January 21, 2013

Ancient North Chinese from 40,000 years ago closely related to modern locals

Tianyuan Cave (source)
The information is sketchy as of now but the news in the press indicate that an Homo sapiens from Tianyuan Cave, near Beijing, whose fragmented remains were discovered in 2003, was closely related to modern East Asians and Native Americans. 

The paper is not yet online but the information released to the media strongly suggests that East Asians were already distinct from other populations some 40,000 years ago. This would seem to be based on the sequencing of the mitochondrial DNA and the explicit mention of Native Americans indicates that the lineage must be A, B, C or D (X, the fifth and less common matrilineage of Native Americans, is not found in East Asians, with some exceptions from Siberia, so we can exclude it safely). 

Ancient DNA from cell nuclei and maternally inherited mitochondria indicates that this individual belonged to a population that eventually gave rise to many present-day Asians and Native Americans, says a team led by Qiaomei Fu and Svante Pääbo, evolutionary geneticists at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. 

This would seem to discard some adventurous hypothesis floating around about tremendous demographic changes in the Paleolithic and afterwards, at least for this region. Probably not even when "mode 4" technology arrived to the region (from Altai) c. 30,000 years ago. 

In other words: the seeds of modern populations were already there c. 40,000 years ago in East Asia (and surely also in most other regions) and, even if they may have changed somewhat, they have remained the same at least to some notable degree.

Furthermore, the autosomal DNA also seems to have been sequenced to at least some degree because the researchers state that Denisovan and Neanderthal genetic inputs are at the same levels as modern North Chinese (i.e. some 0% and 2.5% respectively):

The partial skeleton, unearthed in Tianyuan Cave near Beijing in 2003, carries roughly the same small proportions of Neandertal and Denisovan genes as living Asians do (SN: 8/25/12, p. 22), the scientists report online January 21 in the Proceedings of the National Academy of Sciences.

Or in the words of the Max Plank Institute:

The genetic profile reveals that this early modern human was related to the ancestors of many present-day Asians and Native Americans but had already diverged genetically from the ancestors of present-day Europeans. In addition, the Tianyuan individual did not carry a larger proportion of Neanderthal or Denisovan DNA than present-day people in the region.

This also seems to discard models implying Denisovan admixture happening in Siberia or NE Asia and would indirectly support my own hypothesis of admixture with Homo erectus (for which Denisovans, plausibly an Erectus-Neanderthal hybrid, would be just a proxy) in or near Indonesia.

Sources: Science News, Max Plank Institute.


Update (Jan 22): the paper is already online and is open access (cool!)  I don't think I have time to discuss it today but will do tomorrow without doubt (other than the sky falls on my head, you know).

103 comments:

  1. The Neanderthal admixture percentages are unremarkable. No one supposed that there was significant new Neanderthal admixture at any time after East Eurasian populations had reached China.

    But, the absence of Denisovan admixture in a 40,000 year old modern human is very significant. Many Asian migration wave theories suppose that first wave Asians had Denisovan admixture which was diluted to essentially zero by subsequent waves of migration everywhere but among Papuans, Australian aborigines and certain Phillipino Negritos.

    The results greatly constrain those theories requiring the obliterating waves to have run their course by 40,000 years ago - a time frame that predates by more than ten thousand years, for exampe, the archaeologically attested arrival of the Jomon in Japan, or the archaeologically attested appearance of the earliest modern humans in Tibet, two of the places where Y-DNA D is most prevalent. This sample is only about 5,000 years +/- after the proto-Papuans cross the Wallace line.

    Archaeological evidence for modern humans in China in the timeframe prior to 40,000 years ago is quite sketchy.

    ReplyDelete
    Replies
    1. "The Neanderthal admixture percentages are unremarkable".

      In raw preliminary though, it could be less or even 0% and reach current levels later by arrival of more admixed "moderns". Also some have speculated about distinct admixture events in the East and West... it is not totally unremarkable (depends on what you would expect).

      "But, the absence of Denisovan admixture in a 40,000 year old modern human is very significant".

      Absolutely.

      "The results greatly constrain those theories requiring the obliterating waves to have run their course by 40,000 years ago"...

      Absolutely again.

      Of course, the more we get to know, the better.

      Delete
    2. There might not be a total absence of Denisovan admixture. Different teams are claiming different things at the moment; some claim no evidence of Denisovan admixture on the Eurasian landmass, others claim very small amounts (well below 1 percent) in East Asia.

      In China it also seems that the larger number of modern people are included, the more likely the study is to come up with claims of Denisovan admixture.

      It's kind of same as in Africa with neanderthals. When some teams don't take into account admixture under 1 percent, then we get the claims that there is no Neanderthal heritage in current sub-Saharan African populations.

      Delete
    3. That's because of a reason: the appearance of Neanderthal/Denisovan admixture is subject to minor error because of normal variance (we are not so distant from our other Homo cousins) and that's why only when we look at the big picture we can ascertain with any confidence that there was admixture (the "most Neanderthal" of Nigerians is clearly much much "less Neanderthal" than any Eurasian-plus and only slightly "more Neanderthal" than any other "purebred" African (i.e. excluding populations with significant Eurasian admixture).

      IMO very low levels of apparent "admixture" should therefore be ignored as mere noise (normal variance). Of course we can never be 100% sure that they are meaningless but that's the level of precision we can reach objectively.

      Delete
    4. Erratum: "... is subject to minor error" should better be "... is subject to minor uncertainty".

      Delete
    5. I see reports of very low levels of Denisovan admixture (less than 1%) in SE Asia, but not in Northern China.

      Delete
  2. But maybe they didn't really split off that early as meta-populations, but rather West Asians, and therefore Europeans by extension, had experienced heavy African Admixture since OOA? Recall my K2 ADMIXTURE analysis and the significant (1/3) affinity that Europeans and West Asians had with the African cluster. If Africans were mating with West Asians since OOA, but not with East Asians because they are far away, would that not distort what the ancestral West/East Asian population would have been like?

    ReplyDelete
    Replies
    1. "... but rather West Asians, and therefore Europeans by extension, had experienced heavy African Admixture since OOA?"

      There's something of that indeed but not "heavy", more like "subtle".

      "Recall my K2 ADMIXTURE analysis and the significant (1/3) affinity that Europeans and West Asians had with the African cluster".

      I told you already that such phenomenon happens because of (1) Admixture is forced at K=2 to create exactly two clusters and place all anomalies in the intermediate zone (i.e. "neither this nor that") and (2) to mild (tending to subtle) African admixture of West Eurasians. In other examples I pointed to you in that thread, it is Africans (sampled in lesser numbers) who are forced to appear as "Europeans" or "West Eurasians", because in all A vs B comparisons they happen to be slightly (but not much) closer to WEA pops. If you reduce the sample of East Asians enough, they will surely be forced at K=2 to appear as Western Eurasians, the two poles being Africa vs Europe/WEA.

      Delete
    2. “There's something of that indeed but not "heavy", more like "subtle".”
      By 'heavy', I meant heavy enough to obscure the inference of ancestral populations.

      “I told you already that such phenomenon happens because of “
      Yes you told me but you did not prove it. In either of your points the intermediateness of West Eurasians when 2 poles have been chosen need to be adequately explained, “neither this nor that” is not really an adequate explanation, IMO. In your second point, this is something that can be easily quantifiable and demonstrated if true, and reducing East Asians only defeats the purpose, the point being, that when East Asians/Amerinds, Africans and West Asians are included in an ADMIXTURE run, West Asians always appear intermediate between the two, similar to their placement in global PCA plots.

      Delete
    3. "... but you did not prove it".

      I did not have time nor energies to do it. But above I tell you how to do the test: just put 40 Europeans, 50 Tropical Africans (spare the Pygmies just in case they distort it all) and a mere 10 East Asians and you will see a European and an African cluster and how East Asians fall fully in the European one. That's my prediction and my test setting.

      Anyhow it's common sense re. ADMIXTURE, which after all only checks for affinity and forces each individual into one of K clusters (two in your case). Two clusters is clearly not optimal for all Earth study, three may do depending on which populations you throw in.

      "West Asians always appear intermediate between the two"...

      First, West Asians, especially those from Arabia, have more African admixture than other West Eurasians - whether this is remnants of the OoA or more recent flow, remains to be properly determined in most cases (probably both).

      But more importantly, if you load the sample with East Asians and Africans, West Eurasians will be forced to more or less intermediate positions because they are not too akin to either cluster (but a bit akin to both). So you have to lighten down the sample in either extreme (Africa, East Asia). For example in Behar 2010 (and others), Africans are comparatively undersampled and therefore they cluster inside the West Asian cluster (there's no gradation among West Eurasians and the W-E duality persists at K=3, etc., after Africans are finally separated).

      Please, look at all carefully this because I am certain it is the way I say.

      Delete
    4. This comment has been removed by the author.

      Delete
    5. This comment has been removed by the author.

      Delete
    6. First, West Asians, especially those from Arabia, have more African admixture than other West Eurasians

      Only Arab (including the Druze) and some of Iranian (most probably from the south of Iran, it is not clear whether they are Arabic-speaking or Persian-speaking), Jewish and Samaritan West Asians have discernible Negroid admixture according to the results of Behar et al. 2010 and other ADMIXTURE studies, and among Negroid-admixed West Asians almost only Muslim Arabs and Yemen Jews have Negroid admixture in levels worthy of mention according to the same results. Importantly, Cypriots, Turks, Armenians, Georgians, Assyrians and Kurds show no discernible Negroid admixture while most non-Basque Iberians and southern Italians (including Sicilians) show small levels of discernible Negroid admixture in those analyses.

