov/pmc/articles/pmc1288383/
The phylogeny has been considerably altered since 1999, and so little as to the the significance of O3*' distribution can be gained.
"The signatures do not match".
Why should they? The Liu paper says:
"A rare haplogroup, O3d, was found at the Daxi site in the middle reaches
of the Yangtze River, indicating that the Daxi people might be the ancestors of modern Hmong-Mien populations, which show only small traces of O3d today".
So the signatures do not match for the Daxi either. In fact they also say:
"The absence of O3d in the historical samples from the Daxi site (it might not have been found because of the small sample size), and the migration of
modern Hmong-Mien populations to the southwest might indicate that the prehistoric population in the Three Gorges area has been replaced".
So O3a2b has been completely replaced in the region where it was present 6500-5500 years ago. And the authors obviously accept that the Daxi/Hmong Mien and their Y-DNA have moved south, not north.
And, lastly:
"at least 62.5% of the individual remains (30 out of 48) belong to O haplogroup, which is still the major haplogroup of today’s East Asians".
That is actually somewhat less than the proportion today. The smallest proportion of O shown in the modified Y-DNA O phylogeny paper is 69.2%, in 'Northern China'. So the overall proportion of O in China has changed since the Late Neolithic. And we are talking small sample size in the Neolithic study. Just five individuals from Taosi.
"A small East Asian genetic signature <20% is also present in all Finnic-ancestry peoples of NE Europe".
ReplyDeleteMuch less than 20% is my understanding. Less than 10% even. About the same level as the non-Austronesian Melanesians have. And they don't have the EDAR variation either. Be consistent.
"If you would have read what I said (with informative links) previously about the autosomal genetics of Island Melanesians, you'd knew by now".
I have read a lot more on the subject than just your links and all consistently claim a genetic distinction between Austronesian-speaking Melanesians and non-Austronesian-speaking Melanesians. Of course the genetic boundary is actually in the form of a cline as genes have seeped each way across but the distinction holds. So you cannot compare the level of either Mongoloid or EDAR between non-Austronesina and Austronesian Melanesians.
Correction. You cannot claim the two Melanesian populations are the same. And:
ReplyDelete"Everything eventually changes but, when it does, it is for some reason. So if YOU are claiming such a change, which is very extreme, up to the point that the two genetic pools don't look the same AT ALL, you first need to offer a very good explanation".
Are you not aware that there has been considerable population movement in China over the last few thousand years? otably the Han expansion, but others as well.
You're trying to be overly complicated for things that are fairly simple: the SNPs tested by Liu are not just meaningful for all the paper's data (and that's why it's best to compare within it first) but also fit well with your mention of 16% some other O3 clade whose name I can't recall right now.
ReplyDelete"Much less than 20% is my understanding".
The figure I manage is c. 15%, 10-20% depending on specific populations and methods.
Certain recent study (that I contested) claimed as much as 20-40% East Asian/native American admixture in most European populations anyhow.
"You cannot claim the two Melanesian populations are the same".
They tend to use the same samples once and again but of course I am not sure, nor are you of the opposite. You're just trying anyhow to make complicated (without any clear foundation) something that is very simple in the outline: both Melanesians and Finnics have notable but >20% East Asian admixture but ones display the EDAR variant in reasonable amounts and the others do not, what is puzzling.
You just keep fighting and fighting without foundation nor purpose but the fact is that one.
"The figure I manage is c. 15%, 10-20% depending on specific populations and methods [presumably regarding East asian admixtuer in the Vologda]".
ReplyDeleteBut this is what you wrote in the comments section here some time back:
"It is important enough (many different ADMIXTURE/STRUCTURE studies show it) for the nearly fixated 'red allele' to show up at some 10% levels or at least 5% among Northern Russians".
That is much less than the 20% you are now claiming.
"both Melanesians and Finnics have notable but >20% East Asian admixture"
No. Finnics and non-Austronesian Melanesians have around 10% East Asian admixture. Austronesian Melanesians have around twice as much.
"They tend to use the same samples once and again but of course I am not sure, nor are you of the opposite".
ALL the articles and books I have read on the subject claim a genetic difference between Austronesian and non-Austronesian Melanesians.
"but ones display the EDAR variant in reasonable amounts and the others do not, what is puzzling".
Only 'puzzling' because you are determined to make it so. I think you would agree that we have three points on the periphery of the East Asian EDAR variant's geographic spread, where the variant, the haplogroups and the East Asian genetic influence dwindles away to virtually zero (or four points if we include America). They are: the Munda, the Austronesian Melanesians and Northern Russia. In each region the end result of the mixing of genes and haplogroups is different. The Munda have preserved both the East Asian Y-DNA and the EDAR gene but seem to have largely lost any particular East Asian genetic similarity. The situation in island Melanesia is fairly straightforward. Y-DNA O has almost died out, to be replaced by C2, but mt-DNA B remains prominent. The East Asian genetic similarity has been noticed for some time but a Papuan influence is more prominent in Melanesia. Yet apparently the EDAR mutation survives as a minor presence. In Northern Russia the EDAR gene has failed to survive but East Asian Y-DNA, some mt-DNA, and a substantial East Asian genetic influence survives.
Why would we expect all regions to behave in exactly the same manner?
I can't recall the context of that quote that you bothered searching for but I guess I was younger and naiver back then and I was discussing with a (racism-motivated) admixture-denialist. Whatever the case the figure is c. 15% for Vologda Russians, with minor individual variation. Let's see:
DeleteHere Russians (from Vologda) show exactly 15% (1.5 cm in a 10 cm tall graph in my screen) of East Asian and Native American admixture (50-50). The 75% of 15% is more than 10% but we see 0% of the EDAR thick hair & small breasts variant, when in Eastern Asia and among Native Americans it is at 75-100% frequencies.
It needs an explanation!
Here, the very same sample show ~1/7 of the length of the graph (a bit more in fact) in all K levels, measure it with a ruler, and 1/7 is ~14%.
So it is a very clear case.
"Finnics and non-Austronesian Melanesians have around 10% East Asian admixture. Austronesian Melanesians have around twice as much".
Neither what you say is true (not to any relevant level of exactness) nor it would matter if it would because 10%-20% of EA/NA admixture should show 7-20% of the EDAR variant, and in the case of North Russians it is apparently zero. And that demands an explanation!
According to Wikipedia, the age of haplogroup NO is 34.600±4.700 years BP and the estimated age of EDAR seems to be 30000 years. So, the lack or paucity of EDAR in Siberia could be explained by the more Central-Asian origin of haplogroup NO. If N arose outside China or in Western parts of China at a time when this EDAR gene was not yet around, it is natural that the gene is absent or almost absent in Russians and Finno-Ugric peoples of Northern Eurasia. The interesting thing is that, according to Wikipedia, the age of haplogroup O is 28,000-41,000 years BP, although haplogroup NO must be older than haplogroup O. The gap in these age estimates could be due to the huge difference in population sizes. What if this EDAR gene arose in Central China at a time when the area was dominated by other YDNA haplogroups than O type haplogroups and was passed on to East Asians maternally. This would also explain why it is present in Native Americans who are paternally haplogoup Q.
ReplyDeleteWell, there is at least one key study that estimates the age of O3 in that time-frame (c. 30 Ka ago), so it is most hard for me to believe that NO is so young. Instead I would estimate 50 Ka or more.
DeleteIn any case molecular clock as it has been used so far is not evidence of anything, just an educated guess.
Another thing, maybe more important, Native Americans lack Y-DNA NO altogether, the only specifically East Asian patrilineage they have is C3 and is very secondary to the, originally West/Central/South Eurasian, lineage Q. This one has a branch in NE Asia that is ancestral or directly related to the Native American subclades but it is still not NO. So associating these traits with any particular lineage is going to fail, same for matrilineages. The ancestral population where this allele originated must have included at some point several patri- and matrilineages: at least Y-DNA NO (or its subclades), C3 and D (still very strong among Japanese and Tibetans but not affecting the frequencies of this EDAR allele). On the mtDNA side there are also many different important lineages: M8 (incl. CZ), D, F, B, A, M7, M9, G, etc. which should have been involved since very early.
DeleteSo IMO this points to an ancestral population with diverse lineages in SE Asia (incl. South China), maybe 70 or 60 Ka ago, where this allele would have been gradually established (by drift or selection) and which would have expanded with them all, as they migrated northwards. The fact that Cambodians have relatively low levels of the allele, suggests that they, as well as other peoples from Indochina, etc. would have retained greater diversity, as corresponds to founder populations that have not gone through the founder effects (migrant bottlenecks) of the migrations northwards.
My opinion anyhow.
When you say that in the Americas C3 is very secondary to the lineage Q, it is all the more intriguing that EDAR is present in America but not in Western Siberia. According to Wikipedia, the age of MNOPS node is 35,000-45,000 years BP, and O and N must be younger than that. I would not be surprised if the age of macro-haplogroup O tends to be overestimated and that of N underestimated due to the huge population size of the former group and much smaller numbers of the latter. There is an interesting map on the distribution of NO and different N clades (http://www.nature.com/ejhg/journal/v15/n2/fig_tab/5201748f2.html#figure-title). My guess is that N and NO did not develop in the area where this EDAR gene arose and started expanding. Anyhow, that would be a logical explanation to your question why Russians and Finno-Ugric peoples do not have this gene. In North-Eastern Russia and Finland you have really a tiny amount of Eastern MtDNA present (By the way, my own haplo is I, my husbands's haplo H and my father's haplo U4 and we are all entirely Finnish).
