A doctoral thesis on ancient Ukrainian mtDNA has recently become freely available (h/t Kristiina):
Jeremy R. Newton, Ancient Mitochondrial DNA From Pre-historic Southeastern Europe: The Presence of East Eurasian Haplogroups Provides Evidence of Interactions with South Siberians Across the Central Asian Steppe Belt. Grand Valley State University (thesis), 2011. Freely available → LINK
The key element of this study is table 1:
Location of sites (fig. 3):
Notice that the "Kurgan sites" (D1.8, L8 and L15) are not from the first Kurgan arrivals but rather from a late layer, surely Srubna culture, which is generally believed to be proto-Cimmerian.
The most striking element probably is the presence of relatively high frequencies of mtDNA C since Neolithic times. However this is not inconsistent with previous findings (Desarkissian 2011) of mtDNA C (C1) among NE European Epipaleolithic hunter-gatherers, surely precursors of modern Finnic peoples. It means that the Siberian element of East Asian affinity today best preserved among Uralic peoples, was present in Europe before Neolithic and that it had an impact (21%) even in presumably non-Uralic populations such as Epigravettian derived Dniepr-Don.
This in turn may well explain the subtle Siberian affinity elements sometimes apparent in much of Northern and Central Europe, because these Eastern European peoples made in turn significant demic impacts in those areas, first with the Pitted Ware culture (clearly derived from Dniepr-Don: similar pottery and burial styles) that affected parts of the Southern Baltic, via Belarus, and later with the Kurgan waves of Indoeuropean-speaking invaders.
Maybe a bit more intriguing is the coincidence of C4a lineages in all the three kurgans of SW Ukraine. It may be just a coincidence or a very specific ethnic provenance of the princesses of that sub-group but the thesis argues for these being direct descendants of the Neolithic C4a lineage found in Ya34. I must say I am skeptic but it is not totally impossible. If real, it would imply that all C4a3 and C4a6 haplogroups (at least) are of Eastern European coalescence, what I find a bit difficult to accept, to say the least - but who knows?
An element in favor of such model is that neither of these C sublineages seems to be present in West Siberian ancient mtDNA, while no Oriental lineages altogether have been found in Central Asia before the Iron Age.
See also:
I compared the DnieperDonetz C haplotypes with other ancient specimens and found out that ”DD ya 45” shares the same mutations with Baraba Steppe C (6-3 kya), Karelian Bol’shoy C (3.5 kya), Pazyryk C (Iron Age), Baikal Lokomotiv C (7-6 kya) and Ust’-Ida C (4-5 kya).
ReplyDeleteI also found out that between Dnieper-Donetz U5a and Baraba Stariy Sad U5a there is a 1 step difference. Dnieper-Donetz U3 ( Ya 19) is also found in Baraba Chicha and in Sampula and Daheyan in Xinjiang. Haplogroup H is wide spread in Eurasia, but I could find only one match with Dnieper-Donetz H haplotypes: Dnieper-Donetz H (DD33) is found also in Gawaerk in Xinjiang.
The DD ya 45 haplotype C we find in Dnieper Donetz is the same as the Baraba Steppe Ust Tartas basic haplotype which has 3 mutations and is the most frequent C haplotype. Other Baraba C haplotypes have 1 step differences to Dnieper Donetz C haplotype.
In the early Baraba Steppe culture (2000-4000 BC) hg C is the most frequent haplogroup: 16 out of 52 specimens turned out to be C. There is also quite a lot of diversity, as there are six different haplotypes. Xinjiang Xiaohe C4 haplotype shows 1 step, 2 step or 3 step differences to Dnieper Donetz C. It is very interesting that in Dnieper Donetz there seems to be a lot of diversity in Hg C, as all 6 specimens are different on the basis of HVR1. For example in Xinjiang Xiaohe, we have even 14 C specimens and only two haplotypes (on the basis of HVR1). I would say that it cannot be a coincidence that we have in Dnieper Donetz and Xinjiang Xiaohe a lot of C4. In Xiaohe, ydna was mostly R1a, and we would expect to find R1a also in Dnieper Donetz.
I have my own ideas on the origin of MtDNA C. As this haplogroup is very old in Altai, possibly 30000 years old, it may not be only of Uralic origin. It could have moved already with NO or Ydna Q.
It is also interesting that in Bronze Age kurgans of Central Asia there is no haplogroup C found yet. However, haplogroups T, H and U5a are present.
Bronze Age kurgans: 3xT, 2xU5a, 1xH, 1xHV, 1xI
Late Bronze Age kurgans: 3xH, 1xU1, 1xW
What strikes me most in Dnieper Donetz is the paucity of haplogroup U: one U1 and one U5a. You say that Pitted Ware culture is clearly derived from Dniepr-Don Culture, but by comparing their MtDNA, they have nothing in common. Does this mean that the similarities are not a result of migrations but cultural diffusion.
I added to your previous summary the haplogroup frequencies of the Dnieper Donetz and Swedish ancient MtDNA data.
(PWC=Pitted Ware Culture, TRB=Funnel Beaker Culture, DD= Dnieper-Donetz)
→ H: BB 87%, Unetice 0%, Dnieper Donetz 35%, Swedish PWC 0%, Swedish TRB 33%
→ U(xK): BB: 0%, Unetice: 46%, Dnieper Donetz 12%, Swedish PWC 74%, Swedish TRB 0%
→ J: BB: 13%, Unetice 0%, Dnieper Donetz 0%, Swedish PWC 5%, Swedish TRB 33%
→ T: BB: 0%, Unetice: 31%, Dnieper Donetz 18%, Swedish PWC 5%, Swedish TRB 33%
→ N1: BB: 0%, Unetice 23%, Dnieper Donetz 0%, Swedish PWC 0% (?), Swedish TRB 0%
→ W: BB: 13%, Unetice: 0%, Dnieper Donetz 0%, Swedish PWC 0%, Swedish TRB 0%
→ X: BB: 13%, Unetice: 0%, Dnieper Donetz 0%, Swedish PWC 0%, Swedish TRB 0%
→ C: BB: 0%, Unetice: 0%, Dnieper Donetz 35%, Swedish PWC 0%, Swedish TRB 0%
"I compared the DnieperDonetz C haplotypes with other ancient specimens and found out that ”DD ya 45” shares the same mutations with Baraba Steppe C (6-3 kya)"...
DeleteHow come. A cursory look at Molodin's haplotype tree shows that, while the do share the transitions at 298 and 327 (markers for C in general), the other transitions do not match. I see no 061, 357, 278, 218, 327, etc. in Molodin's data. Can you be more precise? My first impression is that the only correlation is that they are all C.
"The DD ya 45 haplotype C we find in Dnieper Donetz is the same as the Baraba Steppe Ust Tartas basic haplotype"...
Ya45 is underived C-root!!! Can there be anything more ambiguous? I hope you are not imagining all C coalescing in West Eurasia, because I do not think that's acceptable at all.
"I have my own ideas on the origin of MtDNA C. As this haplogroup is very old in Altai"...
Reference, please. AFAIK there is no mtDNA C in Altai at least until the Iron Age. Western lineages in most of it and D only in the Mongolian part.
... "possibly 30000 years old"...
Probably not at all. Not just it has never been sequenced but C, as CZ and all M8, should originate in NE Asia (Manchuria or whatever) and only migrate westwards since the proto-Uralic flows through Siberia (but North of Altai) associated to Y-DNA N1, which are at least post-LGM. Previously there does not seem to be any East-to-West flow via the North.
"I have my own ideas on the origin of MtDNA C."
I see you do. What worries me is how coherent they are with the data.
"It is also interesting that in Bronze Age kurgans of Central Asia there is no haplogroup C found yet."
I agree that this is pretty curious and may tell us of a peculiar genetic pool of Samara Valley PIE (and later Indo-Iranian and Tocharian) peoples, less related to proto-Uralic flows than DD apparently.
"You say that Pitted Ware culture is clearly derived from Dniepr-Don Culture, but by comparing their MtDNA, they have nothing in common".
Some T and U5 but you are right in the general picture, at least for the only documented case, which is in the far edge of Pitted Ware expansion: Götland. But we know almost nothing of the genetic make up of their precursors in Belarus, Lithuania, etc.
Whatever the case pottery and burial practices are very similar and AFAIK they are very directly related in the cultural aspect. Maybe, as DD men (with Y-DNA I1?) moved in NW direction, they took local wives, what would distort the picture we get from mtDNA very much indeed.
"your previous summary the haplogroup frequencies"...
Notice please that I had an error with German Bell Beaker data because the L(xR) (or N(xR)) had not been properly tabbed, so it added up to more than 100%. Both J and H are somewhat less than I reported then.
It seems to me that in Unetice area, Scandinavia and Finland there is plenty of U5a, U5a1, U4. These haplogroups extend to Russia and in Siberia as far East as Lake Baikal. Haplogroup H seems to be frequent in Southern Russia and in BB Culture. I wonder if haplogroup T came from somewhere in Iran and spread, on the one hand, to Central Asia and Volga area and, on the other, through Russia to Northern and Western Europe. Karelian C1 is a mystery and I would really like to know the Ydna that accompanied this haplogroup in Scandinavia.
ReplyDeleteIt seems to me that we have:
LBK: K, J, N1, T (arriving from Caucasus area?)
Scandinavia, Finland: U5a, U4, V, H2, C1
Bell Beaker: H, J, W, X
Unetice: U5a, U5b, T, N1, I
Dnieper Donetz: C, H, T
Baraba Steppe: U5a, U4, U2e, C, Z, D, A
If haplogroups H and U are the oldest in Europe, their distribution is different. It seems that the North is dominated by U5 and U4 and H is more wide-spread in the centre, southeast and southwest (?). If Dnieper Donetz kurgan people spoke an Indo-European language, it is surprising that they do not harbour mtDNA U but instead haplogroup H and C. However, the Central Asian kurgan people probably also spoke an Indo-European language and their mtDNA is somewhat different. This only shows that genes are much older than languages and people change languages quite easily.
NB “Nineteen Pitted Ware samples from Gotland were dominated by mtDNA haplogroups U4 (x4) (or H1b), U5a (x2), U5 (x1), V(x1) though it should be noted that, because of the low resolution of the tests performed, some haplotypes reported as U4 may actually belong to haplogroup H[1b].“
However, I must note that the frequency of H1b is low in Sweden 1.3% and somewhat higher in Karelia (3.3%) and Lithuania (5.5%), and so I would bet that PWC haplotype is not H1 but U4.
"It seems to me that in Unetice area, Scandinavia and Finland there is plenty of U5a, U5a1, U4. These haplogroups extend to Russia and in Siberia as far East as Lake Baikal".
DeleteTrue. And they seem to correlate well with the pre-Neolithic mtDNA pool of Lithuania and NE Poland. So at least in the female lines there seems to be continuity in the Southern Baltic (not sure how they reached Lake Baikal however, must have been via Samara Valley, i.e. the PIE urheimat).
"LBK: K, J, N1, T (arriving from Caucasus area?)"
More like West Asia in general (at least K is very common in Syrian PPNB) or even the Balcans themselves. No reason to think the Caucasus as origin of pretty much anything at all (but rather a refuge area).
"If haplogroups H and U are the oldest in Europe, their distribution is different. It seems that the North is dominated by U5 and U4 and H is more wide-spread in the centre, southeast and southwest (?)".
AFAIK in Paleolithic Eastern Europe H only shows up in the NE so far: Sunghir (Gravettian) and Karelia (Epipaleolithic). Only later they begin to show up also in the South - maybe just a random sampling fluke? The fact is that we have only very limited Paleolithic aDNA and that patterns are not too obvious.
"However, the Central Asian kurgan people probably also spoke an Indo-European language and their mtDNA is somewhat different. This only shows that genes are much older than languages and people change languages quite easily".