      For example in Behar 2010 (and others), Africans are comparatively undersampled and therefore they cluster inside the West Asian cluster

      You must have meant general West Eurasian rather than West Asian, as that cluster includes Europeans and West Asians equally.

      Delete
    7. It doesn't matter which West Asian population has more African affinity @ K2, the point is that they all have it at significant amounts relative to East Asians, this African Affinity of West Asians appears in a clinal manner, Yemenis have it at 43%, while Russians have it at 25%, and even South Asian Dalits have it at 21%, the only populations that do not carry this affinity in appreciable amounts are East Asians and Amerinds, that is the whole point.

      Maju, you are also mistaken in the way you are reading the Behar 2010 ADMIXTURE chart for K2, it is the West Asians that appear in the African cluster and not the reverse, please take a look at it carefully.

      Delete
    8. @Onur: if a population has, for example, some Y-DNA E (and all West Eurasian populations I know of do to some extent) they do have some African-specific admixture. Some of course have more and others less but that's not too important because if the ADMIXTURE algorithm produces very low values of affinity with either cluster, that will weight a lot in comparison in every case.

      What Ethio Helix does is to load ADMIXTURE with various populations from all the Old World and that makes even the Finns to appear intermediate between Africa and East Asia at K=2. I say that that's irrelevant because cross-validation would denounce that K level as mostly meaningless but how it happens is still interesting to understand.

      @Ethio Helix: It is extremely clear in Behar 2010 (PPV), in the global ADMIXTURE graph, how for example Turks appear exactly the same amount of brown at K=2 as blue at K=3 (and yellow for both). That the brown color is used at K=3 to describe Africans and not Eurasians should not mislead you (it does however) because color coding changes following a sequence that is not defined by continents but by cluster number (or arbitrary choice by the authors).

      What I say of Turks is the same for most other West Eurasian populations, although some also display minor African admixture at K=3, etc.

      In that graph it is very clear that Africans (relatively undersampled) are forced to cluster with West Eurasians at K=2 but break apart with their own cluster at K=3.

      You may say: but then Ethiopians are admixed! And I'll answer: of course, you only have to look at patri- and matrilineages but ADMIXTURE confirms it in case there was any doubt. You may not like it (I personally think that admixture is as cool as anything else, or even more, and in the case of Ethiopians very interesting because of its likely antiquity) but it is what it is. Carry your double ancestry proudly because for what they were, you are.

      Delete
    9. Luis (a.k.a. Maju),

      Ethio Helix/Etyopis confusingly often uses the phrase West Asian to denote West Eurasians, i.e., Caucasoids, in general. You have to take that into account when responding to him.

      if a population has, for example, some Y-DNA E (and all West Eurasian populations I know of do to some extent) they do have some African-specific admixture.

      Y-DNA haplogroup E is not a specifically Negroid haplogroup. So it is very normal for a population to carry significant proportion of Y-DNA haplogroup E and at the same time show no Negroid admixture.

      Delete
    10. First a correction on my previous comment, the Behar (2010) run was produced by STRUCTURE not ADMIXTURE.

      Now back to addressing your points Maju,

      #1) The color coding was most certainly not arbitrary Maju, and you know it, in effect what you are saying is that Africans split off from West Asians, which is utterly ridiculous.

      #2) Ok, forget the Behar (2010) global run, I have shown that the data-set is imbalanced anyway, here is another one that resembles my run even better,Schlebusch (2012) , an ADMIXTURE run using 270K SNPs and with a better global representation than Behar (2010). Note again the significant African affinity in all West Asians @ K2.

      #3) Let's try to keep topic of your blog post, which is not about my ancestry but about West and East Asians splitting off from each other > 40KYA. My point is that they while they may have very well split off from each other GEOGRPAHICALLY >40 KYA, what makes these scientists so sure they went their separate ways genetically that long ago? Especially given the fact that modern West Asians today have an African affinity that modern East Asians don't have, and that likely this affinity has always been there and maintained since OOA, thus obscuring any inferred ancestral population for any subset of West Asian populations.

      Delete
    11. This comment has been removed by the author.

      Delete
    12. Ethio Helix, please stop referring to West Eurasians as "West Asians", as that is confusing people and, more importantly, is wrong terminology. Also, the Behar et al. 2010 run is an ADMIXTURE run, not a STRUCTURE run.

      Delete
    13. By West Asia I mean the ENTIRE western part of the massive continent known as Asia, which includes the European peninsula, I do not mean just the Middle East as for example the United Nations defines West Asia. I actually find the term Eurasia confusing and erroneous as Europe is only a tiny fractional part of the Asian continent, but, whatever.
      With respect to Behar (2010), yes it was an ADMIXTURE run, I was correct the first time, but I confused it with another study later.

      Delete
    14. If you do not like the term Europe, then do not use the term Asia either, as Asia by definition exludes Europe since the Antiquity, and only use the term Eurasia. That would not confuse people.

      Delete
    15. I have no personal problem other than confusion with using the shorter and rather realistic geographical term "West Asia" instead of the more common and easy to understand "West Eurasia" but I prefer the latter for the sake of mutual understanding. Also when I say "West Asia" I mean the "Near" or "Middle East", another unfortunate misnomer but...

      But there is no point in discussing "races" as Onur does so aggressively. I wouldn't dare to proclaim that there is a "Negroid race". What is clear is that haplogroup E (Y-DNA) is by origin and much of its downstream processes purely African and that original Africans were more or less "Black", even if later Eurasian inputs have changed that reality in some parts of Africa (the North mostly).

      The presence of Y-DNA E and other lineages (more and more L(xM,N) shows up in West Eurasia, sometimes with very old dates of arrival, others more recent maybe) imply African admixture ("Negroid" if you wish, not my choice of words anyhow) at minor but very real levels. Let's get real!

      @Ethio Helix:

      "The color coding was most certainly not arbitrary"...

      Either arbitrary or automatic it is meaningless. You have used ADMIXTURE: you know how it is: how clusters which are obviously the same change colors from one K-level to the next just because of the change in numeration.

      "... in effect what you are saying is that Africans split off from West Asians"..

      Not at all! What I say is that you cannot reconstruct a genetic tree of Humankind with ADMIXTURE. You should know that too. You need haploid DNA for that, or maybe other autosomal techniques like the not-too-trustworthy TreeMix.

      No time for more. Later...

      Delete
    16. "... here is another one that resembles my run even better,Schlebusch (2012)"...

      That one acually resembles your run in that West Eurasians are split (even if the resulting proportions are quite different) between the East Asian (not "Asian" nor "Eurasian") cluster and the African one. As it happens to you at K=3. What differentiates both from Behar's approach is that in his run, Africans are relatively undersampled: they weight less overall and therefore they become the "victims" of K=2 oversimplism: this or that, being this or that in Behar's case: West Eurasian or East African. And because mutual secondary (mostly minor) genetic influences, all Africans end up as apparent West Eurasians (only at K=2).

      Notice please that unlike in your run or that of Schlebush, in Behar's West Eurasians do not appear as part-and-part. That is very important to know that that cluster is primarily West Eurasian.

      "Let's try to keep topic of your blog post"...

      I beg you, indeed.

      Delete
    17. What is clear is that haplogroup E (Y-DNA) is by origin and much of its downstream processes purely African and that original Africans were more or less "Black", even if later Eurasian inputs have changed that reality in some parts of Africa (the North mostly).

      Incorrect. Africa was racially much more diverse before the relatively recent rise to domination of the Negroid race in its Sub-Saharan part. There were Caucasoids as far south as what is now South Africa (e.g., the Hofmeyr man) during the Upper Paleolithic. Caucasoids arrived much of Africa before the arrival of Negroids to those regions. In fact, the Negroid race seems to be a relatively late formation (originally from western Sub-Saharan Africa? modern-archaic hybrids?). Thus, the original Y-DNA Hg E carriers were probably Caucasoid. The fact that Y-DNA Hg E is descended from Y-DNA Hg DE, which most likely originated somewhere between Africa and East Asia, further supports that.

      Delete
    18. This comment has been removed by the author.

      Delete
    19. @Etio Helix:

      "My point is that they while they may have very well split off from each other GEOGRPAHICALLY >40 KYA, what makes these scientists so sure they went their separate ways genetically that long ago? Especially given the fact that modern West Asians today have an African affinity that modern East Asians don't have, and that likely this affinity has always been there and maintained since OOA, thus obscuring any inferred ancestral population for any subset of West Asian populations".

      Actually the main OoA population (M, N mtDNA haplogroups) migrated to South and SE Asia before colonizing West Eurasia (parts re-colonized, with plausible remix with other minor OoA pops with L(xM,N) lineages, but mostly anew, taking it from Neanderthal hands).

      The Eurasian colonization did not directly spawn from West Asia but from South Asia (and SE Asia). West Asia (parts of it, Arabia Peninsula mostly) only acted as initial corridor or "pump", it was not the real source of the Great Eurasian Expansion.