ReplyDeleteIt is fascinating how diverse East-Asia is! Maternally both M and N are present and paternally you have haplogroups C, D and F (basal F and all MNOPS clades).
I know how difficult it is to link any macro-haplogroup with particular physical looks. MNOPS clades cover almost the whole world, M and S in Indonesia and Melanesia, N in North Eurasia, O in East Asia, P in South Asia, Eurasia, Middle East and the Americas. Also macro-haplogroup DE covers an astonishing variety of ethnic identities on a huge area, African bantus, Afro-Asiatic speakers of Middle East and Africa, Jarawa people and Tibetan as well as Japanese and Ainu. In spite of the deep ancestry, people usually end up looking like their neighbours!
In general age estimates are just that: very rough estimates that have zero value as data or evidence. IMO they are generally too recent and that is partly because of systematic underestimation of the chimpanzee-human divergence date, often cited as 7, 6 and even 5 Ka, when it actually is c. 8-13 Ka, according to several recent studies (Landgraeber for example). With that alone, and assuming all other factors are correctly evaluated, all age estimates should be almost double (minimal error is 15% older, max. 130% older). But there are also other factors that introduce errors. It'd be too cumbersome to discuss here so suffice to say that the molecular clock products so far are pretty much useless, they can be a compass but hardly anything like a reliable map.
Delete"it is all the more intriguing that EDAR is present in America but not in Western Siberia".
We do not know, at least I do not, if EDAR370A is present in West Siberia or how much. (EDAR is the gene, so we all have it, EDAR370A is the variant allele specific of East Asians and Native Americans).
"In North-Eastern Russia and Finland you have really a tiny amount of Eastern MtDNA present"...
But as discussed above, the autosomal admixture signal is of c. 15% in Vologda Russians specifically, maybe a bit larger among Finns from Finland. I don't know too much about the rest but it's easy to imagine considering that it is associated with levels of almost 50% Y-DNA N in many of those populations. You are right in that Oriental mtDNA is nowadays relatively rare but ancient DNA shows that it was more common in the past. The general understanding is that these peoples gradually absorbed other European gene pools but much more intensely by the side of women, retaining identity and language via patrilocality, what is still very visible in the Y-DNA.
"It is fascinating how diverse East-Asia is!"
True. IMO, after Africa, second only to South Asia and not by much. This surely reflects that when the early Eurasians (or Asians if you wish) expanded (probably from South Asia), they quickly colonized also East Asia (and then most of Australasia as well). To my eyes at least mtDNA N and Y-DNA C and D should have been present in very low frequencies and only succeeded in some locations. Another possible explanation might be that Toba supervolcano (c. 74 Ka ago) could have pruned some less common lineages in South Asia, which was the most affected by the ash fall. We know from Petraglia 2007 that people with African-like stone tech, MSA, lived in parts of India before and after the ash rain, so they were already there and at least some survived. But that means (to my eyes at least) that they could also have reached SE Asia, a region poorly understood archaeologically so far, at least for those time depths.
"I know how difficult it is to link any macro-haplogroup with particular physical looks".
I think that the problem is certain "pop" misunderstanding of haplogroups as always equating one single prehistorical population in a 1:1 relation. I think that is very unlikely in most cases, unless you restrict the concept population to that of extended family or clan, what is very extreme. Conversely I understand that, in most cases, populations must have included diverse lineages. Exceptions would be in very low density areas, where genetic drift would be very strong, or areas reached first by a very small founder population (founder effect).
Correction: I did not mean North-Eastern Russia and Finland but North-Western Russia and Finland!
ReplyDelete"According to Wikipedia, the age of MNOPS node is 35,000-45,000 years BP, and O and N must be younger than that"
ReplyDeleteI know Maju is not a fan of 'molecular-clockology' (and neither am I generally) but interestingly Dienekes posted some time ago an age estimate for various basal Y-DNAs:
http://dienekes.blogspot.co.nz/2012/08/dates-of-major-clades-of-y-chromosome.html
He has NO at 33,000 years and the O diversification into the three haplogroups at a little less than 30,000 years 27-28,000). That is stunningly close to the time the EDAR variant first appeared in the modern human species either by mutation or introgression. He doesn't have an age for MNOPS but the Wiki age fits.
"My guess is that N and NO did not develop in the area where this EDAR gene arose and started expanding".
Perhaps not where NO originated, but I think it is extremely likely that the EDAR gene arose where NO diversified. Unlike Maju I see no problem with the EDAR307A's absence in northern Russia.
"but we see 0% of the EDAR thick hair & small breasts variant, when in Eastern Asia and among Native Americans it is at 75-100% frequencies".
As I said, 'Why would we expect all regions [on the margins of a gene's expansion] to behave in exactly the same manner?' I can't understand why you insist that the proportion of each gene should exactly reflect the overall proportion of the whole genetic contribution. Especially when we're talking here of just small proportions. Anything can happen in such situations.
"because 10%-20% of EA/NA admixture should show 7-20% of the EDAR variant"
Why would that automatically be so? In fact it is extremely unlikely the proportions would be exactly the same, or even necessarily close when we're talking about such low levels in the first place.
"Why would that automatically be so?"
DeleteThe actual question is why would it not be so "automatically"?
"What if this EDAR gene arose in Central China at a time when the area was dominated by other YDNA haplogroups than O type haplogroups"
ReplyDeleteThat would require an even greater displacement of haplogroups than the small level I am suggesting. Maju won't have a bar of it. And I have trouble accepting the idea.
"This would also explain why it is present in Native Americans who are paternally haplogoup Q".
I agree that the EDAR variant was carried to America mainly by mt-DNAs. The information I have seen suggests that Y-DNA Q picked up the East Asian mt-DNA haplogroups as they passed to the north of the region Y-DNA O and those mt-DNA haplogroups already occupied. That explanation certainly makes sense to me.
"Native Americans lack Y-DNA NO altogether, the only specifically East Asian patrilineage they have is C3 and is very secondary to the, originally West/Central/South Eurasian, lineage Q".
This is one thing (actually one of many) that Maju and I agree on.
"The ancestral population where this allele originated must have included at some point several patri- and matrilineages: at least Y-DNA NO (or its subclades), C3 and D (still very strong among Japanese and Tibetans but not affecting the frequencies of this EDAR allele)".
We agree on that too, but perhaps I would exclude D from the original grouping. D is present on the Andamans where there is almost no East Asian admixture as far as I'm aware.
"On the mtDNA side there are also many different important lineages: M8 (incl. CZ), D, F, B, A, M7, M9, G, etc. which should have been involved since very early".
Once more we are largely in agreement although I would definitely exclude B and F from the core region. And add N9.
"So IMO this points to an ancestral population with diverse lineages in SE Asia (incl. South China), maybe 70 or 60 Ka ago, where this allele would have been gradually established (by drift or selection) and which would have expanded with them all, as they migrated northwards".
That's where we completely disagree. The EDAR variant almost certainly arose in northern China/Inner Mongolia according to this paper.
"The fact that Cambodians have relatively low levels of the allele, suggests that they, as well as other peoples from Indochina, etc. would have retained greater diversity"
Again I think otherwise. To me it is yet another indication that the gene didn't originate anywhere near there.
"My opinion anyhow".
Thank you.
"That's where we completely disagree".
DeleteAnd in almost everything else, I know. I'm tired today.
Maju, you say that “autosomal admixture signal is of c. 15% in Vologda Russians specifically, maybe a bit larger among Finns from Finland”, but I would bet that it is the opposite. The article on ancient MtDNA: http://www.plosgenetics.org/article/info%3Adoi%2F10.1371%2Fjournal.pgen.1003296 is very interesting and illuminating in this respect. I suppose that you refer to it in your comment. There you can see people moving along different routes, via Volga Ural (aUzPo, Apo) and the northernmost route (aBOO). In figure 3A you can see that there is similarity between aUzPo (with a lot of haplogroup C) and Central Asian people, excluding China. In figure 3B you can see a strong connection between aBOO (with a lot of C, D and Z) and North East Asia. However, it is interesting that Sami people who live today in the same area are not direct descendants of these aBOO people, although haplogroups D and Z have been absorbed specifically to Finnish Sami MtDNA pools. When you say that there are levels of almost 50% Y-DNA N, I can tell you that it is even more, 78% in North Savo (in Eastern Finland) and 60% in South Western Finland (Länsi-Suomi). It would be weird if all MtDNA originally associated with YDNA N virtually disappeared in such a large area as Finland, the Baltic area and North Western Russia. I still think that it is plausible that the archaic N people and their women who eventually reached Finland did not possess this gene. In Finland many people are desperate about their thin hair - I am one of them - and in the whole world the frequency of thin silky hair must be the highest just in Finland.
ReplyDeleteTerry, your justification for exclusion of D from original EDAR370A carriers might be irrelevant, as this gene probably did not yet exist when Andamanese D was separated from the rest of East Asian D http://www.sciencemag.org/content/311/5760/470.2.full. As for your comment on the origin of NO, I would agree with you if you said that it is extremely likely that the EDAR370A gene arose where O diversified! When you want to exclude B and F from the core region, I am wondering what you think were the Y haplos that were expanding with these MtDNA clades.
What we physically look like at the moment is probably a result of quite late local developments, in which a host of different haplos have been involved. I think that 50,000 years back we all looked pretty archaic!
In the comment just before you, Kristiina, began posting (and I was hoping that you had read it but seems not), I linked to two different ADMIXTURE panels which evidence that the Vologda Russians (just "Russians" by label) have 15% admixture from Siberia (or East Asia+Native Americans). The panels are: here and here. Use a ruler to measure the percentages (I did myself). An mtDNA comparison will not help you with that as clearly as a direct autosomal one (probably because in this case it is more strongly related to Y-DNA for reasons of apparent prehistorical patrilocality).