While I do not disagree with the argument of people changing languages with relative ease, let's be cautious because it's very plausible that in many cases it were men who carried the languages and what we are looking at here is the female genetic heritage only.
... "because of the low resolution of the tests performed, some haplotypes reported as U4 may actually belong to haplogroup H[1b]".
That would be interesting and I would say that even quite plausible.
"However, I must note that the frequency of H1b is low in Sweden 1.3% and somewhat higher in Karelia (3.3%) and Lithuania (5.5%), and so I would bet that PWC haplotype is not H1 but U4".
Have you even checked the modern frequencies of U4? For what I know they are very low, especially West of the Baltic: c. 1-2%. East of the Baltic they may be a bit higher, especially in Russia and Azerbaijan → http://u4haplogroup.blogspot.com/2009/10/u4-distribution-map.html.
So no particular reason to think either way.
When I look at Table 4 of the thesis, I see that Ya45 has 3 mutations: 16223T, 16298C, 16327T. In Baraba paper the most common C haplotype shows the mutations 223T-298C-327T and is found in the following specimens: Ut4, Ut7, Ut16, Ut37, Pkr10, TK1,
ReplyDeleteTK2, TA9, TA13, TA14, TA16, Sts7. This ancestral type is present in Siberia and Eastern Europe. The other 5 Dnieper Donetz haplotypes do not match with Baraba C haplotypes as Baraba C haplotypes do not harbour the mutations you mentioned.
I copy here a piece from my previous text on your blog:
Derenko 2002 paper confirms a great variety of C and the existence of M8 and Z in the Altaians and other South-Siberian populations. They say that the coalescence time of all haplotypes to the root of haplogroup C was estimated as 38400 B.P. Then I found out that there is yet another Northern sister branch of C and Z in Udegey, Ulchi and Tofalar.
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&frm=1&source=web&cd=18&ved=0CH0QFjAHOAo&url=http%3A%2F%2Fwww.researchgate.net%2Fpublication%2F10595717_Diversity_of_mitochondrial_DNA_lineages_in_South_Siberia%2Ffile%2F79e415089bd2aed828.pdf&ei=dcuPUdSiEsK14ASjkYGYDA&usg=AFQjCNGSuhIQL-DkpzdTiYhA0o93VqejSg&sig2=9pMpNvfDIuIL0OHXbtjt0A
This ancestral C with mutations 16223T, 16298C, 16327T is present in Altai and Southern Siberia in the following frequencies (not in %):
Altaians 5
Khakassians 4
Buryats 2
Sojots 1
Todjins 4
Tuvinians 10
Tofalars 5
It is more common in the western part than the eastern part of the area.
You say that “AFAIK there is no mtDNA C in Altai at least until the Iron Age.”, but I would add that there are not really many Bronze Age samples from Altai. Haplogroup C has not been found in ancient Kazakhstan kurgans but it has been found in Baraba Steppe. In the oldest Krasnoyarsk samples (8) we have one Z1, and in the oldest Pazyryk, Altai samples we have 3xD out of three samples.
This Eva Fernández Domínguez’s ”package” contains really a lot of information.
http://www.tdx.cat/handle/10803/795
There are interesting findings:
In Southern Spain (Nerja, Pirulejo) we have L1b, J, H, and in Northern Spain (Abauntz, Tres Montes, Atxuri, Garai ) we have L2, H and J, and in Portugal (Toledo) L3a. Consequently, H seems to have arrived very early in Iberia. Also J seems to have arrived early in Western Europe (in one of Eva’s figures, p. 513 results, it looks like T is coming out of J). Here, the most striking thing is the strong presence of L1, L2 and L3 in Iberia.
According to Eupedia, ancient Basques harboured H, J, K, U, T/X, L2. H, J and L2 seem to have been there first, and I would bet that K and X arrived with ydna R1b. This is surely not true if R1b was present in Iberia already in the Mesolithic period.
"... Ya45 has 3 mutations: 16223T, 16298C, 16327T. In Baraba paper the most common C haplotype shows the mutations 223T-298C-327T"
DeleteThat means: CRS → (223) → (L3x)R → (298) → M8 → (327) → C
It is the basic C haplotype as described counting from the CRS (H2a)! There is no mystery to that: every C sequence must have that sequence (unless a mutation towards the CRS haplotype is also present, of course).
It means nothing but apparently underived C (yeah, that happens: mtDNA is that way because mutations accumulate only slowly and irregularly).
"You say that “AFAIK there is no mtDNA C in Altai at least until the Iron Age.”, but I would add that there are not really many Bronze Age samples from Altai".
Enough to question your claim. You're making outlandish claims about C being in Altai since very old, even since 30 Ka BP, without a single piece of evidence!!! In fact with ALL AVAILABLE EVIDENCE AGAINST IT. Not acceptable.
"... it has been found in Baraba Steppe".
Which is NOT ALTAI and does not have any apparent relation with it except via mtDNA D, as we discussed back in the day. The Barabba area is Siberia proper (even if you insist in calling it "steppe" it is actually taiga), while Altai must be considered part of Central Asia for all archaeological and anthropological purposes I can fathom.
"In the oldest Krasnoyarsk samples (8) we have one Z1"...
North of Altai, in Siberia proper.
"... and in the oldest Pazyryk, Altai samples we have 3xD out of three samples".
Only in the Mongolian part of Altai (the rest is 100% West Eurasian). It is still not C.
"H seems to have arrived very early in Iberia"...
DeleteYes.
"Also J seems to have arrived early in Western Europe"
The Nerja sequence is JT(x,J,T) as far as I can discern. I know of no other JT sequences in Paleolithic Europe.
"... in one of Eva’s figures, p. 513 results, it looks like T is coming out of J"...
Not sure if it has any relevance (it could be T coming out of Nerja's JT, after wrongly labeling it as "J"). If so that could be interesting but working with HVS-I sequences only is a total mess, so it can also be absolutely meaningless.
"According to Eupedia, ancient Basques harboured H, J, K, U, T/X, L2".
Eupedia is quite messy and blatantly nordocentric, much of what it says is plain nonsense. This must refer to Izagirre and De La Rúa 1999 (because of the "T/X" label, unique of that paper), which refers to peripheral Chalcolithic sites only, but I have never seen any L2 in ancient Basque samples (nor modern ones to be fair). Some L(xM,N), surely L3, seems to exist since Paleolithic in Portugal and surely other areas of West Iberia but AFAIK not in the Basque Country.
"I would bet that K and X arrived with ydna R1b".
Makes no sense to me. If there's any relation between Y-DNA R1b and some mtDNA it must be with H. K, X etc. are too rare lineages (especially among Basques) and must be related to Neolithic arrivals, what in Y-DNA seems to mean G2a, I2a and E1b-V13, at least in SW Europe (probably also J2b but not detected yet in aDNA).
"Only in the Mongolian part of Altai (the rest is 100% West Eurasian). It is still not C."
DeleteIt is true that in Krasnoyarsk the oldest haplogroups (1800-1400 BC) are the following:
U4, U2e, U5a1, K2b, Z1, T1, T4, H
Pazyryk samples are the only samples coming from Altai proper, and all Pazyryk Bronze Age samples are haplogroup D.
Pazyryk Iron Age samples include the following haplogroups:
2xC, 4xD, 1xA, 1xG2a, 3xK, 2xU5a1, 1xJ, 1Tx1, 1xHV6
Please check Table 2 http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0048904
This comment has been removed by the author.
DeleteThere are no "Pazyrik Bronze Age" samples because Pazyrik is an Iron Age culture. There are Mongolian Altai's Bronze Age samples (n=3, all D, labeled AMGBR). There are also four samples from the Republic of Altai, Russia from Neolithic and Bronze Age (n=4, labeled BRNRA), none of which is Oriental.
DeleteI mean: look at table S2. They are using also samples from previous studies not just the new ones sequenced for that paper.
DeleteWhatever the case there is no C in Altai before the Iron Age, neither in the Mongolian nor in the Russian side (which is what we normally call Altai).
Yes, we cannot say that C was in Altai during the Bronze Age, but we cannot either say that people in Altai were harbouring western haplogroups during the Bronze Age and earlier, if the only evidence are these Mongolian Bronze Age samples that are haplogroup D.
DeleteWhat is this "BRNRA- Rep. Altai, Neolithic and Bronze Age" study? Where did you get the results?
"we cannot either say that people in Altai were harbouring western haplogroups during the Bronze Age and earlier"
DeleteYes, we can say that because that's exactly the kind of data managed by the paper you just linked to. Check table S2 and follow the references please: n=4, all Western lineages for Bronze and Neolithic Republic of Altai!
"Where did you get the results?"
The results are reflected in the map but you will have to follow the bibliographic references to find the exact sources and details (I haven't because I trust that the authors did not cheat on this - it's you who seems to be claiming that they somehow altered the data, or whatever you mean because it seems to be a case of confusion on your side).
Table 3 says:
DeleteBronze Age, Mongolia (Altai), number of samples 3 100% East Eurasian
Neolithic and Bronze Age (Gorny Altai), number of samples 4 100% West Eurasian
Is it this? A paleogenetic study of the prehistoric populations of the Altai
http://link.springer.com/article/10.1134%2FS156301100704012
It is a pity that the article is not free!
They say that:
It addresses human remains from burials associated with various cultures which existed in Gorny Altai from Neolithic to Christian era. The findings reveal that craniometrically, prehistoric populations of Gorny Altai are mostly intermediate Mongoloid and Caucasoid population. It also adds that genetically, individuals from Neolithic and Bronze Age burials display only Western Eurasian mitochondrial DNA (mtDNA) haplotypes in Gorny Altai which attest the possibility of a considerable agreement between morphological and mtDNA data in terms of Mongoloid versus Caucasoid affinities.
There is also this study: http://128.220.160.141/login?auth=0&type=summary&url=/journals/human_biology/v076/76.1ricaut.pdf
According to Gonzalez paper, there is one haplogroup D in Republic of Altai from Pre Iron period.
"If there's any relation between Y-DNA R1b and some mtDNA it must be with H. K, X etc. are too rare lineages (especially among Basques) and must be related to Neolithic arrivals,"
DeleteYes, indeed! I know there is a big disagreement on the age of R1b in Iberia. If it came first, it came with H. If it came during the Neolithic, it probably came with K and X.
However, if it is true that L was present in Iberia in very early date, as it seems it was, I would think that part of it came with E1b-V13. Of course, L may have in part originated also in North Africa.
The reference is #20:
DeleteChikisheva TA, Gubina MA, Kulikov IV, Karafet TM, Voevoda MI, et al. (2007) A paleogenetic study of the prehistoric populations of the Altai. Archaeol Ethnol Anthropol Eurasia 4: 130–142. doi: 10.1134/S156301100704012.
"It is a pity that the article is not free!"
True but you can still check the haplotypes in fig. 3, what looks like 2 H (or similar near CRS haplotype), 1 U5 and 1 K.
"According to Gonzalez paper, there is one haplogroup D in Republic of Altai from Pre Iron period."
In fig. 4 there is one dot that looks like that but this is contradictory with the rest of the data presented in the paper, so I suspect it's a coloring error and nothing else.
In any case it's still not C.
It seems that there has been a lot of MtDNA L in the Near East and Mesopotamia, and recently they identified also Indian MtDNA M in Syria, but these samples are clearly much younger than the samples analyzed by Fernández (8000 A.P.).
ReplyDeleteIn this Fernández paper, we also get an insight into Syrian and Mesopotamian ancient mtDNA.