      In approx. chronology:
      · c. 125-90 Ka preliminary "pump" out of Africa (to Arabia)
      · c. 80 Ka arrival to South Asia and first expansion (M explosion)
      · c. 70-50 Ka secondary expansions in SE Asia and surroundings (Australasia, East Asia in general, backflow to South Asia)
      · c. 55-40 Ka colonization of West Eurasia (mostly in Neanderthal hands before that) and probably also back-flow into parts of Africa.

      So there was a period of at least 30 Ka in which both populations ("Africans" and "Asians") were quite strictly separated. That explains why WEA and EA pops. have relatively high Fst values when compared but not quite as high as either do when compared with Africans. This isolation from Africa has persisted in "Eastern Eurasians" (when I say this I mean to include Oceanian and American aborigines) and mostly also in South Asians as well. But not quite in West Eurasians who regained direct contact with Africans and hence became somewhat "African" (post-OoA Africans) ourselves (not much maybe but still meaningful).

      So the affinity (other than being human) has not "always" been there but was regained with the colonization of West Eurasia (and parts of Africa) from South Asia c. 55-35 Ka ago.

      Delete
    20. @Onur: nonsense! There were no "Caucasoids" in South Africa (before European colonization), that's a total nonsense proper of ridiculous Hitlerian propaganda, nor "Caucasoid" is any sort of absolute category but just a relatively homogeneous range of phenotypes which, logically, is approximately coincident with a zone of relatively homogeneous genetic ancestry.

      You already have warnings (did I not banish you altogether?) for racist-mongering. Beware your steps. Certainly I can pass with none of your racialist babbling.

      Delete
    21. It is scientifically proven that the Hofmeyr Skull (a 37,000-year-old skull from what is now South Africa) is a Caucasoid skull (of the Cro-Magnoid variety of the Caucasoid race):

      http://en.wikipedia.org/wiki/Hofmeyr_Skull

      Delete
    22. The Negroid race seems to be partially descended from the Caucasoid race. This explains the general genetic affinity between Caucasoids and Negroids relative to Mongoloids, Australoids and ASI.

      Delete
    23. Ok I will use Eurasia to avoid confusion, however, I disagree with it as it is simply nothing more than another manifestation of European cultural hegemony rather than any objective geographical description.

      Anyway, first let me mention something with respect to haplogroup E, since it is once again being dubiously implicated as having non-African origins, the Plaster thesis that I had posted on my blog a few months back found two cases (out of 69) where E-M96 (x P147, M75) were found in Ethiopia, specifically amongst the Amhara dataset. This is significant as it proves, again, the antiquity of E* in East Africa, previously, only one find in Southern Africa was found by Karafet (2008) , an additional 2 finds were reported by Abu-Amero (2009) in their Saudi Arabian dataset, but these could have very easily been transported from nearby Africa in recent times. In addition, the database of private genetic testing company FTDNA reports one additional case of E-M96 (x P147, M75) from Ethiopia. So this is in addition to all the subclades of E-M96 that are found in Africa, most of which are exclusively found in Africa. Therefore, people who implicate E-M96 having non-African origins do so, on not only very flimsy data, but also on grounds that are questionably parsimonious.

      I will be getting back to the Autosomal issue a little bit later Maju, after I get some other work done.

      Delete
    24. I already mentioned some of the proofs of the partial Caucasoid origin of the Negroid race, giving some examples from uniparental genetics (Y-DNA Hg E), the fossil record (the Hofmeyr Skull) and briefly touching upon autosomal genetics.

      Here is a clearer autosomal genetics-based proof of the partial Caucasoid origin of the Negroid race:

      http://4.bp.blogspot.com/-unaJATtPLV4/T2jLIx5utbI/AAAAAAAAErY/vtoOrMfBmUA/s1600/1_2.png

      The Late Stone Age of Africa (Africa's equivalent of the Upper Paleolithic) began shortly after the beginning of the Upper Paleolithic in West Eurasia, buttressing theories of migration into Africa from West Eurasia during the same time.

      In summary, autosomal and uniparental genetics and the fossil and archaeological record all indicate partial Caucasoid origin for the Negroid race.

      Delete
    25. This comment has been removed by the author.

      Delete
    26. Ethio Helix,

      East Africa is one of the most recently Negroidized parts of Sub-Saharan Africa, and also the least Negroidized part of Sub-Saharan Africa (especially the Horn region). So an East African origin for Y-DNA Hg E by no means indicates Negroid origin for it.

      Delete
    27. Onur: you had (took) your tribune for your peculiar racialist ideas. We don't give a shit about them (at least I don't). Just to say that the evidence for Eurasian (i.e. "Caucasoid") influence in Africa beyond certain regions is very thin at best (Chad area) or simply negative at all.

      Now please search for a forum where you make people less uncomfortable before I have to start moderating comments again, thanks.

      Delete
    28. Just to say that the evidence for Eurasian (i.e. "Caucasoid") influence in Africa beyond certain regions is very thin at best (Chad area) or simply negative at all.

      Incorrect again. As an example, see this thread:

      http://dienekes.blogspot.com/2012/03/dual-origins-of-sub-saharan-africans.html

      Delete
    29. @Ethio Helix: personally I'd gladly write "Asia" with Europe included because I reckon the arbitrariness (or at the very least very debatable status) of the distinction of an European "continent" that is actually a subcontinent of Asia, a bit like South Asia but larger. It responds however to deeply embedded cultural concepts that come from the Hellenistic period, from a viewpoint centered in the Mediterranean Sea (and secondarily the Black Sea) and, because of the European prominence in global culture to date, has managed to stick. Also Asia proper is the main geographical region of origin of all Eurasians, Oceanians and Native Americans, so I don't mean at all to question the prominence of Asia, just that people will misunderstand my words if I use "Asia" instead of "Eurasia".

      Also I have no question whatsoever about the African origin of Y-DNA E. Otherwise we'd find M and/or N (or some other clearly Asian-centric matrilineage accompanying it). IMO E coalesced together with mtDNA L2"6 or L2'3'4'6 and expanded with it. For some not too well understood reason E displaced all brother lineages in Africa while some of these (C'F, D or rather pre-D) survived in Asia instead - this can probably be best understood if we imagine the L2"6 population as a single one (or a few closely related ones) around the OoA initial timing with E simply drifting out all competitors except at the other side of the Red Sea. If the early OoA phase in Arabia lasted as I think for 35,000 years or more, that allows for a lot of diverse "exiles" to mingle and randomly pick up "champions" (lucky lineages that managed to expand upon arrival to South and SE Asia since c. 80 Ka).

      A variant possibility would be that the L3'4'6 population (with Y-DNA E) would have somehow conquered the L2 population at an early stage and effectively exterminated their male lineages somehow (or even vice versa). But whatever it happened it is a matter of African populations not leaving Africa until much later times. The key is IMO always mtDNA: in Africa as in Oceania or wherever else: because male lineages are much more prone to random sweeps than female ones, it is the matrilineages which give us a coherent and mostly quite clear pattern of the major migration and expansion processes.

      Delete
    30. @Onur: go and tell Dienekes that, according to you, he is right. I'm bored of that nonsense.

      Delete
    31. that nonsense

      Maybe politically incorrect, but not factually incorrect or nonsense. It best explains the relative genetic and physical affinity between the Caucasoid race and the Negroid race.

      Delete
    32. Anthrpometry is not rocket science. It may work somewhat in the right hands and at generic statistical level but individuals way too often break the patterns.

      Delete
    33. My proofs are mostly genetics-based rather than anthropometry-based. But anthropometry, too, is advanced enough not to confuse the races of mankind.

      Delete
    34. Whatever... I don't feel like discussing this anymore here. Do you have a blog where I can spam you instead?

      Delete
    35. Onur: "the original Y-DNA Hg E carriers were probably Caucasoid. The fact that Y-DNA Hg E is descended from Y-DNA Hg DE, which most likely originated somewhere between Africa and East Asia, further supports that."

      A famous haplogroup Y-DNA D population of Asia.

      picture of Andaman islanders

      Nuff said.

      Delete
    36. Do you have a blog where I can spam you instead?

      Don't be so rude.

      Delete
    37. A famous haplogroup Y-DNA D population of Asia.

      picture of Andaman islanders

      Nuff said.


      There are populations high in Y-DNA Hg D that are Mongoloid (e.g., Tibetans, Japanese) and Ainoid (e.g., Ainus). So what?

      Delete
    38. This comment has been removed by the author.

      Delete
    39. Also, Andaman islanders are not racially Negroid.

      Delete
    40. "Don't be so rude".

      Who's the rude one here, Onur? You have been invited to leave several times and keep coming. You just posted four separate single-line comics. Even if everybody else here feels very uncomfortable with your disparate mix of racialist speculation and extremely unlike genetic conjectures (which together denounce your obvious racism and Eurocentrism even if you may not notice in your obtuseness).

      I mean it seriously: write a blog so I can go and spam you for a change. So you can learn how annoying your kind of attitude can be. It's something I recommend not just to you but to everyone with way too "clear" ideas but apparently unable to write a single article: go and write your damn blog. If I feel like I'll go criticize your work. I won't be even a fraction of rude you are, don't worry: I know when to step back and leave - unlike you, it seems.

      Delete
    41. This comment has been removed by the author.

      Delete
    42. Don't worry, I did not return to your blog. I joined this discussion only for a very specific reason and will leave your blog very soon. Your ideological misinformations and name-callings are so annoying and tiresome that I don't want to participate your blog discussions.