Delete"It would be weird if all MtDNA originally associated with YDNA N virtually disappeared in such a large area as Finland, the Baltic area and North Western Russia".
It did not completely disappear but was clearly reduced a lot. The exact mechanisms I can't say but I it seems evident that, in the long run, they outmarried with their southern neighbors once and again within a patrilocal scheme. It's also possible that the more "europeanized" populations from the southern fringes of the Finnic (or Fino-Ugric?) area had some sort of advantage (being better connected) and also they may have been pressed in northward direction by their southern neighbors in a chain effect along time. Speculating a bit here but I hope that you get my point.
"Terry, your justification for exclusion of D from original EDAR370A carriers might be irrelevant, as this gene probably did not yet exist when Andamanese D was separated from the rest of East Asian D"
ReplyDeleteI very much doubt that the Andamans were inhabited as long as 30,000 years ago. From your link:
"The Andamanese sequences of these two haplogroups shared their most recent common ancestor (MRCA) within these haplogroups at time depths 3000 years [confidence interval (CI) 2] and 12,000 years (CI 4), respectively"
But the authors do say, 'the MRCA with continental Indian mtDNA lineages nested within haplogroup M at a time depth of 65,000 years'. But we have no idea if that time has anything to do with the arrival on the Andamans.
"I would agree with you if you said that it is extremely likely that the EDAR370A gene arose where O diversified!"
I'm fairly sure that is in fact what happened.
"When you want to exclude B and F from the core region, I am wondering what you think were the Y haplos that were expanding with these MtDNA clades".
Possibly NO itself along with other K and F clades.
"The actual question is why would it not be so 'automatically'?"
Because migrations do not occur instantaneously. People mix with others along the way diluting and redistributing genes. The proportion of particular genes would alter along the route and we're dealing here with the margins of expansion. Once more you are demonstrating your Garden of Eden obsession. But you know that, because you said:
"I think that the problem is certain 'pop' misunderstanding of haplogroups as always equating one single prehistorical population in a 1:1 relation".
Yet in the matter of EDAR370A in northern Russia you ignore your own comment.
"When you want to exclude B and F from the core region, I am wondering what you think were the Y haplos that were expanding with these MtDNA clades".
ReplyDeleteI thought I'd better clarify what I meant above. Of course in the early days only one mt-DNA B clade made it as far north as where the EDAR variant appeard. That was the B4b branch of B4b'd'e'j. Other B haplogroups look to have stayed in southern China or SE Asia. The same is basically true for F. Just one F clade expanded widely: F1. And only a small proportion of that haplogroup. I'm not sure what branch(es) of F1 are found in such regions as Tibet, North China, Korea and Japan.
Maju, I agree with you when you say that ”two different ADMIXTURE panels which evidence that the Vologda Russians (just "Russians" by label) have 15% admixture from Siberia (or East Asia+Native Americans)”, I just meant that Finns might have less Siberian admixture than Vologda Russians but, actually, this admixture might be the same, little less or more, but it does not change the big picture.
ReplyDeleteTerry, I love you when you say that you’re ”fairly sure that is in fact what happened”. :-) You are also right when you say that Y haplo D could be excluded from the original grouping. It is true that Japan and Tibet are not in the core area where this gene arose according to the map. You were interested in the percentage of N haplo in Volgoda Russians. They have 35.5% of N3 and 3.3% of N2 and the percentage of R1a is 33.1% (http://www.sciencedirect.com/science/article/pii/S0002929707000250).
After a lot of rethinking, I propose that NO arises somewhere in Yunnan from MNOPS node (http://www.nature.com/ejhg/journal/v15/n2/fig_tab/5201748f2.html#figure-title). O arises somewhere between Yunnan and Three Gorges area and starts expanding. OMSY2.2 and OP31 may arise somewhere near this area. OM122 may also arise in Three Gorges area expanding in the first place in the North. NO, N* and different O subclades expand at least with MtDNA N (R9, R11) that will develop into F and B respectively and with MtDNA M9 (?). These O clades or part of them may be present in the area of origin of EDAR307A gene when this gene arises and starts expanding. You give the time frame of 27-28,000 years for the diversification of this hg.
N arises later on in Yunnan (or North of it), and a part of these N men are heading North (There are also important frequencies of N* in Indo-China). North is much less crowded and this is why this haplo becomes an important player in the North and Siberia. According to this recent Chinese research (http://eurogenes.blogspot.fi/2013/01/lots-of-ancient-y-dna-from-china.html#comment-form), there were a lot of N haplo men, as well as Q and R haplo men, in Xinjiang and northernmost part of China 3000 – 5000 YBP. However, it seems to me that they are moving on the outer fringe (or outside) of the core area of EDAR307A gene. I propose that N haplo men heading West did not get this EDAR variant and this leads to the division between more eastern looking Uralic people and Western looking Uralic people.
To get an insight into which Mtdna haplos might have been around where N expands in the North, this research is illuminating: arose http://dienekes.blogspot.fi/2012/01/aapa-2012-abstracts-part-1.html. I would pick up C and Z (M8), D, U2 and U4 as possible candidates. These haplos are still present in Uralic speaking peoples of Eurasia and figure in this recent research I referred to in my previous mail (http://www.plosgenetics.org/article/info%3Adoi%2F10.1371%2Fjournal.pgen.1003296). It is interesting that they found C1 haplo in Northwestern Russia! It is rare in Asia and present in America. As for this process of diminution of Eastern MtDNA in Europe, it seems that these N men took H and U5a women when they came to Europe. In this recent research they confirmed the presence of MtDNA H and U5a in northwestern Russia 7,500 (uncal.) years ago.
I do not know if you have noticed that Davidski has also recently commented this saying ”I think this ancient admixture is something Central Siberian or Central Asian, and it's shared by Europeans, Amerindians and East Asians. It might well be the same thing as what the proto-Uralics were before they mixed with Europeans and East Asian EDAR carrying eastern Siberians.” (http://eurogenes.blogspot.fi/2013/02/ancient-amerindian-like-admixture-in.html)
For Finns specifically, you may want to check Bauchet 2007, specifically fig. 1, where it seems obvious that they have some notable Siberian admixture(represented by the light orange component (Altaians themselves are not purely Oriental, so they also display European components) again in the order of 1/7 or 15%.
ReplyDeleteAnyhow, I see no reason why Finnish would be less Oriental than russified populations. Again arguing on the haploid genetics is not as informative as the autosomal data itself.
This is not the same elusive and contexted Oriental admixture at levels of 20-40% claimed by a recent paper, on very dubious grounds, for all Europeans or at least most. But a much more clear admixture specific of Finnic and related peoples of NE Europe.
As for the origins of N, Kristiina, they are probably somewhere in Central or Southern China (highest diversity), although the highest frequencies are further North in Buryat (a Mongol ethnicity of Siberian) and then in the truly Northern fringes of Siberia.
ReplyDeleteThis paper only gives data on EDAR370A in Eastern Siberia, so indeed that the allele is more diluted to the West of Siberia is a serious possibility and an alternative option to that of "racist selection" in European peoples of minor Siberian ancestry.
"It is interesting that they found [mtDNA] C1 haplo in Northwestern Russia! It is rare in Asia and present in America".
That may have to do with the fact that both share the Siberian connection rather than a strictly East Asian one, in different ways but Siberian anyhow.
"I do not know if you have noticed that Davidski has also recently commented this saying ”I think this ancient admixture is something Central Siberian or Central Asian, and it's shared by Europeans, Amerindians and East Asians. It might well be the same thing as what the proto-Uralics were before they mixed with Europeans and East Asian EDAR carrying eastern Siberians.”"
Naturally. I am subscribed to that comment section. I agree that it is an alternative possibility but only a wider survey of the relevant peoples re. EDAR will give us a proper answer.
"As for the origins of N, Kristiina, they are probably somewhere in Central or Southern China (highest diversity), although the highest frequencies are further North in Buryat (a Mongol ethnicity of Siberian) and then in the truly Northern fringes of Siberia".
ReplyDeleteDiversity is as likely, or even more likely in many cases, to ba a product of multiple enry of a haplogroup rather than being an accurate indicator of origin. I am not claiming that to be the case here but I still think you should consider the possibiliy. As I understand the situation N is quite uncommon in China and where present most prominent in the northwest. And the clades found there are actually widespread clades, not basal ones.
"You were interested in the percentage of N haplo in Volgoda Russians. They have 35.5% of N3 and 3.3% of N2"
Thanks for that.
"I propose that NO arises somewhere in Yunnan from MNOPS node"
Yes. I think it very likely that MNOPS's geographic range extended that far north.
"O arises somewhere between Yunnan and Three Gorges area"
I certainly see O as having originated further north than that. Certainly well within the region where the EDAR370A variant developed and reached fixation.
"OMSY2.2 and OP31 may arise somewhere near this area [Three Gorges]".
That holds reasonably well for O1-MSY2.2, but one clade of O2-P31 is definitely a northern haplogroup. It is very difficult to make a case for O2b having originated anywhere near the Three Gorges region.
"OM122 may also arise in Three Gorges area expanding in the first place in the North"
I can't see a northward movement for O-M122. That region would have early become occupied by N and O2b, and probably C3. That places O3's origin either north or west of the Three Gorges.
"These O clades or part of them may be present in the area of origin of EDAR307A gene when this gene arises and starts expanding. You give the time frame of 27-28,000 years for the diversification of this hg".