Tell Halula: L2, L3a, H, H/T, H/HV/HV1/R/U
Tell Ramad: H/K, K, C
Dja’de: H/T
Mari samples are much younger and are from Sumerian period:
Mari: L2, L3a, U4, J1a
I would guess that U4 came from the North. This ancestral haplotype 356C has been found in the oldest Krasnoyarsk samples (1800-1400 BC) and is today most common in Russia, Volga-Ural, Altai-Saian, Latvia, Bosnia-Macedonia and Hungary.
Fernández left many samples unidentified, but on Internet (http://www.egyptsearch.com/forums/ultimatebb.cgi?ubb=print_topic;f=15;t=005251) someone has identified In Tell Halula also H/K, T2b, H5, K (?), R (?), L2a1
In Tell Halula and Mari, there is again quite a lot of hg L, and in Tell Halula and Tell Ramad also H and K seem to be well presented, but to me the most striking thing is the presence of hg C! You can find a discussion on this haplotype on Internet (http://groups.yahoo.com/neo/groups/HumanMigrations/conversations/topics/4818)
(http://groups.yahoo.com/neo/groups/HumanMigrations/conversations/topics/4815)
This C has even been identified as C1.
If MtDNA C has been present in Western Eurasia, it is possible that it was not only moving with haplogroup NO and Q but also with yDNA R1. In this scenario, of course, ydna N may have picked up MtDNA C only after Q and R. :-)
According to my paper, frequency of haplogroup U4 in Sweden is 2.9%, so it is a little higher than the frequency of H1b.
"It seems that there has been a lot of MtDNA L in the Near East and Mesopotamia"...
DeleteI just had a cursory look at the haplotypes mentioned at the Fernández thesis ("resultados" section → page 490+, just in case someone wants to know where to look at) and I am quite uncertain that her haplogroup adscription is correct. For example you can't attribute an HVS-I sequence with a single transition at the 223 site to "L3a". It is almost certainly L3(xR) but what exactly within that is impossible to say. Probably the same kind of problems affect other lineages in that same list and I imagine that they are more likely to be N and M sublineages, although probably not all of them. In many cases we will never know without coding region data (cheapskates!)
"I would guess that U4 came from the North.".
The oldest one I know is from Morocco (Taforalt 12 Ka BP). Then, in the Epipaleolithic, it is found in Portugal, North Germany, Lithuania, Karelia and North Russia. That it does have a dominantly NE distribution does not mean its origins are there, just its most successful niche (quite different thing).
"In Tell Halula and Mari, there is again quite a lot of hg L"...
My previous look at the matter (checking with PhyloTree) implied some L3 but mostly K (and also some H).
"... but to me the most striking thing is the presence of hg C!"
If it arrived to Ukraine I see no reason why not to Syria. Interesting anyhow.
"If MtDNA C has been present in Western Eurasia, it is possible that it was not only moving with haplogroup NO and Q but also with yDNA R1."
I see no relation with Q, which is original from West Asia. R1a may or not have acted as "vector" but let us leave the matter open by the moment. Whatever the case, considering the absolute lack of Oriental lineages, including C, in Central Asia before the Iron Age, the migration route must have been via Siberia (first) and then Eastern Europe.
How do you know that there is U4 in Taforalt. I cannot find the Kefi study on Internet. I find only this:
DeleteGrotte des Pigeons, Taforalt 10,000 BC H, HV or U? 11 samples of CRS
http://realhistoryww.com/world_history/ancient/Misc/Data/European_genes_table_by_location.htm
Mesopotamian Mari U4 haplotype mutations are
356C
223T, 356C
In Taforalt you have CRS. According to my copy of Krasnoyarsk paper, this U4 356C is not found in Iberia or Africa.
But I agree that La Braña specimens and the fact that V and U5b are found in big numbers in the North and in Iberia, point to North - Southwest links.
The study or rather presentation has vanished from the Internet. But I swear I read it several times back in the day. The U4 result is from contrasting the HVS-I sequences with a recent PhyloTree build.
Delete"In Taforalt you have CRS".
Not just CRS: http://www.ancestraljourneys.org/nafricaadna.shtml
However, checking now again cursorily I might have comitted an error or maybe the defining sequence of U4 (transition at 16356) has changed. Maybe I mistook 355 for 356?
"According to my copy of Krasnoyarsk paper, this U4 356C is not found in Iberia or Africa".
Try reviewing the Chandler, Sykes and Zihao sequences: http://www.bris.ac.uk/archanth/staff/zilhao/dna2005.pdf
You will need to work with fig. 6, as there's no supp. material nor detailed table... but there it is: "356" in the non-R0 zone (left of the branch described as "[073]", which refers to HVS-II: an R0 defining marker). It was originally classified as just generic "U".
"considering the absolute lack of Oriental lineages, including C, in Central Asia before the Iron Age, the migration route must have been via Siberia (first) and then Eastern Europe."
ReplyDeleteI think that you cannot say that haplogroup C cannot originate from Central Asia because other oriental haplogroups do not originate from Central Asia.
Anyway, I am not sure that C is oriental to the same extent as D, and I don't think that Q is from West Asia, as I prefer a more Central Asian origin to haplogroup Q. However, I admit, that we do no know yet what actually happened.
I agree with you that ydna N brought D and perhaps the first A haplotypes to the taiga area, but I think that when ydna Q was moving in North-East Asia, haplogroup C and A may have been there right from the beginning.
With this I mean that MtDNA C was crossing Eurasia very early, 40-30 kya, and during the Ice Age it disappeared from the arid plains of Central Asia.
"... you cannot say that haplogroup C cannot originate from Central Asia because other oriental haplogroups do not originate from Central Asia".
DeleteNO ORIENTAL HAPLOGROUPS, NOT C EITHER! Nowhere it is documented a single Oriental lineage in Central Asia, including (non-Mongolian) Altai before the Iron Age. Only in Siberia and Eastern Europe (and Syria it seems per Eva Fernández - haven't double checked).
"Anyway, I am not sure that C is oriental to the same extent as D"...
I am. M8 in general, CZ in general and C in particular quite clearly originated in NE Asia. Check the data for that region please.
"... and I don't think that Q is from West Asia"...
Y-DNA Q clearly has most of its basal diversity in Iran and nearby areas. The lineages found in Siberia and America are a mere subset of this West Asian diversity.
"I agree with you that ydna N brought D and perhaps the first A haplotypes to the taiga area"...
I think that the link with C is much more clear. As for D and A they may also be related but they may well have traveled with other Oriental Y-lineages like O3 or C3, which expanded at a later stage. I just do not know enough to be sure but mtDNA C must have migrated Westward via Siberia and that is exactly what Y-DNA N1 did.
"... when ydna Q was moving in North-East Asia, haplogroup C and A may have been there right from the beginning".
In NE Asia indeed, not just C and A but also D and surely at least some B (because those are the lineages that were carried to America). But I see no reason whatsoever to thin that happened in Altai or Central Asia, that admixture happened as the Y-DNA Q "clan" migrated Eastward spreading the "mode 4" technology c. 30 Ka BP.
"With this I mean that MtDNA C was crossing Eurasia very early, 40-30 kya, and during the Ice Age it disappeared from the arid plains of Central Asia".
We have no evidence of C being in Europe before the Epipaleolithic. Most probably it arrived only in that period together with the Y-DNA N1 reindeer hunters of the far North, who benefited from the ice sheet contraction in that time.
You say that ydna Q crossed Altai c. 40-30 kya. In your opinion, what MtDNA was there at that time?
DeleteAt the very least X2, because this lineage made it to America. As for the rest (other possible lineages) I do not know. Maybe some U (U2, U7) and some H (H8, H13, etc.) but just speculating here based on some lineages with modern day Central Asian tendencies.
DeleteAnd I think that X2 came only later on. If Q picked up only Eastern MtDNA leaving all Western mtDNA behind, it must have gone through China and East Coast to America.
DeleteI still think that some of the original mtDNA should be left in East Asia (and America), because there are several Q haplotypes found in East Asia and even in Indo-China.
As for the difference between the distribution of D and C in Eurasia, please check these phylogenetic trees Figure 1 and Figure 3
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0015214
"And I think that X2 came only later on."
DeleteYou "think" a lot but why? I do not care so much about your subjective opinion but on what is founded: capricious opinions are of no use to science. The only serious way to explain X2 in Native America is by associating it with Y-DNA Q and the NE Asian spread of "mode 4" from Altai.
"If Q picked up only Eastern MtDNA leaving all Western mtDNA behind"...
Except X2.
"... it must have gone through China and East Coast to America".
Why? It went through NE Asia (plausibly not China as such but we do not know for sure) and once in America it spread to all the continent, including the East coast of North America, of course. X2 is also found in Altai and/or somewhere else in Siberia today. The trail is very clear.
"... there are several Q haplotypes found in East Asia and even in Indo-China."
But they all belong to a single sublineage, while all the upstream diversity is concentrated in West Asia and nowhere else. See the relevant Wikipedia article for example.
"As for the difference between the distribution of D and C in Eurasia, please check these phylogenetic trees Figure 1 and Figure 3"...
In the case of C it is very clear that the basal diversity is concentrated in whatever are the red dots. Let's check the legend: "eastern Asian – in red", bingo!
The other quite diverse area is the green one, which is... "central and southern Siberian – in green".
So that seems to mean East Asia → Siberia → everywhere else.
The same can be said about D but it is an even more clear cut case. If you have any doubt re. C, check also where its relatives Z and M8a are found. Of course these are all NE Asian lineages with wide distribution in Siberia and beyond (Native America, Eastern Europe) but it's not like we can deny them the East Asian origin at all. Not even if we try hard.
BTW the paper has a clear error in the conclusions: "The peopling of northern Asia by anatomically modern humans probably began more than 40 kya, with the first evidence in the Altai region"...
DeleteWe know now for certain it's not the case but that peopling by AMH of South and Central China is much much older, some 50-55 Ka older than that of Altai. We also know that by the time of the peopling of Altai, and some 10-20 Ka before its techno-cultural influence on East Asia, by AMH, mtDNA B4'5 was already to be found near Beijing. What the authors say is biased by the dogma of "modern behavior", which is clearly Eurocentric, quasi-religious and, well, simply wrong, as I have discussed recently.
I agree that at least part of hg C came with yDNA N to Europe, but I am not so sure that all of it came, and in particular C1 makes me doubt that connection.
ReplyDeleteWe disagree on many points, but never mind. Many points will get clearer with time.
ReplyDelete"The only serious way to explain X2 in Native America is by associating it with Y-DNA Q and the NE Asian spread of "mode 4" from Altai."
I do believe that X2 is Native American and came from Siberia, but for linguistic reasons I believe that there were several migrations to America from Asia, and this one which brought X2 was not the first one, and it may have been brought even by R1b.
“But they all belong to a single sublineage, while all the upstream diversity is concentrated in West Asia and nowhere else. See the relevant Wikipedia article for example”
Wikipedia says that possible places of origin for yDNA Q are Central Asia, the Indian Subcontinent and Siberia. It also says that Q* is found with low frequency in India, Pakistan and Afghanistan; this is not West Asia to me. I think that phylogeny of Q is not at all clear yet. We have already discussed that in many papers they use undefined Q1* or even P-45 or K-M9. I have seen a more detailed analysis only in a few papers, e.g. Maliarchuk et al 2011, Zhong et al, 2010.
According to this new paper ”The First Peopling of South America: New Evidence from Y-Chromosome Haplogroup Q”, mongol Q falls in several different clades: M120, M25, L53, L330, and I have no idea to which clade for example Vietnamese Q belong, and also Tibetan Q could be better analyzed.