      Delete
    43. Can very soon be NOW? You are certainly abusing my patience quite a bit in many senses.

      Delete
    44. Don't expect me to not respond to the posts directed at me.

      Delete
    45. Now you will NOT reply to anything else ever again in this blog. You have forced me to use comment moderation ("violence") twice and for that reason I simply do not want to hear of you again EVER.

      Delete
    46. Onur:
      "There are populations high in Y-DNA Hg D that are Mongoloid (e.g., Tibetans, Japanese) and Ainoid (e.g., Ainus). So what?"

      So you actually think the fact that 2 populations living in very different geographical locations, both "alone" in the same side branch of the phylogenetic tree, both offshoots from hg DE, ended up with common phenotypical features by mere coincidence?
      According to you, DE was "caucasoid" but both E and D (see below) would ending up w/ common non-caucasoid/African-like features (and very rare in Eurasians) in very different part of the world, without contacts with each others? What are the odds?

      How convincing....

      The Andaman islanders look like a fossil population (women have steatopygia, for instance).
      It's obvious that the other Asian Ds have become mongoloid through millenary-long admixture (for starters, they are all in population with diverse Y-DNA haplogroup (among them Y-DNA O, which is clearly to be linked w/ mongoloid type) and diverse mtDNA haplogroups, unlike the Andaman populations we're talking of... so who's to be expected to have drifted towards the continental "norm" (so to speak) through the time.

      The high percentage of Ds in Japanese or Tibetans says basically nothing of the initial phenotype of these early hg D-carriers (think of the European-looking N1c1 Finns or of the East Asian-looking R1a1a Altaians).

      "Also, Andaman islanders are not racially Negroid."

      I know (and they're not caucasoid either anyway). It doesn't change a thing from my point of view, given the very ancient processes of settlement/mixing in Africa and Asia that we can infer.
      We can still see several "typical" common features for populations that share the same ancestor in the Y-DNA phylogenetic tree, aside from the main Eurasian group.
      Don't expect me to believe their common ancestor (DE) was caucasoid and even less cro-magnoid (hofmeyr), if that's what you had in mind (your exact vision of all this wasn't so clear to me).

      Delete
  3. I think all those haplogroups are possible in those times in north China: M7, D, M8, M9 maybe even G, or N/N9, B, R9/F, maybe even R11. When they talk about native American, M8(as parent of C), D, B are highly probable. I dont think it will be N (parent of A) because they would not talk about difference from Europe in that case.

    ReplyDelete
    Replies
    1. The articles explicitly mention relatedness to modern East Asians AND Native Americans. The latter have only four matrilineages (excluding X2) so it should be either of them (A, B, C or D). Guess we can be flexible and consider all M8 (which includes C and Z) but all the rest you say should not be (unless both articles has a blatant error, what seems most unlikely), i.e. not any of these you mention: M7, M9, G, N9, R9, R11...

      Delete
    2. PS- As mere bet (i.e. based only on my educated intuition) I would happily guess that not B, which has a more southernly distribution and, being derived from R may have expanded in a secondary wave (together with Y-DNA O). But then of course that secondary wave may well be before 40 Ka. so it's just a hunch after all.

      All the rest (D, A and M8/C) may well be very old in NE Asia. An even thinner bet would be D, which I deem the oldest to expand in Northernmost East Asia.

      Delete
    3. I guess its B, it seems to have some branches in that time, so it probably expanded earlier than D. Haplogroup A - surely not, its younger. M8 is possible but its also too young. I am pretty sure about B.

      Delete
  4. Where's the paper anyway? Isn't it supposed to be out by now?

    ReplyDelete
    Replies
    1. It was suppossed to be out yesterday but I haven't been able to find it yet. Can't do anything about it, sorry.

      Delete
  5. I'm quite surprised of those results, i did not expected them.

    ReplyDelete
    Replies
    1. Can't imagine why. They seem to fit well with my own understanding on the expansion of H. sapiens, creating the basis of modern populations already in what I call the Great Eurasian expansion. The admixture event with Neanderthals happened soon after the OoA while the one with H. erectus must have happened in Indonesia, soon before the colonization of Australasia.

      Delete
    2. Well, it is just that tend to think that post-pleistocene human migrations (i.e. of Neolithic chronology, recent prehistory, etc) are of much importance. For Europe & Mediterranean i recently attended to a conference of Eva Fernández Domínguez & Cristina Gamba, at Residencia d'investigadors (CSIC) in BCN. And she gave a preview of unpublished results on ancient DNA (than i cant really glose as i'm not an expert on genetics) that pointed to an important population remplacement between last mesolithic and first neolithic, in some areas & some continental routes. I'm sending to yout email an extremely low quality cellular phone pics (sorry) and the handout for the conference.

      Delete
    3. It is a common belief (and some people are borderline fanatic about it) but I do not share it. It is still possible that some, even many, of Neolithic and post-Neolithic flows were of great importance in many regions (parts of Europe for example) but not in the simplistic way that some assume (i.e. a unique super-wave from Anatolia that almost totally replaced the pre-Neolithic populations); an intermediate approach is probably the most correct one (in Early Neolithic Europe, for example, there were probably bouts of expansiveness followed by admixture and hiatus and then another expansion wave with a different origin, etc.)

      It is even possible (to be confirmed or not) that there was an specifically Megalithic wave in Atlantic Europe carrying genes from SW Iberia (including some from West Asia and North Africa but not dominantly so) - just saying.

      Also while in some parts of Europe, like the Basque Country (cf. Paternabidea, Navarre), the Neolithic genetic pool is almost identical to modern one (and some pre-Neolithic lineages also appear to have survived to modern times), in others like Central Europe the modern genetic pools do not appear till the Bronze Age (could well be Chalcolithic however, lacking data) - but still some modern elements appear in the Neolithic and may come from more southernly parts of Europe like Italy or the Balcans.

      Each of these aDNA studies give us a bit of more info to reconstruct those processes. In any case massive transcontinental migrations of decisive importance across Eurasia can be safely discarded now, the various "racial" stocks probably formed on early Upper Pleistocene (or even Late Middle Pleistocene) seedings. This should be very obvious for anyone familiar with the genetics of modern populations, for example there is a quite large genetic distance (Fst) between West and East Eurasians, not as large as these two with Tropical/Southern Africans but just clearly large enough to be very very old.

      I got your email, I'll check it now. Thanks.

      Delete
    4. This comment has been removed by the author.

      Delete
  6. Per Dienkes, quoting from the now available paper: "Thus, it is related to the mtDNA that was ancestral to present-day haplogroup B (Fig. 1), which has been estimated to be around 50,000 y old (18) (50. 7 ka BP; 95% CI: 38.1–68.3 ka BP). We note that the age of the Tianyuan individual is compatible with this date." B is, of course, part of the mtDNA N clade rather than the mtDNA M clade.

    ReplyDelete
    Replies
    1. Ah, the paper is already online? Wow! And it's open access - what a luxury!

      Haplogroup B... not my bet but very interesting anyhow. That means that mtDNA R had already expanded c. 40 Ka ago (something we could also infer from the fact that West Eurasia, including Europe, where it's very dominant, was already in full process of colonization by then).

      Very cool, I'll comment more extensively when I read it. :)

      Delete
    2. Looking at the paper itself, it looks to me like there is some very thin, but non-zero non-Neanderthal archaic admixture in Asians that is not just noise. The Europeans and Africans hover around 0.2-0.25 of a percent Denisovan admixture, while New World and Asian populations reach as much as 1% and this sample is right at the top of that range.

      The combined archaic admixture is also right at the top of the modern range for Asians and above the highest European individual by a measureable amount, but drops right into the European range when about 0.8% archaic Asian admixture is taken out.

      If one uses an initial admixture rate of 8% in actually interacting populations found in a prior study for Papuans, this would suggest a roughly 10-1 dilution of a first wave admixing population with a second wave non-admixed population in Asia, which is well short of genocidal. It is on the same order of magnitude of Cro-Magnon v. post-Cro-Magnon admixture in Europe by some estimates.

      There is room in time for an earlier wave. Dates for a first wave into East Asia in the window of about 75,000 to 45,000 years ago are plausible (particularly given a recent SE Asian mountains date possibly in the 60,000ish years ago) even though the archaeological evidence is thin (I have real doubts about 100,000 years ago, but I suppose it isn't entirely ruled out given what we now know about early Out of Africa dates going back to pre-100,000 years ago by a number of thousands of years).

      Also, mtDNA haplogroup B derived from N in a second wave, makes somewhat more sense to me than mtDNA haplogroup B in a first wave (which I would be inclined to expect to be M derived).

      Delete
    3. "... it looks to me like there is some very thin, but non-zero non-Neanderthal archaic admixture in Asians that is not just noise".

      I can see that too but notice that only one African was used as control, a wider sample may reduce those differences to mere noise. Also PCAs are sometimes confusing, in this case the African is only considered in relation to the Eurasian-Oceanian wider sample and their differences, which must not need to be only a matter of archaic admixture.

      "The combined archaic admixture is also right at the top of the modern range for Asians and above the highest European individual by a measureable amount, but drops right into the European range when about 0.8% archaic Asian admixture is taken out".

      Maybe but ask John Hawks. Other data says differently, exactly the opposite in fact, so it's threading a bit too thin.