I agree that the O clade was present in the region when the gene arose but probably not in an already diversified condition. Its expansion carried the gene south while other haplogroups carried it north and into America.
"N arises later on in Yunnan (or North of it), and a part of these N men are heading North (There are also important frequencies of N* in Indo-China)".
I am very sure that N is actually just the northern version of NO while O is the southern branch.
"there were a lot of N haplo men, as well as Q and R haplo men, in Xinjiang and northernmost part of China 3000 – 5000 YBP. However, it seems to me that they are moving on the outer fringe (or outside) of the core area of EDAR307A gene".
On the northern fringe, yes. Have you seen Razib's second map? Maju hasn't posted it for some reason or other:
http://blogs.discovermagazine.com/gnxp/2013/02/is-girls-generation-the-outcome-of-the-pleistocene-mind/
"I propose that N haplo men heading West did not get this EDAR variant and this leads to the division between more eastern looking Uralic people and Western looking Uralic people".
Possible. But it is certainly not necessary to postualte any such thing. The prortion of various genes in a population would change as it expanded and mixed with locals it met along the way. As you say:
"it seems that these N men took H and U5a women when they came to Europe".
Maju, according to this Bauchet 2007 research, Finns have more Asian admixture than Poles and Altai people considerably more than Finns. I think that this level of admixture is well in line with the scenario I proposed above, although Russians do not figure in this study. I did not say that russified populations have more Asian admixture than Finns. I just meant that Volgoda Russians, who have a strong Finno-Ugric substrate, might have more Asian admixture because Finland is further West than Volgoda.
ReplyDeleteI do not know how you get at that bigger southern diversity for HG N. Here http://www.nature.com/ejhg/journal/v15/n2/fig_tab/5201748f3.html#figure-title the Chinese samples are near the root of N*, but there is no geographic indication. In the research they say about N3 that “although the frequency of hg N3 is low in northern China and restricted to a few small populations, its STR variance is higher (0.26, averaged across eight loci: DYS19; DYS389I&II; DYS391; DYS392; DYS393 and DYS439 ...) than in Altai and in Volga-Ural region (0.16 and 0.17, respectively), thus again pointing to northern China rather than southern Siberia as a possible place of expansion of hg N3.” But in fact they continue that “One may notice that while STR variation is relatively low in the Volga-Ural group, some north-European populations have high STR variance (eg, 0.32 in Finns: data from, 36 without DYS385ab). The high STR variation among the latter, however, might not be a result of a long-term in situ differentiation of the founder lineage, but, rather a consequence of an admixture of separate N3 founder types.” I think that Terry referred to this explanation in his comment. The researchers however postulate a possible southern origin when they say “the phylogeography of the NO* and N* lineages ... and the presence of N* chromosomes in southern East Asia (South China and Cambodia ...) suggests that this region could be the source of the initial spread of hg N. In this scenario, the Altay/Sayan/southern Siberia region might have been a place of transition of hg N westward as all major subclades of hg N are still to be found there. “
When you say that the highest frequencies are further North in Buryat, I don’t quite understand it. Here you have an impressive list of frequencies (http://www.docstoc.com/docs/130265674/Research-Institute-of-Medical-Genetics-at-the-Tomsk-Scientific-Center) and the Buryat N3 frequencies are about 30% and 50% and they seem to lack N* and NO* (with only a small amount of N2). Yakuts have the highest frequencies of N3, 81,9% and 94,3% respectively.
Terry, I am much less acquainted with O clades and Chinese geography, so I do not insist. You are right that often it is the basal type that expands first and the diversified clades pop up in different parts of the area concerned. By the way, do you think that O-M122 arose very near the EDAR370A hot spot? Yes, I have seen Razib's second map and I would be interested to know how it is obtained. I proposed in my mail that MtDNA R11 or B is connected to the expansion of O clades in East Asia. Now I am however wondering if B is as old as Wikipedia says, 50 000 years, and NO arises 33,000 years ago, with whom did it arrive in East Asia?
Do you happen to know if it is possible that this EDAR gene disappears in a population where both sexes initially have this gene but later on men marry more and more women who do not carry this gene.
Re. Admixture, Altaians are probably not the best proxy because they are themselves admixed. But if you are so interested just google "Finn admixture East Asia" and you will see a large list of Admixture runs, pro and amateur. With a little patience you can even run your own using Amixture on the 1000 Genomes Project data, as described by Razib here.
DeleteAs for N I won't contest anything of all you have clearly documented very well. I was thinking, wrongly, that N was more common in Buryats than that but still 30% is a quite notable frequency, suggesting a route for the lineage via that region. Otherwise N's great success was clearly in Siberia and even in the North of that region, strongly suggesting that the success of N (mostly N3 and N2) was linked to adaption to the most extreme climates of the Far North. To some extent at least it is also related to Uralic languages. Further South they are instead a small scattered lineage, even if surely more basally diverse.
As for the origin of N, I'd accept "somewhere in or near China" as a perfectly valid answer. The problem of relying on N* is that it can be one single haplogroup ("N4") or many (something like "N4, N5, N6, N7 and N8). Unless that small but potentially diverse fraction is clarified, some uncertainty remains. The presence of NO* is also interesting, although again it is scattered between South China and Mongolia (and even some in Japan and Malaysia, although there's no N over there).
Also notice that I strongly disagree with Terry about the origins of O, which for me are between Indochina and South China, i.e. in the South.
Delete"Do you happen to know if it is possible that this EDAR gene disappears in a population where both sexes initially have this gene but later on men marry more and more women who do not carry this gene".
ReplyDeleteI'd guess it was unlikley unless there was selection against the gene. But I think in the case of northern Russia we are dealing with a situation where a minority of mt-DNA carried the EDAR370A variant and it was drifted out. And even the Y-DNA N men who arrived may have bred with non-EDAR carrying populations along their route.
"thus again pointing to northern China rather than southern Siberia as a possible place of expansion of hg N3".
Following from our discussion re. Y-DNA N I have begun looking at it, at last. I notice that in ISOGG 'N3' from the paper is now 'N1c1'. It is the main N haplogroup that expanded, with N1c2-L666 a long way behind. An origin in North China makes sense.
"The high STR variation among the latter, however, might not be a result of a long-term in situ differentiation of the founder lineage, but, rather a consequence of an admixture of separate N3 founder types".
Most likely through the arrival of large numbers as an already-diverse scattering of haplogroups. That is very often the cause of apparent diversity. I am reasonably sure that individual haplogroups arise in separate regions, often on the advancing front of a haplogroup's distribution.
"The researchers however postulate a possible southern origin when they say 'the phylogeography of the NO* and N* lineages ... and the presence of N* chromosomes in southern East Asia (South China and Cambodia ...) suggests that this region could be the source of the initial spread of hg N'".
As I said, I can accept NO being from there but N looks much more likely to be somewhat north.
"the Buryat N3 frequencies are about 30% and 50% and they seem to lack N* and NO* (with only a small amount of N2)".
In ISOGG 'N2' is now 'N1c2b'. Somewhere I found the information that the haplogroup forms two clusters: one in Ural/Volga and the other one further east.
"do you think that O-M122 arose very near the EDAR370A hot spot?"
I think it extremely likely. And the same holds for the whole O-M175 haplogroup although I could accept that O-M119 and O-P31 arose on the southern boundary of the hotspot if it can be shown to be so. I also think N arose very near the hotspot. The EDAR370A gene is very much associated with N in eastern and northeastern Siberia. And C3 may be involved.
"Now I am however wondering if B is as old as Wikipedia says, 50 000 years"
It could very well be that old, but I'm very sure it is a south China, or even SE Asia, haplogroup originally. As is mt-DNA F.
"NO arises 33,000 years ago, with whom did it arrive in East Asia?"
At least some clades of mt-DNA D and CZ. And possibly the northern clade of B. You may be interested in this link where Maju and I have posted our versions of where the mt-DNA haplogroups arose. We agree on the region each haplogroup arose although we disagree as to the explanation for the distribution. Here is mt-DNA R:
http://ourorigins.wikia.com/wiki/Mt_R_east_to_west
I don't know for sure if they carry EDAR370A or not right now but the native peoples of West Siberia and even the Nenets of far NE Russia look very much East Asian in appearance and autosomal DNA, so my default guess is that they do carry it. These include of course many Uralic-speaking peoples.
DeleteOf course data (but not speculation alone) can show my "default guess" to be wrong. But I'll wait for it to come and you should do too.
"I don't know for sure if they carry EDAR370A or not right now but the native peoples of West Siberia and even the Nenets of far NE Russia look very much East Asian in appearance and autosomal DNA, so my default guess is that they do carry it".
ReplyDeleteThe map above certainly shows a strong presence of the gene in northeastern Siberia. My guess is the same as yours: I strongly suspect that the gene is found in such peopel as the Nenets.
"Of course data (but not speculation alone) can show my 'default guess' to be wrong. But I'll wait for it to come and you should do too".
You're imagining things again. Until now I have made no comment as to the presence of the gene in northern or western Siberia. Only in relation to its apparent absence in the northern Russian sample in the diagram. So I will wait for the data on the region.