In Zhong paper, Fig 3 is interesting in this respect (it uses the old nomenclature). It says that with respect to Q1a1, detection is not done in Europe, or in Nepal (?), South Siberia, and Q1a1 is detected in Pakistan, Southeast Asia, Southern East Asia and Northern East Asia. With respect to Q1a2, detection is not done in Europe, Nepal or Southeast Asia, South Siberia, and Q1a2 is detected in Turkey, Iran and northern East Asia. With respect to Q1a3, detection is not done in Europe (except for 1 study), Southeast Asia or South Siberia, and Q1a3 is detected in Europe, Iran, Pakistan, India (?), Southern East Asia and Northern East Asia. With respect to Q1b, detection is not done in Europe, Southeast Asia or South Siberia, and Q1b is detected in Iran, Pakistan and Northern East Asia. In Maliarchuk paper they in fact analyzed south Siberian Q haplotypes.
Wikipedia says that the possible place of origin of mtDNA C is Central Asia: “Haplogroup C is believed to have arisen somewhere between the Caspian Sea and Lake Baikal some 60,000 years before present.”
I said that if Q left all Western mtDNA behind it must have gone through China and East Coast to America, because it cannot have changed women if there were no East Asian women at sight. I really do not think that Q went through Saha and met "Chinese" B women there when it was getting very cold. As for mtDNA A, Native American A is close to Chukchi A, and NA and Chukchi A is close to Altaian A, so this change of women should have happened with people who are now Altaians, and yDNA should be then N or C3. I am not so sure that C3 was in Altai 30-20 kya (I do not exclude it however) and N is not so clearly linked with mtDNA A. As for mtDNA C1, it is found in East Asia (C1a), but very sparsely (Amur, Japan, Xinjiang, Buryatia) and its age in Native Americans is older. The other branches of C1 are in Europe (ancient Karelians, modern Icelanders), Near-East, India and Siberia. As for mtDNA D, I am enclined to think that it came to America at least in part with yDNA C, but it is more just a hunch.
I'm going to go nuts with unexplained claims like "Native American R1b". Unless you can demonstrate me otherwise, Native American R1 is product of recent European admixture.
DeleteIt has been reported at:
→ http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1180678/ [Bortollini 2003, as P(xQ)]
→ http://mbe.oxfordjournals.org/content/23/11/2161.long [Bolcknick 2006 as mostly R1, but also some P* among the Cherokee]
And then recycled by:
→ http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2584155/ [Singh 2008]
The R1 is generally accepted to be recent European admixture, which in some populations is as high as 50% (Y-DNA), what makes total sense considering their history. Nobody ever AFAIK has shown that that R1 among Native North Americans is anything else, such as a distinctive R1b subclade or whatever: it's French (and other) trapper blood, nothing else. These NE NA populations were already very "whitened" 150 years ago, before racist quasi-apartheid became the norm with the trail of tears, Jim Crow and new very large waves of European immigrants.
More intriguing is some P(xR1) reported in some populations, notably the Cherokee. This I can't say what it is or what it come from. From memory, a discussion we had back in the day (2006-07 probably) suggested that (based on the STR haplotypes) it appeared to be R2, what would be even more intriguing, as it is a mostly South Asian lineage. But nothing certain because no specific study on Northern NA P(xQ) exists.
Unless of course you know something I do not.
Yes, indeed, the problem is that there are so many things that should be better analyzed. We just have to be patient.
DeleteHowever, I like new ideas, and I completely forgot in my considerations above the possibility that yDNA Q met yDNA D in Altai, Northern China 30-20 kya, and mtDNA A was taken from these tribes.
"Wikipedia says that possible places of origin for yDNA Q are Central Asia, the Indian Subcontinent and Siberia".
DeleteDon't read the text, read the data. Following http://en.wikipedia.org/wiki/Haplogroup_Q_%28Y-DNA%29#Subclade_Distribution ... :
The highest phylogenetic tier: Q splits at the root between:
→ Q*: "Found with low frequency in India and Pakistan. Important in Afghanistan, paragroup Q-M242 (xMEH2,xM378) was found at 16.3% in Pashtun people."
→ Q1 (P36) ↓
The second phylogenetic tier: Q1 splits into:
→ Q1*: Found with low frequency in Iran.
→ Q1a (MEH2) ↓
→ Q1b (L275): West Eurasia and South Asia
Third phylogenetic tier: Q1a splits into (not the most updated phylogeny but anyhow):
→ Q1a*: Koryaks, paleo-Eskimo
→ Q1a1a1 (M120): East Asia
→ Q1a1b (M25): Iran, Lebanon, Turkey
→ Q1a2 (M346) ↓
Fourth phylogenetic tier: Q1a2 splits into (again not well updated):
→ Q1a2*: West Eurasia and South Asia
→ Q1a2a (L53): NATIVE AMERICANS
→ Q1a2c (M323): Yemenite Jews
So maybe it can be argued for Q to be originated from South Asia ultimately (around Pakistan maybe?, between Iran and NW India I'd say) but it clearly has an important West Eurasian distribution at all levels. East Asian and Native American subclades, even if numerically more important, are just subsets of the South & West Eurasian diversity.
"We just have to be patient (...) I like new ideas"...
DeletePatient since 2006? New ideas from 2006? Nothing "new" here and I have been patient on this matter (by force) since some six or seven years ago. I just see no reason to speculate wildly while no new data arrives.
"Wikipedia says that the possible place of origin of mtDNA C is Central Asia: “Haplogroup C is believed to have arisen somewhere between the Caspian Sea and Lake Baikal some 60,000 years before present.”"
DeleteAgain it's just literature. M8a is almost exclusive of NE Asia, notably Japan. CZ is more widespread but in any case it looks original from that same region. IF there would be no M8a, which is co-basal to CZ within M8, then we could have some doubt but luckily there's no reason for it.
"I said that if Q left all Western mtDNA behind it must have gone through China and East Coast to America, because it cannot have changed women if there were no East Asian women at sight".
DeleteOf course. But we do not know for sure to what latitude exactly the Eastern Asian colonization of that time reached to. We know that it was at least North China but it can well have been quite further northwards into the Russian Far East (Eastern Siberia). After all if people were in Central China since 100,000 years ago, they had time to very gradually adapt to colder climates. Also the coasts especially offer excellent resources, so I would expect that some 30 Ka ago, when the "Q clan" began moving eastwards, there were already some native peoples in the Russian Far East, not sure exactly where. Otherwise, I'd expect more resilience of the Q founder effects in those regions (Northern China, Mongolia, etc.) something we just do not see (but much further North).
This comment has been removed by the author.
DeleteThis comment has been removed by the author.
Delete"So, also M8 seems to have a northern origin"...
DeleteA Northern origin with East Asia in any case, a NE Asian origin.
... "and I see that it is possible that M8 arose in Central Asia".
There's nothing in that data supporting such idea. What do you have against East Asia? I mean, really...
"The age of M8a variation"...
Nobody knows with any certainty: it's just a speculation, an estimate after many semi-arbitrary assumptions.
Haplogroup M should be c. 100 Ka old, what implies that M8, just 4 coding region mutations downstream (in a lineage that has evolved as many as 22 in some branches) should be from maybe 80-60 Ka ago, M8a and Z 45-65 Ka and C 35-55 Ka (C1 should be just barely more recent, being the only C subclade with no c.r. mutations: at coding region level C1 is identical to generic C). The dates you copy-pasted are totally out of bounds with my most basic understanding of mtDNA history in Eurasia.
Altaians harbour four different M8 haplotypes, 1xM8, 3xM8a. The paper says “In addition, several M8 mtDNA sister haplotypes to the CZ-haplogroup were found in Altaians. The majority of these sequences is determined by a transition at np 16319 and forms a distinct cluster, M8a”.
DeleteThe age of M8a variation is 15700-36700 and the age of C variation is 19400-37400. In M8, M8a2 is much older than M8a1. The age of the former is 11200-25800 and of the latter -400-6300, and the older one is the northern one found in particular in Itelmen and Koryak. So, also M8 seems to have a northern origin, and I see that it is possible that M8 arose in Central Asia well before the Ice Age, and then, C1 back-migrated to the West between 20-10 kya.
According to Tanaka paper, M8 is found as follows:
M8
0.15% Japanese, 0.47% North China, 0.13% Southeast Asia/South China, 1.67% Chukchi, 0.37% Korea
M8a
1.22% Japanese, 4.23% North China, 0.92% China 2, 4.17% Hubei/Quangdong, 1.5% Southeast Asia/South China, 1.47% Kazak/Kirgiz/Uygur, 2.5% Buryat, 1.68% Korea
M8a2
0.23% China 2, 1.08% Kazak/Kirgiz/Altai, 13% Itelmen, 20.9 KAM Koryak, 1.67% Chukchi, 2.78% Tuvinian, 0.37% Korea
China 2 means Xinjiang Han, Yunnan Sali, Qinghai Tu, Xi’an Han, Shanghai Han and Mongolian
I think that it is not without significance that M8 is so strong in Koryak and Chukchi. Both Koryak and Chukchi have haplogroup Q, and moreover Koryak bear this very old haplotype of Q.
I really have nothing against East Asia! Why would you think so? The separation between East and West is not at all so clear-cut and the separation between Europeans and East Asians is said to have happened only 40 kya, if I remember correctly...
"This comment has been removed by the author."
DeleteSeems I replied before you deleted.
"And Derenko paper says that Altaians harbour four different M8 haplotypes, 1xM8, 3xM8a".
Maybe but not documented in aDNA before Iron Age. How does it compare with, say with regions further East. We always see more diversity in East Asia, as the phylogenetic trees previously discussed show. In fact you are the first person ever, to my knowledge, to question the Oriental origin of M8/CZ.
In any case I think you're being extremely partisan in this matter, as if you had phobia to East Asia and tried to dismiss by all means at hand: fighting a battle on it. Is mtDNA C your personal lineage? Why all the fuzz?
"I think that it is not without significance that M8 is so strong in Koryak and Chukchi. Both Koryak and Chukchi have haplogroup Q, and moreover Koryak bear this very old haplotype of Q".
Proves nothing other than C being in NE Asia since very long ago, something we agree with. Actually what I am saying is that it is much older than the "Q clan" migration eastwards (with mtDNA X2), as are all the other non-X Amerindian lineages A, B, C and D, as are other lineages which did not make it to Sibero-America, like N9, as is surely Y-DNA C3, etc.
or was it only 30 kya... I have checked so many details that it is enough. Please correct me.
Delete"This comment has been removed by the author."
DeleteAgain?
It is true that I did recently National Geographic Geno 2.0 test. They say that I am I5a which is very strange indeed. However, they said that I am 4% Native American and it seems that I have a few typical Native American gene sequences. I think that most of us are genetically very mixed.
DeleteThere was an error again! So I am 2% Native American and 4% Northeast Asian!
Delete"In fact you are the first person ever, to my knowledge, to question the Oriental origin of M8/CZ."
DeleteI think that I can't be if Wikipedia proposes a Central Asian origin, but please note that I just consider different models. I do not say that it must have happened in a certain way. The truth is sometimes something completely new to everybody and sometimes it is in between different models.
The Wikipedia generic haplogroup M entry (http://en.wikipedia.org/wiki/Haplogroup_M_%28mtDNA%29), which was largely developed by me (as Sugaar) in 2008 (see history) and mostly expanded and curated by Ebizur (who is extremely knowledgeable of the literature), DOES NOT claim nor propose any "Central Asian origin" for M8. It just mentions it is also found in Central Asia, as happen with many other East Asian lineages (at least A, D, F and G, as well as Y-DNA ones like O3, C3, etc.) If you click the associated external link you get a list of Chinese and Japanese haplotypes, as well as a single Ulchi and the text: "Haplogroup M8 is a small East Asian haplogroup".
DeleteThe brief CZ Wiki-entry reads "Today CZ is found in eastern Eurasian populations. It is most common in Siberian populations".