      "There is room in time for an earlier wave. Dates for a first wave into East Asia in the window of about 75,000 to 45,000 years ago are plausible"...

      I don't see anything in the Sapiens-specific genetic data allowing for two clearly differentiated waves into East Asia, so I'd discard that pretty much. Still the single wave may have been multilayered or otherwise complex (it probably was) but then we see among Papuans pretty much the same great lineages we see in China, India or Europe, so... one single greater wave is the most likely explanation as I see it.

      "Also, mtDNA haplogroup B derived from N in a second wave"...

      Actually it is derived from R (which in turn is derived from N, yes). I see no particular reason in all to talk of a clear second wave but rather complex rapid dynamic and sometimes bidirectional flows between South and SE Asia.

      Personally I think that N coalesced in SE Asia and did it very early in the process of Greater Eurasian Expansion (GEE), otherwise why is it the most basally diverse of all macro-lineages reaching that area?

      Delete
    4. Well, my guess was correct. Its B...

      Delete
    5. Gratz. I though it was the most unlikely bet but still correct in the range of options (consolation prize, I guess). :)

      Delete
    6. My guess was mainly based on this paper: http://www.cell.com/AJHG/abstract/S0002-9297%2809%2900163-3 (you criticized it) I found this paper trustworthy. This case confirmed it again.

      Delete
    7. Don't have time to check it in depth now but my criticism was probably about still too recent age estimates at least in some cases. For example that paper gives an age estimate for the divergence Homo-Gorilla of c. 8 Ma, when that is in my understanding the lowest possible bound for Pan-Homo, which is necessarily more recent than the Homo/Pan-Gorilla split.

      But whatever.

      Delete
  7. "Haplogroup B... not my bet but very interesting anyhow. That means that mtDNA R had already expanded c. 40 Ka ago (something we could also infer from the fact that West Eurasia, including Europe, where it's very dominant, was already in full process of colonization by then)".

    I wouldn't have picked B either.

    "It is still possible that some, even many, of Neolithic and post-Neolithic flows were of great importance in many regions (parts of Europe for example)"

    And in East Asia.

    "in Early Neolithic Europe, for example, there were probably bouts of expansiveness followed by admixture and hiatus and then another expansion wave with a different origin, etc."

    And in East Asia.

    ReplyDelete
  8. Behar et al. 2012 have an age estimate for B4'5 of about 50.000 ybp, and this sample is B4'5 with 4 extra mutations, including 3 in the coding region and 1 in HVR1. This is roughly about what you would expect, on average, for the number of mutations in a 10.000 year period, and with the sample dated at 40.000 ybp, it is consistent with Behar's age estimate. Of course there is large uncertainty for a single sample, so it will be interesting to see more very ancient mtDNA results. The paper seems to suggest that the method described will allow more very ancient samples to be tested.

    Regarding Onur's racial commments - we know that both mtDNA and y-DNA uniparental haplogroups are deeply rooted in Africa. You have to question either the motives or the sanity of anyone who argues otherwise.

    ReplyDelete
    Replies
    1. There's an age estimate of 50 Ka for B or B4'5 floating around and in this case it could be close to the truth. But in general I simply disregard age estimates because they are often presented as "facts" when they are just more or less "good guesses", sometimes even horrible-looking hunches in fact. Hopefully the method will be refined but, only for mtDNA, the method looks deeply flawed to me because branch lengths are extremely unequal and to that the answer from the usual models is to assume greater times, much larger in fact, for SNP-coalescence in the shorter branches, while I'd argue that instead they become "fossilized" early on for reason of demographic dynamics in large populations, in which the dominant clade will almost invariably "drift out" the small ones. Instead in peripheral clades with small numbers the "clock" may actually tick close to max. speed because the chances for replacement (by mere drift) by the "mutants" is much closer to even. It's slippery terrain in any case.

      Delete
  9. A few patterns I've noticed in the data:

    1. The affinity to the Denisovan sample is greatest in the Papuan sample, followed by the samples from China (Dai and then Han). The European, American, and African samples do not order clearly (the Karitiana, Yoruba, and Mandenka samples actually fall between the French, who exhibit greater affinity to the Denisovan than the Karitiana, Yoruba, or Mandenka exhibit, and the Sardinians, who exhibit less affinity to the Denisovan than the Karitiana, Yoruba, or Mandenka exhibit), but the order would be Americans followed by Europeans followed finally by Africans if the French and Sardinian samples are averaged to represent Europeans as a whole and the Yoruba, Mandenka, and Dinka samples are averaged to represent Africans as a whole. Overall, the pattern seems to suggest the greatest amount of Denisovan-related admixture in Australasia, followed by China with about 1/3 as much Denisovan-related admixture as the Papuans, and finally followed by America, Europe, and Africa with approximately equally low or non-existent levels of Denisovan-related admixture.

    However, the sample of ancient DNA from Tianyuan Cave in northern China exhibits the same level of affinity to Denisova as the modern Papuans do, which raises the question of why, if the mtDNA haplogroup of the Tianyuan Cave individual really does belong to a specifically East Asian-American ("Mongoloid") clade, the modern Mongoloid samples (and especially the Karitiana individual) exhibit levels of Denisovan affinity that are much closer to the levels exhibited by modern Europeans and Africans.

    2. In regard to the European samples, the expected pattern is observed, with the insular Sardinians exhibiting greater PND with almost all populations relative to the cosmopolitan French. The only exceptions, populations with which the Sardinians exhibit less PND than the French, are the Mandenka [-732], the Papuans [-662], and Tianyuan [-262].

    3. The French-Dai PND is much less than the Sardinian-Dai PND, even more than should be expected according to my observation of the pattern mentioned above in #2. Could the often conjectured protohistorical Indo-European influence in Yunnan be responsible?

    ReplyDelete
    Replies
    1. Notice please that, I stated above in response to Neanderthalerin, Tianyuan is a very old individual, so all the mutations accumulated in these 40,000 years do not exist for him/her. That means that Tianyuan is necessarily somewhat closer to the ancestral genotype of Homo sapiens that any of the modern samples and that ancestral genotype must be somewhat closer to the Denisovan one, who is a rather close relative of our species, than any of the modern ones. We have to correct for that.

      Apparently the authors already did in their other analysis like fig. 3.

      Delete
  10. 4. Among the samples of modern non-African populations, the American (Karitiana) sample exhibits the least PND with the sample of ancient DNA from Tianyuan Cave, followed by the Han, Sardinian, French, and Dai samples in that order. The Papuan sample exhibits the greatest PND with Tianyuan among the non-Africans, but the African populations exhibit even greater PNDs with the Tianyuan sample, in the expected order (Dinka < Mandenka < Yoruba < Mbuti < San, i.e. following the order expected according to the general Eurasian affinities of the various African populations). However, both the Tianyuan sample and the Papuan sample exhibit a similarly high affinity for the Denisovan sample, and the Karitiana, Europeans, and Africans exhibit a similarly low affinity for the Denisovan sample, despite the closeness between the Karitiana and Tianyuan and the distance between the Papuan and Tianyuan.

    In other words,

    Karitiana: close to Tianyuan, far from Denisovan, far from Papuan
    Tianyuan: close to Karitiana, close to Denisovan, far from Papuan
    Papuan: far from Karitiana, close to Denisovan, far from Tianyuan

    To me, this suggests a two-prong colonization of "Greater East Eurasia" (including America and Australasia), with an initial split between a "Northern" (proto-Karitiana/Tianyuan) branch and a "Southern" branch (proto-Papuan), followed by at least two distinct episodes of admixture with Denisovan-related populations, at least one in East Asia (proto-Tianyuan) and at least one in Southeast Asia/Australasia (proto-Papuan). This makes me suspect that the Tianyuan population was a proto-American population that had experienced some admixture with a Denisovan-related population, hinting that the proto-Americans were probably the first population of modern humans to arrive to East Asia (thus, it was possible for them to interact with Denisovan-related archaic humans). This Denisovan-influenced proto-American population probably has formed the substrate upon which modern populations of East Asia have been formed. However, ...