The nenets haplos are: 40% N1c and 57% N1c2b, on maternal side 8.4% H, 25% U, 12.6 % C2, 20.8% C3, 8.3% D, 16.7% D3, 4.2% D4, 4.2% Z, and I agree that Edar variant should be found there. I also think that this gene is obviously present in N-haplo men of Northeast Asia. As it is very difficult to me to see any reason for racist selection against this gene, I would say that if in later studies they show that Finno-Ugric people in the West carry that gene at frequencies of more than 20%, I agree that N1c may have expanded to Finland from Northeast China. If they are less than 10%, then the ancestors of the majority of Western Uralics turned West in West China without mixing with Edar variant-carrying people of the Northeast. If the frequency is between 10 and 20%, it is open to debate. These figures are of course indicative.
ReplyDeleteI checked that the first people came to Japan 30,000 years ago and the Jomon period started 14,000 BCE. On Asiafinest.com someone proposed that the first inhabitants were hg C1 and Jomon people hg D. I suggest that hg C1 was accompanied by Mtdna N9 and hg D by mtDNA M10 and D. They also proposed that N and NO came to Japan only 1 – 200 AD, but I think that it might have come earlier, with O3 and O2b1, 800 BCE. With this time frame in mind, the EDAR variant may have been present in Korea and China from the first colonization of this area. I do not know the route of hg D to Korea and Japan, but the Ancestors of Jomon could have passed through the area where EDAR370A arose or expanded.
I would say that Mtdna D (age 40,000 - 60,000 YBP in Wikipedia) is connected with Ydna D and Mtdna B (age 50,000 YBP in Wikipedia) with Ydna O or even K. IMO, Mtdna F (age 43,400 YBP) is a good match with hg O, but, of course, in the southern range of hg O. Mtdna M8, ancestor of C and Z, could have arisen and wandered North with hg N, although these clades, later on, almost completely disappeared in the West. In fact C is not any more found in the West, although it was detected in a 7,500-years old burial in northwestern Russia. I think that hg N men and U4 women met in Xinjiang and Altai.
It is fascinating to look at this map on the distribution of hg C: http://www.google.com/imgres?um=1&hl=fi&lr=&biw=1233&bih=680&tbm=isch&tbnid=ezFuZ5UZ_ut_pM:&imgrefurl=http://dienekes.blogspot.com/2010_05_01_archive.html&docid=KU5i4l1MUYZGZM&imgurl=http://3.bp.blogspot.com/_Ish7688voT0/S-aCMKf-JyI/AAAAAAAACYY/npKxBWTQC-U/s1600/hapC.jpg&w=757&h=468&ei=pAQ7UZCNEuyL4gT0j4DoBw&zoom=1&iact=hc&vpx=258&vpy=386&dur=260&hovh=176&hovw=286&tx=136&ty=151&page=3&tbnh=132&tbnw=215&start=56&ndsp=30&ved=1t:429,r:58,s:0,i:255
I now wonder if I saw once someone proposing on a blog that hg C went to China along the Northern route skipping India. This route looks even likely, when you look at the map. Of course, there may have been Neanderthals on the way and the climate may have been quite harsh, but our species should have been very smart already at that early time.
When I postulated the origin of NO in Yunnan (based on the frequency map), I had in mind the origin of MNOPS somewhere in Burma, as I think that Hg K must have come to South East Asia through Burma. In my mind, hg P arose in Eastern India or Bangladesh and went up to Eurasian Plate in Pakistan and not in China or Tibet. Now, I suppose that you have in mind that MNOPS coalesced further South in Indochina or even further South. That, of course, influences the way you see how different clades arise and expand.
I just cannot concur, Kristiina: the apparent centroid of Y-DNA C is in SE Asia (let's not forget C2, C4, C5 and C*) and also, when the main Asian expansion happened, before 50 Ka ago, the only "northern route" plausible corridor, Altai, was occupied by Neanderthals and "Denisovans" with a technology (Mousterian) mostly unknown of in East Asia.
ReplyDelete"the apparent centroid of Y-DNA C is in SE Asia"
ReplyDeleteSorry Maju. You're carefully avoiding facing the facts again:
http://www.pnas.org/content/106/48/20174.full.pdfhttp://www.pnas.org/content/106/48/20174.full.pdf
And the supplementary material:
http://www.pnas.org/content/suppl/2009/11/16/0910803106.DCSupplemental/pnas.200914264SI.pdf
Coordinates of centroid for Y chromosome haplogroup C: 109.62 degrees longitude, 40.28 degrees latitude. That places its centroid well to the north of SE Asia.
"when the main Asian expansion happened, before 50 Ka ago, the only 'northern route' plausible corridor, Altai, was occupied by Neanderthals and 'Denisovans' with a technology (Mousterian) mostly unknown of in East Asia".
C was almost certainly not involved with the 'main Asian expansion' and we have no idea when it might have arrived in East Asia. And at 50,000 years ago we actually have no idea if Neanderthals and Denisovans and their Mousterian technology were long established in the Altai or were recent arrivals.
"With this time frame in mind, the EDAR variant may have been present in Korea and China from the first colonization of this area".
Although the map maju posted shows a strong presence of the gene in Japan the other map at Razib's shows Japan was not part of the early fixation of the gene. Later even than southern China. That would support the idea that the first people to japan were not of the 'Mongoloid phenotype', whatever their haplogroups.
"I now wonder if I saw once someone proposing on a blog that hg C went to China along the Northern route skipping India".
Possibly me.
"I had in mind the origin of MNOPS somewhere in Burma, as I think that Hg K must have come to South East Asia through Burma".
MNOPS's ancestors at least must have come through Burma. However MNOPS may have coalesced further south as both M and S are trans-Wallacean haplogroups as are several 'K' haplogroups, leaving just P and NO west of Wallace's Line. As you say, 'Now, I suppose that you have in mind that MNOPS coalesced further South in Indochina or even further South. That, of course, influences the way you see how different clades arise and expand'.
"hg P arose in Eastern India or Bangladesh and went up to Eurasian Plate in Pakistan and not in China or Tibet".
I agree P did not go through China or Tibet but as to its origin see above.
terryt wrote,
ReplyDelete"Coordinates of centroid for Y chromosome haplogroup C: 109.62 degrees longitude, 40.28 degrees latitude. That places its centroid well to the north of SE Asia."
The location indicated by those coordinates should be somewhere in Inner Mongolia near the border with Shaanxi, slightly northeast of Yinchuan, Ningxia Hui Autonomous Region. This is close to the ancient homeland of (at least one important population ancestral to) the Han Chinese (Hua-Xia), as well as the historical seat of the Tanguts (Xi-Xia "West Xia"), a population that spoke an archaic Tibeto-Burman (or Sino-Tibetan) language.
The modern Chinese-speaking Muslim inhabitants of this region are known to contain a great diversity of lineages within Y-DNA Haplogroup C-M130, including C3-M217, C5-M356, and C-M130(xC1-M8, C3-M217, C5-M356), so it does seem like a decent candidate for the region of prehistoric diversification of my putative Northern branch of C-M130, which includes East Eurasian-American C3-M217, Central-South-West Asian C5-M356, and Japanese C1-M8.
I think it is possible that the ancestors of my putative "Southern" branch of C-M130, which includes at least Wallacean-Oceanic Austronesian C2-M38 and Southeast-South Asian C*-M130, may have entered Southeast Asia by a different route than the ancestors of the Northern branch of C-M130 took to North-Central China. In other words, C-M130 may have split into at least two subclades prior to the migration into Greater Eastern Eurasia (from Southwest Asia or wherever the ancestors of C-M130 parted ways with their CF-P143 relatives).
As for C, I do not say that that was what happened. However, it is a fascinating possibility and it would make easier to understand the division between Mtdna N and M. If we stick to the conventional view and argue that ydna C and Mtdna N (nearly) disappeared from India because of the Toba catastrophe, the expansion of C must have started soon after OOA, i.e. 100,000 years ago. Recent findings of an old clade of ydna C in Italy are also exciting and would point at C not taking only the straightest road to India.
ReplyDeleteIt seems that the temperatures started plummeting 50,000 years ago. What if hg C went through Central Asia to China during the warm period before 50,000 and then, when the climate got colder, the Neanderthals came back. The climate was very warm between 90,000 and 70,000. http://en.wikipedia.org/wiki/File:Phanerozoic_Climate_Change.png
If N and NO took Mtdna M8, ancestor of C and Z, to Northeast Asia and mtdna C was then carried to America by hg Q, the dates must match. According to Wikipedia N is from 15,000 to 25,000 years BP, and in the light of the scenario above, it is preferably closer to the upper end, depending on the entrance of people in Beringia. This is, of course, speculative, as M8 could have been carried North earlier with another ydna O clade. However C and Z seem to be quite rare in northern China, Japan and Korea and frequent in Altai and Central Asia.
My insistence on this early arrival of EDAR variant to Japan is linked to my everyday experience of the typical hair of Japanese women. With a lot of Finnair’s Asian flights going on, we have plenty of Japanese (and more and more Chinese) in Helsinki, and I always look very envious at their beautiful thick hair which is so different from the average Finnish one. This is in fact why the title of this post and Maju’s comment caught my eye and I started this discussion. Anyway, the Paleolithic and Jomon ydna (C1 and D1/D2) cover 38.1% of the modern Japanese ydna pool and the typical Ryukuan and Ainu Mtdna haplogroups (A, N9, Y, M7, G, D4) 46.00% and 51.66% respectively (according to Tanaka et al.).
"What if hg C went through Central Asia to China during the warm period before 50,000"...
DeleteThat all the archaeological evidence for Central Asia cries "Neanderthals, Denisovans, Mousterian!" Only since c. 43 Ka BP (c. 47 Ka calBP) we can begin speaking of H. sapiens in Central Asia, what is in line with the rest of West Eurasian early UP colonization.
So the "what if" fails from the beginning.