It is true that the C entry says unreferenced nonsense about C being originated between the Caspian Sea and Lake Baikal but again this is just blah-blah. When it comes to the real (and sourced) stuff it says: "Haplogroup C is found in Northeast Asia[2] (including Siberia). In Eurasia, Haplogroup C is especially frequent among populations of arctic Siberia, such as Yukaghirs and Nganasans[3]", what hardly has to do anything with the Caspian Sea, although it may be related to Lake Baikal or some other NE Asian region indeed.
The unsourced claim about the origin was added by an anonymous user with IP 82.6.29.26 on February 2009. It has never been sourced and certainly merits removal (but I do not collaborate with Ziopedia anymore).
The separation between East and West is not at all so clear-cut and the separation between Europeans and East Asians is said to have happened only 40 kya, if I remember correctly...
DeleteIs 40 or 30 kya recent as a time for the separation of Caucasoids and Mongoloids ("European" and "East Asian" are not the correct terms here; the correct terms would be "Caucasoid" and "Mongoloid", but of course you can instead use the somehow more vague terms "West Eurasian" and "East Eurasian" if you think that "Caucasoid" and "Mongoloid" are "politically incorrect")?! In modern human scales, 30 or so kya is old enough for a Caucasoid-Mongoloid separation time considering the relatively recent origin of modern humans as a whole and their relatively high mobility and fast evolution throughout much of their existence.
I also agree with Luis that Central Asia is not much of a source of anything but largely a population sink.
DeleteThen there is this possibility that yDNA O (or NO) was already there in the North (30-)20 kya with mtDNA B and C (and A and D) when yDNA Q arrived, and most of the O went back to the South during the Ice Age. Of course, it is also possible that yDNA O, C3 and D were all (or some of them) there at that time and receded to the South when it got too cold. Then, YDNA (N)O and D may have made their way to Northern China either from the South (as most people say) or from the West. On the basis of Zhong paper, I would say that C took the southern route to China, but some people have argued for a northern route for C too.
DeleteIf we agree that mtDNA M8 was up in the North 30-40kya as we know mtDNA B was, it must have been accompanied by a Y line. If M8 was not with Q, it must have been with (N)O, D or C. Many possibilities indeed!
West and East Eurasian populations, Onur. I don't think that anthropometric "races" are in most cases valid categories (for example I disagree with the notion of a "Mongoloid" valid grouping on merely cranioopometric grounds, although I agree that those populations share a similar ancestral genetic pool) and in any case it would be a morphological issue like blond hair or prominent cheeks, whose relation with the overall genetic background is not clear at all.
DeleteWe're talking populations and genetics, not "races and faces". 100,000 or even 50,000 years ago the "races" (looks) were surely not yet conformed because it is a process of homogeneization that takes time and not just founder effects.
... "if you think that "Caucasoid" and "Mongoloid" are "politically incorrect"".
Not just politically incorrect (that may depend on sensibility) it is conceptually incorrect: it's making assumptions about looks that we do not know how real were in the past. I strongly prefer to stay put in objective and neutral terms as populations (West Eurasians for example) and genetics, rather than to speculate on the looks of people many millennia ago or to begin a discussion on whether the concept "Mongoloid" has any bearing in its own purely anthropometric terms (I do not think so or very weakly at most).
Seriously...
"In modern human scales, 30 or so kya is old enough for a Caucasoid-Mongoloid separation time considering the relatively recent origin of modern humans as a whole and their relatively high mobility"...
DeleteIn strong disagreement here too. The timeline for the separation of West Eurasians (from South Asians essentially) can't be more recent than 50 Ka and is probably a bit older (independently of possible Neolithic backflows into Pakistan-India), at least 55 Ka (Emirian).
The "recent origin" of humans as a whole is at least 200,000 years, not sure what you had in mind.
The "relatively high mobility" unless clearly specified to archaeologically defined cultures is such a vague term that is meaningless. Some people seem to have moved a lot (from Altai to America for example, but they "swapped race" in the process, which was at least 10 Ka long), others not. For example West Eurasians seem to have remained relatively the same (speaking of genetic lineages) since the beginnings of the Upper Paleolithic.
"I also agree with Luis that Central Asia is not much of a source of anything but largely a population sink".
DeleteActually it seems to be at the origin of Native Americans (and related populations of Siberia carrying Y-DNA Q). But this seems to be the only case before the Indo-Iranian and Turkic expansions.
However we still do not know well what role Central Asia may have played in the Neolithic or the still open R1a question.
What is quite clear to me is that all "Soviet" Central Asia, including Altai, belonged culturally and genetically to West Eurasia ("Caucasoid" in your racial terminology) until the Iron Age or so. Siberia instead rather belonged to East Asia or (West Siberia) was an intermediate zone.
Also there seems to be a well defined West Siberian-Central Asian specific autosomal component (half-wat between East and West in terms of Fst), what implies that ancient Central Asians were at least largely ancestral to its modern ones, regardless of other flows. In the case of Kazakhs or Uyghurs this local ancestry is like 60%, in the case of the Khanty almost 100%.
"If we agree that mtDNA M8 was up in the North 30-40kya as we know mtDNA B was, it must have been accompanied by a Y line. If M8 was not with Q, it must have been with (N)O, D or C. Many possibilities indeed!"
DeleteI would argue for C3, although it is also very likely that NO lineages like O2b or O3 (or also some ancient forms of N) were also around back then. Let's not forget about D either. However, of all these, only C3 and N1 would really push northwards into the Siberian taiga. Assuming patrilocality, this seems to say something about the different eco-cultural stands of the various paleo-populations of NE Asia but I would not dare to enter into a too detailed speculation because chances are that I would be wrong.
I don't think that anthropometric "races" are in most cases valid categories (for example I disagree with the notion of a "Mongoloid" valid grouping on merely cranioopometric grounds, although I agree that those populations share a similar ancestral genetic pool)
DeleteMongoloids as a group (whether you call them "race" or "population") are more united genetically than morphologically, but still, their overall morphology is distinct and united enough to separate them from all other modern human groups and group them in a single category (whether you call it "Mongoloid" or "East Eurasian + Amerindian"). The other modern human races too have their own distinctive and united genetic and morphological characteristics sufficient to validate them.
and in any case it would be a morphological issue like blond hair or prominent cheeks, whose relation with the overall genetic background is not clear at all.
At present, the relation between individual genes and morphological traits are newly beginning to be resolved, but overall autosomal genetics and haplogroup genetics have a fairly well degree of correlation with overall morphology in differentiating between the modern human races.
We're talking populations and genetics, not "races and faces". 100,000 or even 50,000 years ago the "races" (looks) were surely not yet conformed because it is a process of homogeneization that takes time and not just founder effects.
That is why I do not use racial categories when talking about modern human groups from such distant periods.
Not just politically incorrect (that may depend on sensibility) it is conceptually incorrect: it's making assumptions about looks that we do not know how real were in the past. I strongly prefer to stay put in objective and neutral terms as populations (West Eurasians for example) and genetics, rather than to speculate on the looks of people many millennia ago or to begin a discussion on whether the concept "Mongoloid" has any bearing in its own purely anthropometric terms (I do not think so or very weakly at most).
That is why for much of the Upper Paleolithic period I prefer to use racial terms that begin with "Proto-" for specific groups (e.g., Proto-Mongoloids).
In strong disagreement here too. The timeline for the separation of West Eurasians (from South Asians essentially) can't be more recent than 50 Ka and is probably a bit older (independently of possible Neolithic backflows into Pakistan-India), at least 55 Ka (Emirian).
DeleteIt is Kristiina who claimed a 30 or 40 kya separation time for Caucasoids and Mongoloids. I did not comment on the validity or non-validity of that separation time but just on whether it is recent or not in modern human evolutionary scales.
The "relatively high mobility" unless clearly specified to archaeologically defined cultures is such a vague term that is meaningless. Some people seem to have moved a lot (from Altai to America for example, but they "swapped race" in the process, which was at least 10 Ka long), others not. For example West Eurasians seem to have remained relatively the same (speaking of genetic lineages) since the beginnings of the Upper Paleolithic.
I meant "high mobility relative to archaic humans", for which I am sure you will agree with me.
Actually it seems to be at the origin of Native Americans (and related populations of Siberia carrying Y-DNA Q). But this seems to be the only case before the Indo-Iranian and Turkic expansions.
DeleteThe role of Central Asia in the origin of Amerindians is far from clear. The Turkic expansion began from what is now Mongolia and its environs, thus from East Eurasia rather than Central Asia; it was largely an expansion into Central Asia rather than from it. But in the Indo-Iranian expansion Central Asia indeed seems to have served more as a source than a destination.
"That is why I do not use racial categories when talking about modern human groups from such distant periods".
DeleteWell you did and that's why I went all through those comments.
"It is Kristiina who claimed a 30 or 40 kya separation time for Caucasoids and Mongoloids".
My misunderstanding then. Sorry.
"I meant "high mobility relative to archaic humans", for which I am sure you will agree with me".
I do not know. It is true that H. sapiens were effectively faster and did tend to use somewhat larger territories for the same ecological conditions (i.e. the Aquitanian H. sapiens used larger territories than Aquitanian Neanderthals but not than German ones), and it is true that this kind of different "mobility" may have been one of the key Sapiens advantages in the long term competition with Neanderthals. However this probably does not apply to the more slender but less large-brained H. erectus.
In any case I don't understand its implications on ethnic or "racial" differentiation among humans. I thought you meant that they could have moved from here to there once and again and again and again with no pattern nor even mild stability (I've heard that before from others and I quite disagree), but obviously it was not what you had in mind.
"The role of Central Asia in the origin of Amerindians is far from clear".
If you look at the origin of Y-DNA Q, as we did above, it should be quite clear that it coalesced in South or West Asia before spreading, via Altai, to NE Asia and America. This is pretty much in agreement with the pattern of Aurignacoid tech ("mode 4", "Upper Paleolithic") in Altai (since c. 47 Ka BP) and Mongolia and NE China (since c. 30 Ka BP, clearly derived from Altai).
See for example:
http://ejournal.anu.edu.au/index.php/bippa/article/viewFile/84/75
http://forwhattheywereweare.blogspot.com/2012/08/upper-paleolithic-of-north-china-c.html
Also there is that dog genetic data, with the oldest sequenced dog ancestral to modern ones being from Altai (c. 30 Ka BP) and being mostly ancestral to Native American dogs: http://forwhattheywereweare.blogspot.com/2013/03/33000-years-old-dog-from-altai-is.html
I have absolutely no doubts about the ultimate Altaian origins of the Native American (and related Siberians) major patrilineages within Q (and therefore some cultural elements surely). However they overall genetics (autosomal) as well as almost 100% of their matrilineages are East Asian.
"The Turkic expansion began from what is now Mongolia and its environs, thus from East Eurasia rather than Central Asia; it was largely an expansion into Central Asia rather than from it".
Originally for sure but then Central Asia must have acted as trampoline and therefore as source (secondary source if you wish) for further Turkic expansion. Similarly with Indoeuropeans in my understanding, although in this case the ultimate origin is in the West (Samara Valley and such).
My point is just that it was indeed a destination first but later also a secondary source. Another case would be West Asian flows, be them Paleolithic (as with the Altai-Amerindian issue) or Neolithic (at least the southern part of Central Asia was influenced by West Asian Neolithic as far as I know, although more research is needed on that area's Neolithic). We can agree in any case that the region is essentially a melting pot, especially nowadays.
Well you did and that's why I went all through those comments.
DeleteWrong. I did not use any racial categories for the periods before the Upper Paleolithic.
In any case I don't understand its implications on ethnic or "racial" differentiation among humans. I thought you meant that they could have moved from here to there once and again and again and again with no pattern nor even mild stability (I've heard that before from others and I quite disagree), but obviously it was not what you had in mind.