    ReplyDelete
  11. 5. ...the Denisovan-influenced proto-American population in East Asia must have received some later influence from a Euro-African population. Note that the Karitiana and the Papuan samples are, to an approximately equal degree, the most distant from modern Africans of all sampled modern non-African populations; considering the Denisovan-related admixture that is evident in modern Papuans, the lack of evidence of such Denisovan-related admixture in modern Karitiana, and the great PND between modern Africans and Denisova, it must be assumed that the proto-Papuan population prior to Denisova-related admixture (i.e."Southern branch of the proto-Greater East Eurasian population") was somewhat closer to modern Africans than the proto-American population (i.e."Northern branch of the proto-Greater East Eurasian population"). Despite this fact, and the fact that both Tianyuan and modern East Asians share a close affinity with Karitiana, modern East Asians exhibit much closer affinity to modern Africans (but, surprisingly, not so much to modern Europeans, except in the case of the anomalous French-Dai affinity mentioned in #3 above) than either Tianyuan or Karitiana exhibit to modern Africans. This must mean one of the following three things: (A) a population of African affinity has influenced the proto-East Asian population (i.e. "post-Denisovan-influence proto-American remnants in East Asia") after the proto-Americans had populated America; (B) a population of East Asian affinity has influenced Africa after the proto-Americans had populated America; (C) an extinct or unsampled population has influenced both East Asians and Africans after the proto-Americans had populated America. Scenario (A) is nice in that it can explain the slightly greater PND between European and East Asian populations relative to European and American populations (except the noted French-Dai anomaly) without requiring an extra episode of American > European or European > American gene flow (Africans are more distant from any non-archaic Eurasian population than any pair of non-archaic Eurasian populations are from each other, so African admixture in East Asians should tend to pull them away from Europeans relative to the Americans, who lack such African admixture), and it explains the lower affinity of modern East Asians for Denisova despite the high Denisovan affinity of the ancient sample from Tianyuan (Africans have a very low affinity for Denisova, so any African influence on East Asians would tend to lower the latter's Denisovan affinity). However, where is the African-related haploid DNA in East Asia? Y-DNA haplogroup D might be suspected, but there does not seem to be any matrilineal counterpart. Furthermore, the European samples, like the East Asian samples (and actually to a slightly greater degree than the East Asian samples on average) also exhibit an affinity for modern Africans relative to the very anti-African Americans and Papuans; if the slight affinity between the Europeans and the Karitiana sample were due to a lack of the hypothetical African admixture found in East Asians, then why do Europeans exhibit African affinity like the modern East Asians (and perhaps to an even higher degree)? However, I suppose that the illusion of slight Karitiana-European affinity could also be explained by gene flow between Americans and Europeans or by East Asians' being dragged away from Americans due to East Asians' residual Denisovan-related admixture (bequeathed from the Denisovan-admixed branch of the proto-American population in East Asia approx. 40,000 YBP, represented by Tianyuan).

    ReplyDelete
    Replies
    1. Or noise. The differences are so thin that I don't dare to say only on this data.

      "why do Europeans exhibit African affinity like the modern East Asians (and perhaps to an even higher degree)?"

      Europeans have some minor African admixture (Y-DNA E and some L(xM,N) matrilineages) that East Asians lack. West Eurasians have been in contact with "aboriginal Africans" since the colonization of the region (earlier in Neanderthal hands essentially) c. 55 Ka ago and successive millennia, what makes for some unavoidable genetic flow in both directions. In addition to that there were surely residuals from the preliminary OoA-related migrations into Arabia, Palestine, North Africa... that may have also influenced the issue (but harder to evaluate).

      "... the illusion of slight Karitiana-European affinity"...

      Should not be any illusion because the main Native American patrilineage, Q, clearly originates from somewhere in the Western Eurasian region (possibly Iran or somewhere nearby) and arrived to Beringia via Altai and Siberia (along with matrilineage X2, also original from West Asia). Some of that genetic legacy, even if the main genetic substrate is East Asian (matrilineages A, B, C, D and patrilineage C3, and obviously most of the autosomal data).

      Delete
    2. I don't see any African-East Asian affinity in the PND table: French are clearly "more African" than any Eastern population; Sardinians depending with which African population you compare with. Anyhow threading too thinly you are...

      Delete
  12. Scenario (B) also suffers from a lack of haploid DNA evidence (only Y-DNA haplogroup E and mtDNA haplogroup M1 spring to mind). Furthermore, Scenario (B) would require that ancestral "proto-Africans" (prior to the hypothetical East Asian admixture) be even more distinct from Denisovans than modern Africans are, which, actually, does appear to be the case if San and Mbuti are taken as being somewhat representative of "proto-Africans" and Yoruba, Mandenka, and Dinka are taken as being representative of "East Asian-admixed Africans." Also, if the East Asian-African affinity relative to American-African and Papuan-African comparisons were due to gene flow from East Asia to Africa, then why do Europeans also exhibit African affinity?

    In my opinion, the most likely explanation is (C): modern Africans and modern East Asians both have been influenced by a population that is no longer extant in pure form, but which has contributed to the formation not only of modern Africans and modern East Asians, but also to modern Europeans. The ones who lack detectable influence from this hypothetical population are ancient East Asians like Tianyuan, modern Native Americans like Karitiana, and modern Australasians like Papuans. There is an obvious candidate for such a population in the form of Y-DNA haplogroup DE, whose distribution is perfectly coincident. However, what is the mtDNA counterpart of Y-DNA haplogroup DE? What enabled this population to expand over (or influence) the northern branch ("proto-Americans") of the proto-Greater East Eurasians in East Asia, the proto-Europeans in West Eurasia, and the proto-Africans in Africa?

    Or is it not really another "mystery population," but just proto-West Eurasians who are responsible for this pattern? Did ancient West Eurasians colonize both Africa and East Asia but fail to reach America or Australasia? In this case, the European-Karitiana affinity would have to be explained by Denisovan admixture in East Asians and Papuans pulling them away from Europeans and Karitiana. Is there any logical reason for opposing this? (I mean a truly logical reason, not appeals to the ghost of White racism.)

    ReplyDelete
  13. I just recalled that the Europeans are closer to Tianyuan than the Papuans are to Tianyuan despite the sharing of Denisovan affinity between Tianyuan and Papuans. This must mean that modern Europeans contain an element related to the northern branch of the proto-Greater East Eurasian population; this is where the recent claims of admixture between a "proto-American" population and an "ancient West Eurasian aborigine" population lying at the origin of modern Europeans (or modern West Asians and Europeans in general) come into play. Maybe Y-DNA haplogroup DE and the aforementioned patterns of autosomal affinity reflect "ancient West Eurasian aborigine" influence on East Asians and Africans to the exclusion of Americans and Australasians?

    ReplyDelete
  14. Maju wrote,

    "I don't see any African-East Asian affinity in the PND table: French are clearly "more African" than any Eastern population; Sardinians depending with which African population you compare with. Anyhow threading too thinly you are..."

    Here you go again, conveniently ignoring the facts when they don't suit your pet theory. I really should not need to point this out to you so explicitly:

    Dinka-French 27,135
    Dinka-Sardinian 27,245
    Dinka-Yoruba 27,624
    Dinka-Dai 27,704
    Dinka-Mandenka 27,906
    Dinka-Han 27,923
    Dinka-Karitiana 28,187
    Dinka-Papuan 28,931
    Dinka-Tianyuan 30,297
    Dinka-Denisovan 46,208

    The Dinka vs. French and Dinka vs. Sardinian comparisons produce fewer PNDs than even the Dinka. vs. other African comparisons. Even the Dinka-Dai (Thai-speaking population from Yunnan Province of southwestern PRC) comparison produces fewer PNDs than the Dinka-Mandenka comparison. Does this not suggest gene flow from some ancestors of Europeans into several already largely differentiated proto-African populations? Is there any other way to explain this pattern?

    Mandenka-Yoruba 26,684
    Mandenka-Sardinian 27,331
    Mandenka-Dinka 27,906
    Mandenka-Han 27,981
    Mandenka-French 28,063
    Mandenka-Dai 28,332
    Mandenka-Papuan 28,871
    Mandenka-Karitiana 29,077
    Mandenka-Tianyuan 30,333
    Mandenka-Denisovan 45,850

    The Mandenka-European and Mandenka-Asian comparisons produce values that are interspersed, whereas the Mandenka-Papuan, Mandenka-Karitiana, and Mandenka-Tianyuan comparisons all produce high PND values that cluster toward the bottom of the list.

    Yoruba-Mandenka 26,684
    Yoruba-French 27,559
    Yoruba-Dinka 27,624
    Yoruba-Sardinian 27,901
    Yoruba-Dai 28,092
    Yoruba-Han 28,633
    Yoruba-Karitiana 28,917
    Yoruba-Papuan 29,023
    Yoruba-Tianyuan 31,059
    Yoruba-Denisovan 45,876

    Both the Dinka and the Yoruba exhibit the same pattern of affinity for extra-African populations: Dinka/Yoruba-vs.-French, followed by Dinka/Yoruba-vs.-Sardinian, followed by Dinka/Yoruba-vs.-Dai, followed by Dinka/Yoruba-vs.-Han, followed by Dinka/Yoruba-vs.-Karitiana, followed by Dinka/Yoruba-vs.-Papuan, followed by Dinka/Yoruba-vs.-Tianyuan, followed by Dinka/Yoruba-vs.-Denisovan.

    The pattern of affinities for the Mandenka sample is slightly different: Mandenka-vs.-Sardinian (rather than French), followed by Mandenka-vs.-Han (Han closer than Dai in this case), followed by Mandenka-vs.-French, followed by Mandenka-vs.-Dai, followed by Mandenka-vs.-Papuan, followed by Mandenka-vs.-Karitiana, followed by Mandenka-vs.-Tianyuan, followed by Mandenka-vs.-Denisovan.

    On average, there seems to be a pattern by which the European populations exhibit the most affinity for the African populations, the East Asian populations exhibit the second most affinity for the African populations, the American Karitiana population (a non-Denisovan admixed branch of Northern Greater East Eurasians) exhibits the third most affinity for the African populations, the Papuan population (a Denisovan-admixed branch of Southern Greater East Eurasians) exhibits the fourth most affinity for the African populations, the Tianyuan sample (ancient proto-East Asian, or a Denisovan-admixed branch of Northern Greater East Eurasians) exhibits the fifth most affinity for the African populations, and the Denisovan sample exhibits by far the least affinity for the African populations. Considering the great difference in patterning between the Tianyuan (ancient East Asian vs. modern African comparisons and the modern East Asian vs. modern African comparisons, it must be supposed that East Asians have received a great deal of gene flow from some population that shares affinity with modern Africans at some time during the past 40,000 years.