Anyhow, as I say in my other comment, the real centroid of Y-DNA C is in SE Asia, not NE Asia at all.
As for your speculations about pairings of Y-DNA and mtDNA, I cannot say because I have never looked the matter in such depth for NE Asian mtDNA/Y-DNA possible correlations (and never ever used scholastic "age estimates" in any case as foundation for such speculations). I suspect that mtDNA M8 coalesced relatively early in the southern fringes of NE Asia, but some time after D and M7, and it should have an age (my rough estimate) similar to that of the earliest West/Central Eurasian colonization by our species (c. 55 Ka ago). MtDNA R and therefore Y-DNA MNOPS (NO) had not yet arrived to NE Asia (but would soon after) - always in my opinion (because estimating this stuff is not rocket science).
@Terry (and secondarily Ebizur):
ReplyDelete"Coordinates of centroid for Y chromosome haplogroup C: 109.62 degrees longitude, 40.28 degrees latitude. That places its centroid well to the north of SE Asia".
That is not a fact but a miscalculation by the authors, who only considered frequency and not basal diversity. When you begin at the level of subclades: C1, C2, C3, C4, C5... you end up with a number of subclade centroids: Japan for C1, Wallacea for C2, NE Asia for C3, Australia for C4 and whichever for C5 (I don't have enough data for this clade). The resulting trapezoid (not using C5) has a centroid East of Philippines, after correcting for the actual geography and the weight of C5, it ends up near Guangzhou or Tonkin.
Try doing it yourself, Terry. Parroting the errors of others is not appealing to "facts". Be critical, mercilessly rational, scientific.
I expect that time will tell what routes hg C took on its way to East. I think it is better to be open and not take a definite stand at this stage.
ReplyDeleteTerry, when I am talking about this gene, I am not talking about 'Mongoloid phenotype’. These two things are surely interconnected and mostly expanded together, but initially they could have had their origin in several different ethnic groups. As I understood it, this gene is only about hair thickness and breast size and not about the form of face or eyes or teeth. Some Finns have small and round eyes, a round face or a small nose, and these features we surely carried from the East, at least from Central Asia.
"That is not a fact but a miscalculation by the authors"
ReplyDeleteOf course. I should have known. You know far more about the subject than do the authors.
"Try doing it yourself, Terry. Parroting the errors of others is not appealing to 'facts'. Be critical, mercilessly rational, scientific".
Why don't you take your own advice? I have long noticed that you always begin with your own long-established pre-existing belief. And if any new evidence fails to fit that belief you go through an exact same process. First of all you concoct all sorts of hypothetical scenarios, based on no evidence at all, in an attempt to make the evidence fit that belief. If even after all the contortions involved you are still unable to construct an even slightly believeable scenario you simply dismiss the evidence entirely. That is what you call 'science'. You never consider the possibility that your pre-existing belief may be incorrect. Exactly as you have done here:
"That is not a fact but a miscalculation by the authors, who only considered frequency and not basal diversity".
I rest my case.
"That all the archaeological evidence for Central Asia cries 'Neanderthals, Denisovans, Mousterian!'"
It does not. No, or very few, human remains have been found anywhere in the region and 'Mousterian' does not automatically mean 'pre-sapiens', as I've consistently tried to point out.
"Only since c. 43 Ka BP (c. 47 Ka calBP) we can begin speaking of H. sapiens in Central Asia"
Wrong. You should have said, 'Only since c. 43 Ka BP (c. 47 Ka calBP) we can begin speaking of Upper Paleolithic in Central Asia'. I was sure that even you were becoming aware of the fact that the Upper Paleolithic arose quite some time after modern humans had emerged from Africa.
Do you see what the problem is with that pseudo-centroid? Do you see how one MUST calculate centroids properly? If so, don't come with me with that fascist rubbish of appealing to "authority" without further reasoning: neither I nor this blog welcome that crap.
Delete"No, or very few, human remains have been found anywhere in the region"...
Of course they have been found. You are being knee-jerk.
You are not constructive, not realist, not factualist and also a bore. Quit it. Write your own blog instead of bombing this one with your knee-jerk nonsense.
Delete"The location indicated by those coordinates should be somewhere in Inner Mongolia near the border with Shaanxi, slightly northeast of Yinchuan, Ningxia Hui Autonomous Region. This is close to the ancient homeland of (at least one important population ancestral to) the Han Chinese (Hua-Xia), as well as the historical seat of the Tanguts (Xi-Xia 'West Xia'), a population that spoke an archaic Tibeto-Burman (or Sino-Tibetan) language".
ReplyDeleteYes, the first part of that was my conclusion also. And close to where the EDAR370A mutation arose. Some vector was obviously involved in its spread and the only candidate that I can see is Y-DNA O, especially O3. O3 is also closely associated with Sino-Tibetan speakers.
"it does seem like a decent candidate for the region of prehistoric diversification of my putative Northern branch of C-M130, which includes East Eurasian-American C3-M217, Central-South-West Asian C5-M356, and Japanese C1-M8".
I am very much inclined to agree.
"C-M130 may have split into at least two subclades prior to the migration into Greater Eastern Eurasia (from Southwest Asia or wherever the ancestors of C-M130 parted ways with their CF-P143 relatives)".
I am very much inclined to the idea that all the Y-DNA C's were originally part of the same expansion. They don't 'obviously' split into two groups which is what I would expect if that scenario was correct. Of course it is quite possible such will be revealed eventually. My guess is that the northern branch got left behind as C moved south through eastern Eurasia to Wallacea and Australia. But a detailed Y-DNA C would reveal much.
"it would make easier to understand the division between Mtdna N and M".
Exactly.
"It seems that the temperatures started plummeting 50,000 years ago. What if hg C went through Central Asia to China during the warm period before 50,000 and then, when the climate got colder, the Neanderthals came back".
That has long been my suspicion, although I think temperatures started plummeting more like 70,000 years ago.
"Terry, when I am talking about this gene, I am not talking about 'Mongoloid phenotype’. These two things are surely interconnected and mostly expanded together, but initially they could have had their origin in several different ethnic groups".
I'm certain thay what you say is correct. Australian/Wallacean C is notable for the absence of the EDAR variant except where it has been introduced by Austronesian-speaking groups. So C reached that region long before EDAR370A did.
"As I understood it, this gene is only about hair thickness and breast size and not about the form of face or eyes or teeth".
But the genes probably expanded as a package deal. They coincide geographically to a very large extent. The genes would have become disconnected at the geographic margins of course.
Sorry, the climate link that I posted was not good, because the timescale was wrong. That is what happens when you are in hurry. According to this site http://www.dandebat.dk/klima5.htm, the temperatures were relatively high between 110,000 and 78,000, and then went down and there was a cold period from 70,000 to 60,000. After that, temperatures went up and down, but the trend was downward. The sea level reached its lowest level aproximately 65,000 years ago, so that was the best time to land in Australia. In this scenario hg C must have done its big march pretty early. It would also have been hit by Toba between 69,000 and 77,000 years ago.
ReplyDelete"Do you see what the problem is with that pseudo-centroid? Do you see how one MUST calculate centroids properly?"
ReplyDeleteBut by 'properly' you mean a calculation that comes up with the answer you want it to. Centroids are not a useful method of calculating origin anyway.
"Of course they have been found. You are being knee-jerk".
Which ones?
"the temperatures were relatively high between 110,000 and 78,000, and then went down and there was a cold period from 70,000 to 60,000. After that, temperatures went up and down, but the trend was downward. The sea level reached its lowest level aproximately 65,000 years ago, so that was the best time to land in Australia".
Very useful information. Thanks Kristiina. Althopugh no evidence has been found for an arrival in Australia as early as that it may be that was the time when the haplogroup first crossed Wallace's Line. Once it had managed to do that (almost certainly by some sort of boat) it was only a matter of time before it reached Australia (by boat). Certainly by 46,000 years ago.
By "properly" I mean obviously considering the phylogeny and not just frequency. If you can't do that, you are not worth discussing with, really.
DeleteKristiina, I'm sure ou will be interested in this post at Dienekes (if you haven't already seen it):
ReplyDeletehttp://dienekes.blogspot.co.nz/2013/03/genomewide-structure-of-populations.htmlhttp://dienekes.blogspot.co.nz/2013/03/genomewide-structure-of-populations.html
Terry, the link did not work.
ReplyDeleteJust a quick remark - my supposision is that in these paleolithic times people crossed the sea when they saw land on the other side. My hunch is that people crossed over to Australia at a time when the sea level was so low that the Australian cost was actually visible. I think that the first settlers were not yet true navigators, as they did not settle in the Pacific islands. Of course, I do not know what is the maximum distance between the two costs allowing still to see the other side and when this condition might have been met.
It all depends on the height: at sea level the horizon is about 5 km away, but on a 100m tall hill, it is at almost 40 km, and at 900 m it is more than 110 km. See: Wikipedia: Horizon - Distance to the horizon.
DeleteThere are mountains in Timor well above 2000 m., the highest peak being Tatamailau with almost 3000 m above current sea level. The horizon at such height is very far away: some 158 km.
Anyhow, the archaeology right now can only confirm human presence in Australia since c. 55 Ka ago. Although I agree that it is probably an underestimate. Some people have claimed as much as 70 Ka. or even earlier.
DeleteThank you Maju! You are good at geometry! On Google map the distance between East Timor and Melville Island seems to be c. 400 km. So, the land was visible during the colder periods. In the link I posted you (http://www.dandebat.dk/klima5.htm), there is a map showing that during the last Ice Age, you could almost walk from Sundaland to Australia. It seems that during the previous cold period between 70,000 and 60,000 the distance must have been almost as short. Actually, 55 Ka ago there was a warm period, so I think the colonizationed happened before 60,0000 years. The gap between 65 Ka and 60 Ka is really not much.