Of course there is a pattern, but that pattern is more flexible than that of Neanderthals and this must have some implications on the distribution of racial/population groups of modern humans relative to those of Neanderhals.
Originally for sure but then Central Asia must have acted as trampoline and therefore as source (secondary source if you wish) for further Turkic expansion. Similarly with Indoeuropeans in my understanding, although in this case the ultimate origin is in the West (Samara Valley and such).
My point is just that it was indeed a destination first but later also a secondary source.
In both the Indo-Iranian and the Turkic expansions Central Asia was not the ultimate source of the expansions but a secondary source. But in the Indo-Iranian expansion Central Asia greatly exceeded in importance the ultimate source of the expansion and thus served more as a source than destination. In contrast, in the Turkic expansion Central Asia never exceeded in importance the ultimate source of the expansion so that Central Asia served more as a destination than as a source.
"I did not use any racial categories for the periods before the Upper Paleolithic".
DeletePrecisely. I think that using racial categories in general is highly questionable, unless you are talking craniometry specifically, but more so in such remote times as the (early) Upper Paleolithic when the modern phenotype differences were surely still in very early stage at best. I do not think it's technically correct to talk of "Caucasoids" for example regarding the Aurignacian (very few human remains, some semi-archaic like Pestera or with ambiguous features like Emirian ones, sometimes described as "Mechtoid" or whatever). It would be quite debatable even for a later period like Gravettian and Solutrean, when we begin to find more remains (so called "Crô Magnon" type). It is in any case a different debate than genetics and I find it quite distracting.
On the racial affiliation of Cro-Magnons, see:
Deletehttp://racialreality.blogspot.com/2011/02/cro-magnons-were-caucasoid-not-negroid.html
Also, North Africans seem to have been pure or almost pure Caucasoids until recent times.
They are outliers tending towards Mongolia...
DeleteI don't think that kind of graph is too useful nor informative but, for whatever meaning it can have, it's outstanding that they are halfway between modern Germans or Italians and Bronze Age Mongols.
"North Africans seem to have been pure or almost pure Caucasoids until recent times".
I don't think so. African specific genetic influence in the region (and also to lesser extent in Arabia) seems very very old, it can well be from the time of the OoA (Aterian in the North African case) and may have been reinforced later on, for example in the Neolithic semi-pluvial. However I agree that the bulk of the ancestry (2/3 to 3/4, with local variations) seems to be European or West Asian.
Craniometrically we can find all kind of affinities and probably also local peculiarities, for example they tend to be dolicocephalic (not meso, dolico), what is a clearly African element. You find even very extreme and striking dolicocephaly in Egyptian mummies, so it's not recent. However skulls like Taforalt seem more European-like, probably reflecting the important Gravetto-Solutrean → Oranian genetic flow.
It's a complex matter in any case, so we better talk genetics and leave craniometry for the old-fashioned freaks.
Also CM-1 may or not be representative of Gravettian peoples usually known as "Cro-Magnons", he may be peculiarly robust in fact, an individual trait no doubt. But he also has some Oriental-like traits like prominent cheeks not so common in West Eurasia nowadays.
DeleteFor the pure or almost pure Caucasoid racial affiliation of ancient North Africans, see for example:
Delete"MITOCHONDRIAL DNA AND PHYLOGENETIC ANALYSIS OF PREHISTORIC NORTH AFRICAN POPULATIONS
North Africa is located at a crossroad between Europe, Africa and Asia and has been inhabited since the Prehistoric time. In the Epipaleolithic period (23.000 years to 10.000 years BP), the Western North Africa has been occupied by Mecha- Afalou Men, authors of the Iberomaurusian industry. The origin of the Iberomaurusians is unresolved, several hypotheses have been forwarded. With the aim to contribute to a better knowledge of the Iberomaurusian settlement we analysed the mitochondrial DNA (mtDNA) of skeletons exhumed from the prehistoric site of Taforalt in Morocco (23.000-10.800 years BP) and Afalou in Algeria (11.000 to 15.000 BP -Algeria). Hypervariable segment 1 of mtDNA from 38 individuals were amplified by Real-Time PCR and directly sequenced. Sequences were aligned with the reference sequence to perform the mtDNA classification within haplogroups. Phylogenetic analysis based on mitochondrial sequences from Mediterranean populations was performed using Neighbor-Joining algorithm implemented in MEGA program. mtDNA sequences from Afalou and Taforalt were classified in Eurasiatic and North African haplogroups. We noted the absence of Sub-Saharan haplotypes. Phylogenetic tree clustered Taforalt with European populations. Our results excluded the hypothesis of the sub-Saharan origin of Iberomaurusians populations and highlighted the genetic flow between Northern and Southern cost of Mediterranean since Epipaleolithic period."
This is from the ISABS 2013 abstracts, so the the full text is not publically available at present. But there are already publically fully available analyses (including craniometric) of ancient North Africans with similar results.
What the fuck with racial language?! Seriously! Things are not black and white: even if there are clusters there are also clines. Also I am very proud of my (practically certain) African ancestry, even if it is minimal. Fuck endogamous purebreeds! Up with clinality and racial diffuseness! It really annoys me to pretend that things must be white or black and not grey or multicolor as they really are most of the time. Your racialist obsession is kidnapping again reasonable discussion here.
DeleteGenetic affinity is not "race". I know perfectly that Taforalt has only (or almost only) West Eurasian lineages but Taforalt is not all North Africa (and even if you include Affalou, that is not enough). In fact the most interesting persistence of apparent deep ancestry (Aterian?) is today in Southern Morocco, in the area around Marakesh. See:
http://forwhattheywereweare.blogspot.com/2011/12/north-african-genetics-through-prism-of.html
The South Moroccan component is equidistant at high Fst values (of inter-continental level) from the West Eurasian and Tropical African components and that's why I am almost certain it is an ancient Aterian remnant. A similar case is found in Egypt and Saudi Arabia: http://forwhattheywereweare.blogspot.com/2012/01/egyptian-genetics-in-regional-context.html
Also in Epipaleolithic Portugal we have African "L" lineages (at least one likely L3d2 and another unclear L(xR), what can only be explained, I understand, because of back-migration from North Africa at the time of the Oranian genesis, which is probably also at the origin of tanged points in Iberian Gravetto-Solutrean, a concept probably borrowed from Aterian. This African signature is still quite visible in Western Iberia, at least in the genetics but also arguably when you look at the phenotype, which sometimes approaches the North African typology like the so-called Berid type.
How you can read my writings as racialist obsession is beyond me. I am just making a scientific discussion here, basing my statements on the results of scientific papers.
DeleteADMIXTURE/STRUCTURE components do not come with age. Some components, particularly those that are largely restricted to a single population or a few closely related populations, are recent in origin and a result of high inbredness. The "Kalash" component seen in many ADMIXTURE and STRUCTURE analyses is a classical example to such components. Your "Southern Moroccan", "Egyptian" and "Saudi" components seem to have little difference from the "Kalash" component in this respect.
As for the L lineages in West Eurasia, many of them seem to be very ancient in West Eurasia and might even be from the time of the initial Out of Africa.
"How you can read my writings as racialist obsession is beyond me."
DeleteI cite: "Caucasoid", "Mongoloid", "racial affiliation of Cro-Magnons", "North Africans seem to have been pure or almost pure Caucasoids", etc.
But then in most cases we are talking population genetics, not craniometry, which is what those categories imply. The only case where it would apply, it happens that the "racial [i.e. craniometric] classification of Cro-Magnons" is clear as mud (half-way to Mongolia). But you still insisted on racializing the issue against Africa with the use of the false "racial" category "Negroid". Africa is way to diverse to belong to any single "race" but indeed most populations share some traits, what is a perfect example on how "race" is constructed by homogeneization through time and therefore "purity" is an absurd concept: all "races" are mixed.
All that language of the early 20th century has way too many implications of apartheid, Jim Crow and then of course the Nazis and is not really scientific, as demonstrated in the famous "The Mismeasure of Man" and many other data we can see easily. Talking happily of "races" all the time is not acceptable in polite society, as they say, nor in the worst slums either. You seem to live in a bubble about this but I mean that even in Turkey your way of talking raises some eyebrows.
You pretend it is scientific terminology but it's not: and you do not normally find such terms in scientific literature anymore, at least not the one I read such as human population genetic studies. So if scientists find no problem (or even quite useful and convenient) in using other language, why would you claim that your extremely and obsessively racialized language is "scientific"?
You are continuously making a point about your personal beliefs on the matter of human differences. You have demonstrated to be stubborn like a mule and refractory to others' opinions. You don't care and hide behind the word "scientific", which you abuse more than a Brezhnevian neo-Stalinist.
You made a promise on this matter anyhow (that was a condition to lift your ban) but you are not fulfilling it. So... what do we make of it?
"ADMIXTURE/STRUCTURE components do not come with age."
DeleteThey come with Fst distances, what is about the same, at least as I see it: components with inter-continental level differences to every other "racial cluster" around mean OoA age. Similarly all West Eurasian clusters are within a relatively tight bracket in terms of Fst to each other, what means that they diverged in a specific time bracket, which would be between 1/2 and 1/4 (rough est.) of the OoA.
Whatever the case with an inter-continental Fst distance to both "Caucasoids" and "Negroids", using your racialized wording, they are neither, and they are not a mixture either: the component represents a small North African specific "lost race", which has indeed been spotted in phenotype studies as "pseudo-Mongol" or "pseudi-Khoisan" but is so diluted that only one individual out of many, even in the refuge areas, shows some of those traits anymore.
Even more interesting is that another such component seems to exist around the Red Sea.
As racialist, I thought you would be interested in this but I see you are not, what makes me wonder about your "scientific" attitude, based on open-mindeness and not just stubborn parroting of old books - that's religion, not science - and even as religion is a very poor attitude of extremely low spirituality, I'd say: one can't see if he closes his eyes.
"Your "Southern Moroccan", "Egyptian" and "Saudi" components seem to have little difference from the "Kalash" component in this respect."
As far as I know the Kalash component has low Fst distance values, just like the many Tunisian Berber ones, etc.: those are product of recent endogamy and suggest that the population lacks interest for this kind of analysis (actually is noise). You don't seem to understand how important are relative Fst values when understanding what autosomal components actually are. No wonder: most papers ignore them (between population Fst values yes but between components only rarely, when it's actually much more useful in this case).
"As for the L lineages in West Eurasia, many of them seem to be very ancient in West Eurasia and might even be from the time of the initial Out of Africa."
Indeed, that's precisely my point. But the N and M lineages in West Eurasia only arrived since the Aurignacian (senso lato), what means almost 50,000 years of staying put in North Africa and Arabia before the remix. That's the difference.
West Eurasian autosomal components are closer to East Asian ones than to the South Moroccan or Red Sea ones, which are clearly OoA remnants (and are not closer to Tropical African ones either). Maybe this breaks your racialized schemes to pieces but in fact it should not unless those are too simple and not really scientific.
But then in most cases we are talking population genetics, not craniometry, which is what those categories imply. The only case where it would apply, it happens that the "racial [i.e. craniometric] classification of Cro-Magnons" is clear as mud (half-way to Mongolia).
DeleteAs I said before, there is a fairly high degree of correlation between the morphology-based racial categories and the genetics-based big clusters of modern humans. That justifies the application of the same racial categories on the genetics-based big clusters. This is in fact what population geneticists such as Cavalli-Sforza did back in the day. If they do not use racial categories anymore, this is not due to any scientific development since then but due to the decades-long campaigns of the politically correctist camp.
But you still insisted on racializing the issue against Africa with the use of the false "racial" category "Negroid". Africa is way to diverse to belong to any single "race" but indeed most populations share some traits, what is a perfect example on how "race" is constructed by homogeneization through time and therefore "purity" is an absurd concept: all "races" are mixed.