    ReplyDelete
    Replies
    1. "The Dinka vs. French and Dinka vs. Sardinian comparisons produce fewer PNDs than even the Dinka. vs. other African comparisons".

      On one side we must understand that the Africa-Other distances should not be larger than intra-African distances because Africa hosts the parent diversity of all Humankind. Dinka being as distant from, say, Yoruba as from French is just normal to my understanding, especially because the Dinka, who live in the Nile Basin, may belong to the pre-OoA population by many ancestral branches, while the Yoruba no or by many less (let's not forget this is autosomal genetics and therefore ancestral branches are more like a hyper-complicated web than any kind of tree-like structure).

      On the other side, most of the African genetic influence into Europe should come from the are where the Dinkas inhabit (Nile) and not that of Yorubas. And, vice versa, the area of the Eastern Sudan was clearly more affected by Eurasian backflow than that of West Africa, at least if we exclude some zones of the Sahel (Fulani, Chadic peoples).

      "Both the Dinka and the Yoruba exhibit the same pattern of affinity for extra-African populations"...

      I have to admit to that. I previously confused the Sardinian with the adjacent Papuan comparison. I don't see how it matters however.

      "On average, there seems to be a pattern by which the European populations exhibit the most affinity for the African populations, the East Asian populations exhibit the second most affinity for the African populations, the American Karitiana population (...) exhibits the third most affinity for the African populations, the Papuan population (...) exhibits the fourth most affinity for the African populations, the Tianyuan sample (ancient proto-East Asian, or a Denisovan-admixed branch of Northern Greater East Eurasians) exhibits the fifth most affinity for the African populations, and the Denisovan sample exhibits by far the least affinity for the African populations".

      It may well be but you are still threading very thing and the result, unless confirmed somehow by other means are probably of no statistical significance.

      Does that happen also with the San? Your sequence breaks apart. San affinity outside Africa per the PDN table 1:

      1. Han 34.0 Kb
      2. French 34.1 Kb
      3. Sardinian 34.1 Kb
      4. Dai 34.2 Kb
      5. Karitiana 34.6 Kb
      6. Papuan 34.7 Kb
      7. Tianyuan 35.9 Kb

      Distances to other Africans are in the 33.2-33.8 range (suggesting some admixture since the OoA), distance to Denisovan is 47.5. This may be caused by issues like similar phenotype (both have epicanthic fold and all that, maybe caused by a founder effect) but whatever the case, it blurs your argument quite a bit.

      ...

      Delete
    2. ...

      "... it must be supposed that East Asians have received a great deal of gene flow from some population that shares affinity with modern Africans at some time during the past 40,000 years".

      The percentage of differential apparent African affinity between Tianyuan and modern Han looks indeed significant on first sight (5-8%) but how can you be so sure that this is not a fluke caused by individual variation (at least 80% of the variance is individual, which if subtracted makes the percentages just 1-2%) or the extremely poor quality of the Tianyuan sequence (only 3% could be sequenced directly, what induces huge error margins).

      I think that you should leave Tianyuan out from the equation and compare only modern populations. Then the difference between Han and Papuans respectively vs. Africans is much lesser: 2% vs San, 1.3% vs Yoruba.

      Assuming this is not just a fluke, re. Papuans and Native Americans there can be issues of isolation but also of initial bottlencking (founder effects). Certainly it suggests that there was no back-migration from Melanesia as you were arguing in a recent discussion.

      So let's assume for the sake of the discussion that all Eurasians (i.e. excluding Australasian and American aborigines) have an appearance of 1-1.5% greater African admixture as it could seem from this PND table. Can't this be just a fluke? Do we have other data supporting it? Contradicting it?

      Delete
  15. Maju wrote,

    "Should not be any illusion because the main Native American patrilineage, Q, clearly originates from somewhere in the Western Eurasian region (possibly Iran or somewhere nearby) and arrived to Beringia via Altai and Siberia (along with matrilineage X2, also original from West Asia). Some of that genetic legacy, even if the main genetic substrate is East Asian (matrilineages A, B, C, D and patrilineage C3, and obviously most of the autosomal data)."

    I am immune to your delusions of a West Eurasian origin for haplogroup Q (let alone haplogroup R), whose origin is deeply rooted in the eastern half of Eurasia (Y-DNA haplogroup MNOPS-M526/PF5979), and whose major representatives among extant populations, the indigenous populations of the Americas, are quite literally the anthropological textbook definition of "East Eurasians" (taken as a euphemism for "Mongoloid").

    Even academic geneticists finally have accepted that they cannot ignore the East Eurasian origin of Y-DNA haplogroup R any longer (cf. the recent publication of several papers explicitly indicating proto-American admixture in Europeans, or in modern West Eurasians in general).

    ReplyDelete
    Replies
    1. Isn't P, Q and R greatest and almost only real variance in South and West Asia? I don't care what you're immune about of what makes you cry, I care about data.

      "cf. the recent publication of several papers explicitly indicating proto-American admixture in Europeans, or in modern West Eurasians in general".

      You mean this junk?

      I care about what the field data says about diversity at the highest phylogenetic levels re. Q, R and P (and any other well researched lineage). I care about archaeological data showing how "mode 4" technology clearly expanded from Altai to NE Asia in which is probably the clearest indirect archaeological evidence behind the proto-Amerindian expansion through Siberia, not yet directly mapped. Tianyuan B people used flake industries (MP type, it's debatable why but is a constant East of Bengal) but some millennia later, since c. 30 Ka ago "mode 4" blade industries show a clear track Eastwards through Mongolia and Northern China, later appearing further South in SE Asia.

      Q and all the rest must be integrated in a generally valid continental or global understanding of population flows. This understanding must be founded in a holistic (multi-angle) approach via mtDNA, Y-DNA and archaeology. I don't know the answers for every single strange detail or ill-understood data bit but I do have managed along the years to understand better and better how the process of Eurasia+ colonization actually must have ensued. And in the case of Q it must imply a backflow SE Asia (MNOPS) → South Asia (P) → West/Central Asia (Q) →→ Siberia and Native Americans. This process is also apparent in subclades of N (incl. R), although it suggest rather dynamic and sometimes bidirectional South Asia ←→ SE Asia flows in the early times of the Eurasian+ colonization.

      I can only feel that you're being horribly Sinocentric here.

      Delete
  16. Regarding Onur's racial commments - we know that both mtDNA and y-DNA uniparental haplogroups are deeply rooted in Africa. You have to question either the motives or the sanity of anyone who argues otherwise.

    Africa's racial history is not as simple as it looks (based on modern-day populations). For instance, East Africa (the candidate region of origin of Y-DNA Hg E in Africa) Negroidized relatively recently.

    ReplyDelete
    Replies
    1. Y-DNA E is hypothesized to have originated in East Africa based on it being a subclade of proto-Eurasian CT, and perhaps its likely relation to the spread of mtDNA L3. But the diversity of E is actually greatest in West-Central Africa. On the other hand, East Africa harbors the greatest diversity of E1b1, which is dominant in most Niger-Congo populations. In Ethiopia, Y-DNA E1b1 peaks in isolated Omotic speakers from Southwestern Ethiopia, who are less Eurasian-admixed than any other Ethiopians (excluding Nilo-Saharans), and also carry the most divergent subclades of Y-DNA E1b1. The autosomal evidence is quite clear that Omotics are descended from the oldest inhabitants of Ethiopia, with some minor Eurasian admixture. They were later dominated by Cushites migrating from Northeast Sudan in the past 10,000 years, who were a mixture between West Eurasians and a Sudanese-like population.

      The delusions of some that a possible non-African origin of Y-DNA DE would have come from a population even remotely close to modern West Eurasians, either genetically or phenotypically, is a big joke. DE must have coalesced in a population that was closely related to the early L3 carriers who migrated to Eurasia, and would have no particular relation to modern West Eurasians who descend from Y-DNA F and mtDNA N carriers. Another joke is that a possible Eurasian origin of DE would have any impact on the much more recent exit of Y-DNA E1b1b1 into Eurasia. The lengths to which some will go to deny ancestral connections with inhabitants of the African continent are truly amazing, albeit misguided.

      Delete
    2. You are making inferences based on modern-day populations and disregarding the ancient modern human specimens.

      Delete
    3. What ancient specimens? I'm not going to base my whole impression of African population history based on some supposed morphological similarities between ancient East Africans and West Eurasians.

      The Hofmeyr South African specimen is often used for this kind of rhetoric. As it turns out, Hofmeyr shows some similarities to Upper Palaeolithic Europeans, and is dissimilar to modern Africans. As there has been great morphological change in human populations all over the world in just the past 20,000 years, this is no great surprise. Due to the lacking African archaeological record, Hofmeyr can't be compared to contemporaneous and slightly earlier populations from Eastern-Southern Africa. Unless an intrusive population is clearly identified in the archaeological record, there is no way to conclude that Hofmeyr was descended from Eurasian migrants.

      Disregarding all of the indications to the opposite and acknowledging the unlikely possibility that Hofmeyr is actually descended from Eurasian migrants... What exactly would be the relation to modern West Eurasians?? Hofmeyr lived in the far south of Africa during a time frame when we even find Neanderthal settlements in the Levant.