ReplyDeleteI noticed that on Dienekes' Blog, there was a link to this study (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0058552) Is the admixture analysis in figure 4a what you wanted to see? There you have a comparison between Finns, Western Uralics, Russians, and other Europians and Chinese. How should we understand it in your opinion?
There's another possible route via the smaller islands (North Malukku) between Philippines, Sulawesi and New Guinea, which, as you surely know, was part of the Australian (or Sahul) continent in the Ice Age. In that case the "island-hopping" distances to save may be a little smaller but I'm not sure about mountains.
DeleteAnyhow there's evidence of open ocean fishing in Timor before 40 Ka., what means that the inhabitants were seafarers in general. But the evidence for periods as early as 50, 60 or 70 Ka is short to come (partly maybe because seafarers would have lived in the coasts and these are now submerged, partly because evidence is generally harder to find the older it is and partly because archaeology in SE Asia still allows for many discoveries - a lot of work to do over there, I'm sure).
"How should we understand it in your opinion?"
No idea. I have it filed for comment it here but I had no time to read nor meditate on it yet. Let's see if I can get 30 mins or an hour now for it. Put up with my limitations please.
Here: http://forwhattheywereweare.blogspot.com/2013/03/autosomal-dna-of-ne-europeans.html
Delete"at 900 m it is more than 110 km"
ReplyDeleteBut even during the deepest ice age the distance between Timor and Australia was almost certainly much greater than that:
http://www.kimberleysociety.org/oldfiles/2008/ABORIGINAL%20ART%20AND%20CULTURE%20IN%20THE%20KIMBERLEY%20AND%20ADJOINING%20AREAS%20Apr%2008.pdf
Quote:
"That gap is nearly 500 km wide today, while the distance to the horizon at sea level is less than 5 km. When the first people reached Australia, sea level was 50 m lower than it is today, and there was still a gap of more than 400 km between Timor and the Australian mainland. Yet, because of the lower sea level , many present-day reefs southeast of Timor were exposed as small islands, the closest being along Hibernia Reef, about 120 km southeast of Timor. People from Timor may have first crossed the gap to those islands by chance, being swept there in a boat during a storm. They can be expected to have found rich animal life (fish, shellfish, birds, and turtles) on and
around those islands, and would have taken that information back to Timor,
prompting others to make the crossing".
If sea level was at 100 metres lower at the time of crossing the distance would still be greater than 100 km, perhaps 150 km.
"My hunch is that people crossed over to Australia at a time when the sea level was so low that the Australian cost was actually visible".
That is what is so interesting. The Australian coast was never visible from Timor, for example. However, as Maju said, we know that people on (and presumably near) Timor were fishing offshore by around 40,000 years ago. If a fishing party was blown further afield they may have reached Australia. They would realise fairly readily that they were south of where they started out and so would move back north as soon as conditions were favourable. By doing that they would almost certainly run into the lesser Sunda Islands somewhere along their east/west spread. From there easily finding their way home and telling their relations of a large uninhabited island to the southeast.
"Terry, the link did not work".
I see you've found Dienekes' blog so you presumably have now found the relevant post.
"There's another possible route via the smaller islands (North Malukku) between Philippines, Sulawesi and New Guinea, which, as you surely know, was part of the Australian (or Sahul) continent in the Ice Age".
To me that seems to have been thr route that Y-DNA K-derived haplogroups took, as well as the mt-DNA M haplogroups.
I repeat: "There are mountains in Timor well above 2000 m., the highest peak being Tatamailau with almost 3000 m above current sea level. The horizon at such height is very far away: some 158 km".
DeleteA fact you conveniently ignored.
It's almost certain that at lowest sea levels Australia was visible from the peaks of Timor, not very neatly but visible in any case.
There are islands (Yamdena) East of Timor that are even closer than Timor itself.
Said that, we do not know which route or routes they took, nor how they made their journey.
"I repeat: 'There are mountains in Timor well above 2000 m.'"
ReplyDeleteTalk sense for a change Maju. And what would inspire a people who were primarily coastal-adapted venture so far up a jungle-covered mountain? I know from personal experience it is almost impossible to see very far while climbing such mountains until you emerge above the tree line which, in Timor, is presumably quite high.
"The horizon at such height is very far away: some 158 km".
So even at the top of the mountain Australia was still invisible for much of our prehistory.
"There are islands (Yamdena) East of Timor that are even closer than Timor itself".
Very low islands. Stop making things up, as usual.
"Said that, we do not know which route or routes they took, nor how they made their journey".
True, but almost certainly via Timor. The greatest diversity of Australian Y-DNA haplogroups is in Arnhem Land according to information Ebizur provided some time back. That diverstity is presumably also true of the Kimberly region, the other part of Australia closest to Timor.
I don't know in your country but in mine people go up and down the mountain all the time: it's a passtime and in the past probably a living (hunting wild caprids, later pastoralism). Believe what you want about ancient Timorese but from experience when sea and mountain are so close people use both (logically, why to disdain an eco-niche).
Delete"Stop making things up, as usual".
I'm not making anything: I just stated the facts: there are some islands East of Timor that are quite closer to the Australian coasts and were so in the past as well. I said nothing about mountains.
"The greatest diversity of Australian Y-DNA haplogroups is in Arnhem Land"...
How reliable is that? Remember that they have been driven to near-extinction. We can only study the remnants.
Said that, I also think that they arrived via Timor or the nearby islands but they did so following, as The Dream says, the great crocodile.
But we do not know if the also saw the coast from the mountain before daring that most uncertain journey. Mystery shall remain.
"I'm not making anything: I just stated the facts"
ReplyDeleteYou are not 'stating facts'. You are stating what you believe. For example:
"I don't know in your country but in mine people go up and down the mountain all the time"
You fail to realise that the vegetation on mountains in your region is totally different from that of mountains in the tropics. And I doubt that the vegatation on mountains in your country is anything like the vegetation that was there when people first arrived. I have been 'up and down the mountains' all through my country as well (it used to be part of my job). And I can tell you that unless you're taller than the trees all you can see is the trees. It is rare to even be able to see the ridge next to the one you're actually climbing. So you are not 'stating facts'. You're imagining things.
"there are some islands East of Timor that are quite closer to the Australian coasts and were so in the past as well".
And the hills on those islands are nowhere near as high as those on Timor itself. You're using your imagination again.
"How reliable is that? Remember that they have been driven to near-extinction. We can only study the remnants".
Ebizur's data from Leherensuge some time back (sorry, I can't track down the actual post):
10.0% C-M130(xC2-M38,C3-M217,C4a-DYD390.1del) in 'Australia/Arnhem' and 53.3% C4a-DYS390.1del. In 'Australia/Desert' no C-M130(xC2-M38,C3-M217,C4a-DYD390.1del), only 68.6% C4a-DYS390.1del. Both region s have about the same small level of F-M89(xK-M9) and R1-M173 but 'Australia/Arnhem' has 30.0% K-M9(xNO-M214,M2-M353,K1-M147,S-M230,M1-M4,P-M45) while 'Australia/Desert' has 17.1% of that haplogroup. Quite possibly this last haplogroup is actually MNOPS. I agree that the sample is small but it may be significant. Perhaps Ebizur can enlighten as as to more recent information.
"But we do not know if the also saw the coast from the mountain before daring that most uncertain journey. Mystery shall remain".
I remain convinced that, like the discovery of the outer Pacific islands, the first discovery was accidental. The discoverer returned home and told his (usually 'his') relations about an unoccupied paradise and off they went.
Boring, pointless. Just trying to fight where there's nothing to argue. We do not even disagree... except in your emphasis on picking fights all the time.
DeleteDoes the Australian F* mean the root type that is very rare in East Asia or something else? There is root type K* scattered in East Asia, for example in the Filipines (see Genetic Diversity and Evidence for Population Admixture in Batak Negritos from Palawan). When we add to this R1, does this mean that there could have been four different waves of people coming to Australia? I just found this interesting study on Australian Mtdna: http://www.sciencedirect.com/science/article/pii/S1040618211002278. Also the Mtdna hg pool could be a result of four waves, as there are the following haplogroups represented: N13/14, S, O; P4; M14/15; Q2; M42; and Australian M42 is a sister clade of M74 from South China.
ReplyDeleteF* does not mean "the root type" but just F-other. Other than what? It depends which other clades were tested for (usually the ones reported). Where did you found that piece of data? Without knowing the context we cannot judge what it means.
DeleteIf it comes from the above list then F-M89(xK-M9) is clearly F but not K. It could in pure theory be G, H or IJ but it's probably something else like the F* that is found at low levels in East and South Asia (and other places, I believe that also in Europe and even Africa). We can't say much about it by the moment.
Delete"Does the Australian F* mean the root type that is very rare in East Asia or something else?"
ReplyDeleteMy guess is that it is the same as the F in the Philippines, almost certainly F2-M427. This haplogroup is also present in southern China, especially the Lahu and Yi. As you mention, Australia has mt-DNA M42a, a subset of M42'74. M42 is also found in the Munda of India while M74 is found in South China and Vietnam. This indicates some connection between South China and Australia, which makes sense.
"There is root type K* scattered in East Asia, for example in the Filipines"
That could be MNOPS, also present in the Lesser Sunda Islands. Or possibly K1-P60, now called 'K1' at ISOGG as the earlier South Asian K1-M147 has been shown to be downstream of M256, whatever that is.