There is nothing wrong with the category "Negroid". Sub-Saharan Africans are genetically and morphologically closer to each other than to any other modern human group. But then again, both genetics and physical anthropology confirm the existence of sub-racial groups there, and you may just call them racial groups if you wish.
As for the issue of purity, I always use the word "pure" in a relative sense. There is no absolute purity but degrees of purity. For instance, Basques are among the purest Caucasoids today.
All that language of the early 20th century has way too many implications of apartheid, Jim Crow and then of course the Nazis and is not really scientific, as demonstrated in the famous "The Mismeasure of Man" and many other data we can see easily.
No sane person would put me and people with similar thoughts in the same category as the Nazis, apartheidists, etc. You always politicize discussions when I mention racial categories. You are so hyper-sensitive that you treat people who use racial categories as racist. By your definition, many mild people from the genetics and physical anthropology communities would be classified as racist. I sometimes feel like talking to a witch hunter or a McCarthyist when discussing with you.
Talking happily of "races" all the time is not acceptable in polite society, as they say, nor in the worst slums either. You seem to live in a bubble about this but I mean that even in Turkey your way of talking raises some eyebrows.
I very rarely talk of races in daily life, but in any case, I see nothing wrong with talking about races as long as we stick to the scientific facts and not abuse them. You cannot solve a problem by ignoring its causes. Denial of races and racial differences do no good to people who suffer from racial discrimination.
You pretend it is scientific terminology but it's not: and you do not normally find such terms in scientific literature anymore, at least not the one I read such as human population genetic studies. So if scientists find no problem (or even quite useful and convenient) in using other language, why would you claim that your extremely and obsessively racialized language is "scientific"?
DeleteYou are continuously making a point about your personal beliefs on the matter of human differences. You have demonstrated to be stubborn like a mule and refractory to others' opinions. You don't care and hide behind the word "scientific", which you abuse more than a Brezhnevian neo-Stalinist.
As I told you, if the racial categories are being much less frequently used in the scientific literature today, this is solely due to the efforts of the politically correctists and has nothing to do with any scientific development. Scientists today refrain from using racial categories solely due to the pressures of the politically correctist lobbies. The politically correctist lobbies are so strong and influential in the academic world today that scientists feel a strong need to submit to their demands in order to attain and retain tenures.
You made a promise on this matter anyhow (that was a condition to lift your ban) but you are not fulfilling it. So... what do we make of it?
I only promised not to us the term "hybrid" to designate human racial mixes in your blogs.
This comment has been removed by the author.
DeleteAs for the ADMIXTURE/STRUCTURE components issue, you have no proof that those components are as old as you assume. They are always found in populations that are in the transition zone between the races that constitute the building blocks of those components. This can't be a coincidence. Same with the Central Asian-specific component of Xing et al.
DeleteYou promised not to abuse the racialist language and try to avoid stepping on race-sensitives toes such as mine.
Delete"... if the racial categories are being much less frequently used in the scientific literature today, this is solely due to the efforts of the politically correctists"...
That's not completely true: it's a matter of scientific correctness as well. It is an issue not just on whether races are real or just subjective (there is a strong subjective component to them, of course) but also about the many nuances that racialist ideology seem unable to face. Racialists are trapped in a sink of ancient concepts that are clearly obsolete and at least partly incorrect, extremely contaminated by subjective and ethnocentric (when not outright racist) biases.
The very concept of race imposes a preconception to the multiplexed reality of population genetics: it is a virtual box that traps many in a wrongly limited kind of though. It's much better to think outside the box, mostly because the box is pretty much unreal. Racialism is for dummies, I say sincerely, and almost completely irrelevant and confusing for population genetics and general anthropology.
"As for the ADMIXTURE/STRUCTURE components issue, you have no proof that those components are as old as you assume."
DeleteYes I do have some proof: the Fst distances. You don't get inter-continental Fst distances without a long time behind.
"They are always found in populations that are in the transition zone"..
It has nothing to do with the transition zone: transitional, admixed components like the Ethiopian or the Fulani ones are equidistant to the two "racial" poles involved but at levels of intra-continental Fst values, not inter-continental ones. You haven't even looked at the matter: you are just confronting it without any meditation.
"Same with the Central Asian-specific component of Xing et al."
That's like the Ethiopian component: obviously mixed. But not at all like the South Moroccan or Red Sea components I found: these are genuine "other races" lost in time. There's a clear difference in Fst distances, which are almost double in the latter.
This comment has been removed by the author.
DeleteLuis, I strongly disagree with you on the race issue for reasons too many to elaborate that the thread would be completely drifted away from the main topic (already we have drifted too far). All I want to say for now is that I request you to be more tolerant towards ideas that you disagree with. I very well know what is racist and what is not, so please do not worry about me.
DeleteAs for the component issue, as I told you long ago, those high Fst components you mention are minor components even in the populations in which they peak and this alone is enough to explain their high Fst distances. So they are most likely artifacts of the peculiarities in the algorithms of the ADMIXTURE and STRUCTURE softwares.
"those high Fst components you mention are minor components even in the populations in which they peak"...
DeleteYeah.
"... and this alone is enough to explain their high Fst distances".
Nope. Other minor components (for example those caused by endogamy) don't behave like that.
They are not "artifacts", at least I do not see any reason to think so.
"All I want to say for now is that I request you to be more tolerant towards ideas that you disagree with".
All I want is that commenters in this blog do not unbury that jerk of Coon every other day just because. It's not a matter of racism (not only at least) but a matter of scientific seriousness.
You sound like a Stalinist talking of "scientific materialism", what is another nonsense by the way. Just because you introduce the word "science" arbitrarily it does not make it any more scientific. Coon is dead and his Nordocentric nonsense is as well (and yeah, he was extremely nordocentric and totally unable to discern differences inside populations other than his own).
Nope. Other minor components (for example those caused by endogamy) don't behave like that.
DeleteWhat minor components are you talking about?
They are not "artifacts", at least I do not see any reason to think so.
Most likely they are.
All I want is that commenters in this blog do not unbury that jerk of Coon every other day just because. It's not a matter of racism (not only at least) but a matter of scientific seriousness.
You sound like a Stalinist talking of "scientific materialism", what is another nonsense by the way. Just because you introduce the word "science" arbitrarily it does not make it any more scientific. Coon is dead and his Nordocentric nonsense is as well (and yeah, he was extremely nordocentric and totally unable to discern differences inside populations other than his own).
Who said I am a Coonist? Just because I use terms such as "Caucasoid" and "Mongoloid" and make a racial classification of modern human groups neither makes me a Coonist nor implies that my definitions of Caucasoid, Mongoloid, etc. and race are the same as those of Coon. Just as in our discussion of hybrids, you are fixating too much on words and ignoring their content and context.
Finally, scientific materialism has no resemblance (either by content or methodology) to what we are talking of here. So please do not distort the subject with such inappropriate comparisons.
"What minor components are you talking about?"
DeleteHave you bothered reading the relevant threads (linked previously), looking at the components found, the Fst distances? Obviously not judging on this redundant question (etcetera). In those same entries there are other minor components, both "racially mixed" and "racially pure" (in your kind of language, I would use "continental region" or "macro-population" instead of "race" because I have no idea how they relate to phenotype, if at all) which show intra-continental Fst values rel. to other components. These two I highlighted do not: the show inter-continental Fst values. Just go over there and study the results before you write any other line. Thanks.
You can also experiment with ADMIXTURE yourself: it's rather easy (although it may take some time). All I say is: don't stop at low K values (too shallow) and pay attention to Fst distances displayed by the components, because those values are very important.
Have you bothered reading the relevant threads (linked previously), looking at the components found, the Fst distances? Obviously not judging on this redundant question (etcetera). In those same entries there are other minor components, both "racially mixed" and "racially pure" (in your kind of language, I would use "continental region" or "macro-population" instead of "race" because I have no idea how they relate to phenotype, if at all) which show intra-continental Fst values rel. to other components. These two I highlighted do not: the show inter-continental Fst values. Just go over there and study the results before you write any other line. Thanks.
DeleteYou can also experiment with ADMIXTURE yourself: it's rather easy (although it may take some time). All I say is: don't stop at low K values (too shallow) and pay attention to Fst distances displayed by the components, because those values are very important.
Of course I read them. None of the "Tunisian", "Mozabite", "Kalash" and the Xing et al. "Central Asian" ("Khanty" to be more precise) components are minor components in the populations they peak. You obviously do not have a clue about how the algorithms work. All things being equal, the lower the level of the component in the peak population, the higher the Fst distance of the component to the other components.
Please, if you wish to continue this discussion head to:
Deletehttp://forwhattheywereweare.blogspot.com/2011/12/north-african-genetics-through-prism-of.html
Which is the relevant thread.
The three Tunisian components as well as others are all minor in their peak population and they produce intra-continental values. It seems it's you who have no idea how the algorithms work in practice. It also seems that you are not looking at the data I produced but at other works.
When I look at Derenko phylogenetic trees, I see that the clusters are the following:
ReplyDeleteC1a: Southern Siberia, East Asia
C4a1a: Southern Siberia, East Asia
C4a1b: India
C4a2a: Southern Siberia, India
C4a2b: India
C4b: Southern Siberia, Northeastern Asia
C4 (unspecified): Europe
C4e: Southern Siberia
C5a: Southern Siberia
C5 (uspecified): East Asia
C5b: Southern Siberia
C5c: Southern Siberia, Europe
C5d: Southern Siberia
C7a: India
C7b: India
C7 (unspecified): East Asia
D3: Central and Southern Siberia
D4b1a1: East Asia
D4b1a2: Northeast Asia, Central and Southern Siberia, East Asia
D4b1b: East Asia
D4b2a + D4b2b: East Asia
D4b2c: India
D4b2 (not specified): East Asia
D4b2d: East Asia
D4c + D4d: East Asia
D4e1: America, Europe, East Asia (in particular Korea)
D4e3: India
D4e2: East Asia
D4e4: Central and Southern Siberia, Russia
D4e5: Central and Southern Siberia, East Asia
D4f: Central and Southern Siberia, East Asia (in particular Japan)
D4g + D4h + D4i: East Asia
D4j1: India
D4j2 + D4j4 + D4j5 + D4j6 + D4j7: Central and Southern Siberia
D4j3 + D4j8 + D4j9 + D4j10 + D4j (unspecified): East Asia
D4 (unspecified): Central and Southern Siberia
D4k’o’p: East Asia
D4l + D4m: East Asia, Central and Southern Siberia
D4n: East Asia
D4q: India
D4a: East Asia
D5a’b + D5c: East Asia
D6: East Asia
I know that your fundamental point is the migrations from South to North and ancient out of Indo-China migration. I believe that there were more focus areas and multiple routes.
This comment has been removed by the author.
ReplyDeleteAfter all this pondering about, I came up with another solution. In this model, M from which M8 derives comes from East Asia, but it still means a completely new way of thinking about the human migrations and it resolves many oddities. However, I do not want to argue over it yet. :-) I prefer to wait for new research to possibly back up this model.
ReplyDeleteM almost certainly expanded from South Asia beginning some 100 Ka ago, although it sent many "tentacles" to East Asia, in agreement with the fast migration model. One of those many East Asian branches was M8. There are many others but there are even more in South Asia.
DeleteIn the end, Bronze Age East-Asian MtDNA has been found in Altai proper. I found this study "Bronze Mitochondrial DNA of a late neolithic woman from Kaminnaia cave (Gorny Altai)". It seems to be the same place where Neolithic H/K haplogroup was found. It is again pay per view, but the abstract says
ReplyDelete"The results of molecular genetic analysis of mitochondrial DNA of a Late Neolithic woman (middle of the 4th millennium BC) from Kaminnaia cave (Gorny Altai) are presented. It was determined that the studied sample belongs to the East Eurasian A4 haplogroup. A phylogeographic analysis indicates an informativity of the A4 variants for the reconstruction of early stages of ethnogenetic processes on the south of Western and Eastern Siberia."