      Delete
    4. This comment has been removed by the author.

      Delete
    5. Regarding the fossil record, read the book below. The archaeological record does imply migration from West Asia into Africa during the transition to the Late Stone Age in Africa. Add to these the concurrent spread of Y-DNA Hg E in Africa, and everything begins to make better sense.

      Delete
    6. Y-DNA E has all basal variance in Africa. Not just that most DE* has been found in (West) Africa as well. E is African by origin no matter how you look at it: nothing of that exists in West Asia or Europe. If you look at mtDNA (some people always seem to want to ignore the most reliable marker: mtDNA) there's no way that there was any population going back to Africa from West Asia, except restricted to some areas of North and East Africa. This last one is easily identifiable by West Eurasian Y-DNA lineages (J, R, etc.) and mtDNA ones. However, as there was a thing known as Afroasiatic expansion, in many of these areas E subclades mostly replaced F ones. You can only understand all this comparing the various Y-DNA but also mtDNA layers (autosomal and what-not also), and not just looking at one side of the issue.

      The archaeological record does suggest cultural flows in various directions but they do not necessarily mean large migrations: when you see your neighbor using a better technique, you copy it, if not you because you're too old to innovate anymore, your son or daughter will no doubt. We do not see genetic evidence of migrations related to the LSA except surely in some buffer areas like North Africa, The Horn, parts of the Sahel (these last may well be Neolithic flows in fact).

      Similarly we see 'mode 4' flowing to East Asia from Altai via Mongolia and North China but that does not mean that West Eurasian lineages reached out in that direction in great numbers with the already mentioned exception of Amerindian Q. How do we know that? Because of Tianyuan mtDNA quite remarkably. Nobody can come now and say that mtDNA N/R expanded to East Asia in the UP because we have already found one individual with B in Beijing and can show the middle finger to all those Eurocentric speculators.

      And similarly we see today's technological advances like computers and mobile phones flowing (quite faster) around the world without any specific related migrations, just some loosening of ethno-political borders.

      Why should I accept that, with no apparent genetic signature (mtDNA please, it's safer), there was a massive flow of people into Africa but not into East Asia or nowadays with techno-globalization around the world? I will not.

      Now please, Onur, move on because you are hitchhiking the thread and you were already once forced out of this blog, what is always a bother. You are abusive by not respecting the directives of the owner of the blog (moi). It's like if you get into my home and I invite you to leave and you don't and I insist (when I should not need to) and you insist on staying on any pretext, so you're really forcing me to use violence and everything getting very ugly.

      For the last time: move to some place where your annoying and unlikely racialism and Eurocentrism is welcome.

      Delete
    7. Respecto a Hofmeyr skull, we can see in this interesting thread at Mathilda's Anthropology Blog (now in indefinite hiatus but holding a good deal of materials on African and other anthropology) how its classification overlaps several racial/continental categories, falling in graph 1 between WEA and AFR (out of modern SAN however), and in graph 2 overlapping all three pops: WEA, AFR and SAN and also the range of European UP Skulls, which are a bit anomalous re modern humans by that measure method.

      Of course all these are mere approximations but we see nothing in Hofmeyr skull to declare it an intruder from the far North. Similarly other skulls from UP Europe back in the day were claimed to be "Negroid", "Eskimoid" and what not, claims that have been dramatically turned down as time pass and more and more people look at the matter.

      Delete
    8. Earlier suggestions that "mainstream" (Cushitic, Semitic) Ethiopians are descendants of the oldest inhabitants of the region have now been proven wrong by genetics. Omotic speakers are not descendants of some recent intrusive population into the region, which was previously dominated by Cushitic "Hamites". It is actually the other way around. This is evident both in the autosomal evidence and the data on haploid DNA provided by Plaster's study on Ethiopian Y-DNA/mtDNA.

      Despite the heavy Eurasian mtDNA M/N contribution in Ethiopian Cushites and Semites (belonging to L3), the Omotics sampled by Plaster actually had the highest frequency of mtDNA L3 in Ethiopia, and likely all of Sub-Saharan Africa. Along with the highest frequencies of Y-DNA E in Ethiopia. They also have low to nonexistent levels of Y-DNA A3b2, which is significant both in Sudanese and Cushitic/Semitic Ethiopians. In addition, the rare A3b2 found in Maale Omotics from Plaster's study belonged to the A3b2* paragroup, whereas in all of the Cushitic/Semitic samples and most of the Sudanese, the more recently derived A3b2b was found.

      The genetic evidence clearly indicates that Omotics are the most ancient inhabitants of Ethiopia. Farther south, the Hadza are probably a better candidate, and there were probably some remnant Khoisan/Hadza-like peoples in Ethiopia as well.

      Delete
  17. This comment has been removed by the author.

    ReplyDelete
  18. On the recency of the Negroidization of East Africa, read:

    http://www.amazon.co.uk/Whos-Who-Skulls-Identification-Measurements/dp/0873652096

    ReplyDelete
  19. "To me, this suggests a two-prong colonization of 'Greater East Eurasia' (including America and Australasia), with an initial split between a 'Northern' (proto-Karitiana/Tianyuan) branch and a 'Southern' branch (proto-Papuan)'

    Similar to the situation I support. Basically it seems to me likely that Y-DNA C and mt-DNA N carried the Denisova element along the northern path through eurasia and then south to Australia/New Guinea. The southern branch was Denisova-free and is represented by the expansion of Y-DNA F and mt-DNA M. Of course subsequent population movement and mixture has shuffled the mixtures. For example New Guinea has very little of either Y-DNA C or mt-DNA N (apart from P) so must have received the Denisova element before they managed to reach New Guinea.

    "On one side we must understand that the Africa-Other distances should not be larger than intra-African distances because Africa hosts the parent diversity of all Humankind".

    But the situation has not been static since the OoA, either in Africa or outside it. And humans outside Africa have become mixed with older OoA populations. The same mixing with older populations now appears to be true within Africa too.

    "Isn't P, Q and R greatest and almost only real variance in South and West Asia?"

    I'm with Maju here, 'a backflow SE Asia (MNOPS) → South Asia (P) → West/Central Asia (Q) →→ Siberia and Native Americans'. To me too it looks most likely that although P originated in SE Asia its diversity began as it moved west through South Asia before emerging onto the Eurasian steppe.

    "Tianyuan B people used flake industries (MP type, it's debatable why but is a constant East of Bengal)"

    Here you disappoint me. I would have thought you'd understand why by now.

    "but some millennia later, since c. 30 Ka ago 'mode 4' blade industries show a clear track Eastwards through Mongolia and Northern China, later appearing further South in SE Asia".

    It seems at least as though you are now prepared to accept that these people carrying mode 4 industries were not the first 'modern humans' into East Asia.

    ReplyDelete
  20. "But the diversity of E is actually greatest in West-Central Africa".

    I agree here too. E almost has to be West African in origin, presumably along the northern forest/savannah margin of the Congo Basin.

    "The delusions of some that a possible non-African origin of Y-DNA DE would have come from a population even remotely close to modern West Eurasians, either genetically or phenotypically, is a big joke".

    I disagree with that comment. As I said before: 'the situation has not been static since the OoA, either in Africa or outside it'. Almost anything is possible, and D certainly did not originate in Africa. My guess is that DE had spread through a large proportion of northern Africa/Eurasia before the onset of the ice age and increased aridity separated the two populations with E surviving in Africa and D in East Asia.

    "DE must have coalesced in a population that was closely related to the early L3 carriers who migrated to Eurasia, and would have no particular relation to modern West Eurasians who descend from Y-DNA F and mtDNA N carriers".

    Why have you chosen to ignore mt-DNA M and Y-DNA C?

    "The archaeological record does suggest cultural flows in various directions but they do not necessarily mean large migrations"

    But probably do represent haplogroup movement. A man carrying a particular Y-DNA who had children with a local woman would pass on his Y-haplogroup but only half his aDNA. The process would be repeated progressively with his children and grandchildren, and so on. I believ you over-estimate the ability of populations to adopt new techniques without the adoption of at least some newly arrived individuals. That explains, 'with the already mentioned exception of Amerindian Q'.

    "similarly we see today's technological advances like computers and mobile phones flowing (quite faster) around the world without any specific related migrations, just some loosening of ethno-political borders".

    We are in a totally different environment today though. We basically have a single culture throughout the world with the exception of some people who hold onto primitive religious beliefs and continue to be a threat to the rest of us.

    ReplyDelete
  21. Since I am banned, this will be my last post, and as this is my last post, I'll be brief.

    As the craniometric analyses show, the Hofmeyr Skull is closest to Cro-Magnons. What is the racial affiliation of Cro-Magnons? They are Caucasoids in a general sense. If you don't believe me on that issue, see:

    http://racialreality.blogspot.com/2011/02/cro-magnons-were-caucasoid-not-negroid.html

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

    The Caucasoid racial affiliation of Cro-Magnons (and consequently also the Hofmeyr Skull) is also in congruence with the early divergence between Caucasoids and Mongoloids according to the results of Fu et al. 2013.

    Agur!

    ReplyDelete

Please, be reasonably respectful when making comments. I do not tolerate in particular sexism, racism nor homophobia. Personal attacks, manipulation and trolling are also very much unwelcome here.The author reserves the right to delete any abusive comment.

Preliminary comment moderation is... ON (sorry, too many trolls).