"When we add to this R1"
R1 is most likely European although it could be Indian I guess.
"does this mean that there could have been four different waves of people coming to Australia?"
Certainly more than just one. My best guess is Y-DNA C with mt-DNAs N13, N14, S, O and P in the first wave. Not just 'P4'. P's basal diversity strongly suggests P coalesced there. Australia has P5, P6, P7 and P9 to itself and shares P3 and P4 with New Guinea. P1 is hardly present in Australia as it is a New Guinea/Melanesia haplogroup. Subclades of P2'10 reached the Philippines. M14 and M15 may also have been part of the first wave, or part of the next: Y-DNA MNOPS with mt-DNA M27, M28 and M29'Q some time after the first wave. The F/M42 could be an even later arrival.
"I just found this interesting study on Australian Mtdna"
Extremely interesting. Thanks. I had seen the one on Palawan before. I can only agree with this comment from the abstract:
"Mt haplotypes indicate ancient ancestry for both Australia and New Guinea peoples, with evidence of some shared genetic connection and other unshared haplogroups apparently specific to both places".
The two regions are surprisingly different if we blindly accept the ice age land connections between the two. And Maju won't like this:
"Migration into Sahul from south-east Asia may have been by more complex routes than only along a ‘southern coastal route’, raising the question of possible common ancestry in central or northern Asia for some Australian and American peoples for which current genetic evidence is tenuous".
I've been trying to tell him that for years.
That Australian mt-DNA map is extremely interesting. The four regions are quite different from each other. The Indian-related M42 makes up a substantial proportion of the Riverine Aborigines yet is almost absent through the rest of Oz. Only noted in the Northern Territory/Arnhem Land, the most diverse region as we would expect. Other interesting comparisons also emerge. Thanks for the link.
ReplyDeleteIt was pay-per-view, so I printed the diagrams and tried to read the haplogroups. Fortunately, I still see quite well with my right eye. :-) Did you read the entire article? If yes, were there any age estimations? Is it so that even if the time scale might differ from the reality, the relations between the age estimations of different haplogroup are true?
ReplyDelete"Did you read the entire article?"
ReplyDeleteNo. I'm too Scottish to pay. Like you I copied over the map which is how I got the above information. Consequently I don't know the age estimations although the authors do make a cryptic comment about disentangling the pattern of migration within Australia.
"The authors argue for a selective pressure but I find hard to imagine which one could it be. Therefore I rather lean again towards a random founder effect."
ReplyDeleteThe selective pressure in all organisms that have adaptively evolved is the nutrient-dependent thermodynamic regulation of intracellular signaling and internuclear interactions that must be controlled by organism-level thermoregulation. Hair and sweat glands = organism level thermoregulation in mammals associated with pheromone-controlled reproduction. Sexual selection for nutrient-dependent social odors (e.g., pheromones) associated with cell cycle thermodynamics and thermoregulatory fitness facilitate adaptive evolution. Random mutations in a founder don't.
Besides that purely theoretical discourse, can you pinpoint which could be the specific environmental differences between East Asia and the rest of Eurasia that would be the necessary selective pressure to actually justify evolutionary selection as a reasonable explanation.
DeleteI can find not one. Hence: founder effect.
I came across some information by accident the other day that suggests that white people are the descendants of African or East Indian albinos and that Asian people are some sort of sub species from an African tribe called the Sans people. I also watched a video where an Asian man confirmed in his laboratory that Asians are directly related to Africans and that they are not a unique race. So let's assume this is true for a minute. If you look at the Sans people, they have the exact same facial characteristics as Asians. The only difference is that they are black and have curly hair. So how do we get from a black face with curly hair to an Asian, same face, but now with straight hair? What materials do we have genetically to create this genetic selection?
ReplyDeleteI will say that being mixed with black and white, people have thought I had Filipino blood, and I know a girl who is mixed with black and white who has a strikingly Asian appearance. However many ppl of my mix look nothing like Asians, so the idea that Sans plus albino equals Asian is ridiculous. There has to be more digging done to figure out this mutation, if that's what it is. Otherwise I may still continue to believe they are unique, despite that they do look like the Sans people.
"So let's assume this is true for a minute".
DeleteIt's all just a modern myth. As every myth it has a seed of true, so yes: Asians (all Asians, not just East Asians, and derived peoples: Europeans, Australasian and American aboriginals) are descendants of emigrants from Africa some 100-80,000 years ago but such an ancient population has no known name (and are not the San or Bushmen, although there may be an less direct relation of some sort (with other Africans too or even more closely however).
In this sense you have to understand that all non-Africans (and even to some extent some Africans because of back-migration) descend from a single population with two male and two female lineages only (and these are very irregularly mixed, not forming any obvious pattern that may allow us to talk of two different populations).
It is even possible that depigmentation has some oblique relation with albinism but it's not the same result at all: depigmented or "white" types do tan, albinos don't. "White" populations (including East Asians, whose genetic path is surely different to that of West Eurasians but have a very similar result) show a gradation of shades among individuals and also in the same individual depending on solar exposition: they retain the melanin function, even if attenuated or in flexible form. Almost no "whites" are unable to tan, although some tan better and more easily than others. Instead albinos (white or black) can't tan at all and must guard themselves very tightly from sunlight. Theirs is a failure of the genetic activation of melanin altogether, while in normal depigmented (more or less "white") individuals it's a process with much more gradual and flexible results.
The real cause for depigmentation in high latitudes is purely evolutive in the most genuine sense of survival of the fittest. We humans need good doses of vitamin D and this vitamin is almost absent in food (fish excepted) so we photosynthesize it in the skin. Brown or black pigmentation instead protects us from the sun (and may have a relation not just with sunburn and skin cancer but with protecting another key vitamin: folate). Vitamin D is critical, we known now, for the proper brain development of infants, even in the womb, so it was critical to adapt to high latitudes' relative darkness: those who did not had faulty descendants, who had much more difficult lives and maybe never survived at all.
We see at least two such patterns of depigmentation: one in West Eurasia and another in East Asia. But to a lesser extent we also see it in Southern Africa, precisely among the Khoisan (=Khoi+San) aboriginal peoples, which has a subtropical climate. So there were at least three different populations who were pushed to adapt that way more or less strongly. Our skin is the best evidence that evolution works (but one has to understand why: vitamin D synthesis in the skin).
More:
Delete"If you look at the Sans people, they have the exact same facial characteristics as Asians".
Well, not quite but they do have the epicanthic fold. East Asians and Khoisans are not the only ones anyhow, if you look carefully you will notice that partial epicanthic fold is also quite common in Northern Europe (some say it's an East Asian influence, others that it's just a remnant from a shared past in Tropical Asia, etc. - nobody knows for sure, maybe both).
Whatever the case it's indeed possible that the trait originated from an East African genetic pool partly related to the Khoisan and randomly survived in some populations and not others. There are other such traits like sinodontia (most visually described as shovel-like front teeth), not uncommon in Africa but not precisely among the Khoisan. Etcetera: one has too look at the whole picture: a single trait is not determinant (being either random distribution or even, as happens with skin depigmentation, an evolutionary adaption with various origins and pathways - convergent evolution).
"I will say that being mixed with black and white, people have thought I had Filipino blood, and I know a girl who is mixed with black and white who has a strikingly Asian appearance".
Maybe but, as you surely know, that's not a rule. Say Obama: he's a true 50-50 first generation mix and he does not look Asian at all. Individuals vary, among "pure whites", for example, there's people who look more oriental. In my country they'd be often nicknamed "Chino" (=Chinese) but they had no Asian ancestry whatsoever, just the normal local variation. From your profile photo, you are very pale for an Afrodescendant, so I guess that confuses people also: you do have a hard to define ambiguous "multiracial" look and if by chance you have somewhat slanted eyes (can't see well in the photo). But it's not the Philipino look to my eyes, really: something more ambiguous instead. Your man (husband?) instead looks to me as likely Euro-American Mestizo - he does have some quite clear Native American look but also looks somewhat European (at least to my eyes and with the difficulties from such a small-sized photo - beautiful photo nevertheless).
"... so the idea that Sans plus albino equals Asian is ridiculous".
Exactly. What happens is that we are at individual level much more alike across "races", just that some traits have been either selected (skin color clearly, maybe others) or have been homogenized by reason of founder effects and the resulting concentration of some genes (some would also argue for social selection for likeness but I have all kind of doubts about it - maybe it helps a bit). But other traits are retained in great variety instead, so you have Africans (especially, because of their much greater genetic diversity) and others with variation for traits that have been fixed elsewhere. More or less fixed because you can find, for example, the occasional East Asian without epicanthic fold as well (but all other "racial" traits like thick hair or whatever), exactly the same that happens with pure Native Americans and every other population. So when you begin mixing (or sometimes even without any mix) you can end up with all kind of variations.
Maybe you have Khoisan ancestry anyhow, like, for example Nelson Mandela. Or maybe it comes from your European ancestors, or maybe some unknown other (Native American?) one... Hard to say. Quality genetic testing may help you find out something if you are very interested.
Look for example at this Basque woman (a TV journalist). AFAIK she's as Basque (i.e. purely European) as anyone can be but she has a very East Asian look to her. She's not the only one, although one of the most shocking cases I know. Not this particular look but other oriental-looking traits are common also in Northern Europe and even more towards the NE (if you have Russian or Finnish ancestors that could well be a simple explanation, for example, for slanted eyes). But my Italian grandfather also had very slanted eyes also being blond with blue eyes, so it can really be from anywhere.
Delete