I am not at all sure about it, but two Baraba Late Krotovo and Andronovo samples look like A4. By contrast, the Earliest samples seem to be A10 which is said to be autochtonous to Baraba area.
It should be "Mitochondrial DNA of a late neolithic woman from Kaminnaia cave (Gorny Altai)"!
ReplyDeleteNot a suprise the 7,500ybp C1 in Russia near Finnish border is defintley connected with the spread of Mongliod Y DNA N1c1 in that area, Uralic languages, and Kunda culture. So there was Mongliod inter marraige also not a surprise it is shocking how popular Caucasin mtDNa haplogroups are from Chinese dyntic samples. We know it is from inter marriage with Indo Iranians and other Indo Europeans who lived around west china for almost 5,000 years like Tocherians.
ReplyDeleteCentral Asia believe it or not in the bronze and iron age was mainly white. The Indo Iranian tribes like Sycthians and their ancient bronze age cultures like Andronovo, Sinshta, Abashevo dominated central asai from about 4,500-2,000ybp. Their DNA samples show Y DNA R1a1(most likley Indo Iranian branch r1a1a1b2 Z93) typical west Eurasian(Europe and mid east) mtDNA haplogroups more typically European like U5, What surprised me is they had about 10% T1 which is almost non exitnct in Europe almost all T is T2. They had pale skin and mainly fair hair and eyes no doubt they were Europeans. And they match ancient Greek, Roman, and Chinese descriptions of these tribes. Groups like Sycthians were the main people of central Asia.
The Turks the main ancestors of modern central asians arrived around 300-400ad and conquered the Indo Iranians up to the middle ages. So central Asia has a complicated history. I don't think Caucasian mtDNa in dynastic China way above modern average and Mongliod mtDNA in Neolithic and bronze age Ukraine way above modern average means that much so people inter married big deal. Figuring out when this inter marriage happened and how long ago is important bit it should not be a surprise.
MtDNA C1 is not found in Uralic people. Its frequency is at the moment 0% everywhere to my knowledge, so I am not at all sure about that connection.
ReplyDeleteWell, being white is not a privilege of Caucasoids, and many Caucasoids are not very white. Northeastern Asians are usually quite pale, e.g. the Japanese.
People have been mixing with each other from the beginning of time and there has never really been any pure race, or perhaps for a short while when a group broke out on a virgin territory, but in the Palaeolithic time depth, in that case, there were usually other species on the way.
The Caucasoids or Mongoloids are not two uniform races that have had both their origin in one narrow area and then spread to the whole Eurasia respectively. This is evident when you look at the yDNA tree, MNOPS is widespread in whole Eurasia and Indonesia, and is also common in Africa.
there has never really been any pure race
ReplyDeleteNo one has claimed in this blog AFAIK that races have to be pure. Purity is not in the definition of race. All that is required for the formation of races is long term relative homogenization through mixing in a certain territory (not necessarily a small territory) and relative long term long term isolation from other territories.
The Caucasoids or Mongoloids are not two uniform races that have had both their origin in one narrow area and then spread to the whole Eurasia respectively.
Tell that to Terry. According to him the Mongoloid race originated in a narrow piece of land and expanded from there.
This is evident when you look at the yDNA tree, MNOPS is widespread in whole Eurasia and Indonesia, and is also common in Africa.
I would not put forward Y-chromosomal haplogroups as examples to the non-purity of races, as haplogroups, and especially Y-chromosomal haplogroups, can easily swap race through intense mixing with a race other than their original race without leaving any detectable signature of their original race so that they may have no detectable effect in the dilution of the level of purity of their new race or a branch of their new race.
What a boredom!
Delete"the definition of race"
Which definition of race?! Being a human construct with only very limited scientific support, this concept has no standardized definition. It seems to come from Italian razza and nobody knows what it originally meant nor where it comes from. In Romance it's mostly applied to animals. So pitbull is a race (i.e. breed) of dogs, Merino is a race of sheep, etc. Humans are not bred like cattle so it's not really easy to apply.
"Tell that to Terry".
That guy is banned for being a polemicist jerk, twisting and manipulating (or even just making up) others' words and ideas, etc.
"... the Mongoloid race originated in a narrow piece of land and expanded from there".
Not likely at all because genetics simply does not support that model, certainly not for a Neolithic time-frame. But in any case it does not seem like the very concept of "Mongoloid" is too solid either.
Which definition of race?! Being a human construct with only very limited scientific support, this concept has no standardized definition. It seems to come from Italian razza and nobody knows what it originally meant nor where it comes from. In Romance it's mostly applied to animals. So pitbull is a race (i.e. breed) of dogs, Merino is a race of sheep, etc. Humans are not bred like cattle so it's not really easy to apply.
DeleteWe are not discussing the etymology of the word "race". The only kind of race we are talking about here is the scientific-based race. Not just race but all of taxa (species, order, genus, etc.) are human constructs, but that does not change the fact that they are all scientific-based. Race is not an arbitrary concept, it has a concrete basis in biology.
Not likely at all because genetics simply does not support that model, certainly not for a Neolithic time-frame. But in any case it does not seem like the very concept of "Mongoloid" is too solid either.
It seems I neither agree with you nor with Terry on the structure of the Mongoloid race. But this is a side issue, and we have already drifted too far from the main subject of this thread. So we'd better stop here.
"... the scientific-based race".
DeleteThere's no such thing. Scientific classification with any meaning stops at species level (and even there there are some doubts). Categories like subspecies, variants, breeds, races and cultivars are mostly subjective, even if sometimes conventional. Reality does not come conveniently boxed for classification purposes, much less downstream of the species level.
"Race is not an arbitrary concept, it has a concrete basis in biology".
Which one? I never heard of such "concrete basis". Reality is much more more clinal and irregular than you seem to believe.
In fact, when you ask racialists to describe their neatly packed boxes, there's always a total indefinition in the details. There may be some archetypes but real people almost never conform to them: they are just very broad and diffuse tendencies. Instead when you compare a lion and tiger, there is no doubt: each individual fully complies with its species characteristics (unless hybrid, of course).
There's no such thing. Scientific classification with any meaning stops at species level (and even there there are some doubts). Categories like subspecies, variants, breeds, races and cultivars are mostly subjective, even if sometimes conventional. Reality does not come conveniently boxed for classification purposes, much less downstream of the species level.
DeleteWhich one? I never heard of such "concrete basis". Reality is much more more clinal and irregular than you seem to believe.
In fact, when you ask racialists to describe their neatly packed boxes, there's always a total indefinition in the details. There may be some archetypes but real people almost never conform to them: they are just very broad and diffuse tendencies. Instead when you compare a lion and tiger, there is no doubt: each individual fully complies with its species characteristics (unless hybrid, of course).
The above-species level taxonomic ranks are no less artificial than the below-species level taxonomic ranks. The least artificial of the taxonomic ranks is species. Basically, the closer to the species level, the less artifial a taxonomic rank. Races are not that far from the species level and thus not so much artificial.
"Races are not that far from the species level and thus not so much artificial. "
DeleteRaces are not any taxonomic level, they are not close to the notion of species (that's genus) and they are highly artificial, based on nothing but appearances. The genetic legitimacy of the notion of "race" or more precisely continental grouping is about less than 10% of the variance: it's ridiculously low (80% is merely individual and another 10% is inside your racial or continental groupings, among more localized groups).
In any case I'm getting very bored of having to put up with this "medieval" level of discussion, really.
I do not diagree with the concept of long term relative homogenization through mixing in a certain territory leading to a shared physiognomy, but I would add that the border lines are always blurred.
ReplyDeleteI do not disagree with Terry either if the point is that at a certain point of time, e.g. in connection with the cultivation of rice, certain physiognomy and features spread to a great extent in East Asia. However, my point was that these features resulted from a previous mixture of multiple elements. East Asia is genetically diverse: there are yDNA D, C and MNOPS, even F, and as for mitochondrial DNA, there are mtDNA lines M, N and R which may have arrived along several routes. In addition, we have the possible archaic mixtures with other subspecies of Homo.
Europe is also diverse: there are yDNA G, IJ, E, R, Q and N, as well as some T, and as for mitochondrial DNA, there are mtDNA lines N and R, and it seems that generally haplogroups have arrived to Europe from the East. It also seems that in Europe there was more mtDNA L and M in prehistoric times than today.
"It also seems that in Europe there was more mtDNA L and M in prehistoric times than today".
DeleteNot much anyhow. Just some C1 in the NE and some L(xR) in West Iberia, which may in one case be L3 (the other may well be N*). Almost every single pre-Neolithic lineage of Europe is R (either U or R(xU), including H and JT*, but R in any case).
but I would add that the border lines are always blurred
DeleteThey are what are commonly called hybrid zones. Though Luis forbade me to use the word "hybrid" in denoting racial mixes in his blogs. So I stop here.
Not just areas of admixture but individuals in the core areas are also extremely varied and with weird tendencies that may or not approach other such categories. South Sudanese with sinodonty and slanted eyes, Chinese without epicanthic fold, Scandinavians with broad nose and epicanthic fold, Australian Aborigines that look like dark Vikings, etc. It's blurry all around, even in the core zones.
DeleteA good example are Native Americans, who are extremely diverse even if they have a marked and relatively recent founder effect. And I'm talking only of the unmixed ones, of course. You apparently seem unable to see the massive differences inside your artificious groupings, but I do and they are just brutal and often totally unexpected.
"They are what are commonly called hybrid zones. Though Luis forbade me to use the word "hybrid" in denoting racial mixes in his blogs".
DeleteI'm being extremely patient. But my patience is not infinite. I truly believe that you should look for other forums and quit this one. Because you don't belong here, really.
Not just areas of admixture but individuals in the core areas are also extremely varied and with weird tendencies that may or not approach other such categories. South Sudanese with sinodonty and slanted eyes, Chinese without epicanthic fold, Scandinavians with broad nose and epicanthic fold, Australian Aborigines that look like dark Vikings, etc. It's blurry all around, even in the core zones.
DeleteA good example are Native Americans, who are extremely diverse even if they have a marked and relatively recent founder effect. And I'm talking only of the unmixed ones, of course. You apparently seem unable to see the massive differences inside your artificious groupings, but I do and they are just brutal and often totally unexpected.
If you focus on single traits, you cannot see the big picture. Races are fairly homogeneous entities in their overall race-specific morphology. Moreover, races are more homogeneous entities genetically than they are morphologically. Genetics strengthened, not weakened, the biological basis of races.
I'm being extremely patient. But my patience is not infinite. I truly believe that you should look for other forums and quit this one. Because you don't belong here, really.
DeleteIt is you who started this race debate. My first posts in this thread were all on topic posts.
This comment has been removed by a blog administrator.
Delete"It is you who started this race debate."
DeleteYou use and abuse racialist terms in an anti-racist blog and you expect not to be discussed (or worse)? You are an arrogant prick!
I'm going to review all my policies, enable comment moderation for the time being (not just for you: but your language seems to act as vortex of attraction of the worst trolls) and surely enter into blogging hiatus until I can finish a migration to WordPress where I can more easily moderate commenters individually, it seems (also no .es anymore in my address).
But you bet that apology of racialism and excessive racialist language will be banned when I return to normality. I'll ponder carefully how to outline that.
This comment has been removed by a blog administrator.
ReplyDeleteThis comment has been removed by a blog administrator.
ReplyDeleteThis comment has been removed by a blog administrator.
ReplyDeleteThis comment has been removed by a blog administrator.
ReplyDeletemtDNA and yDNA ancestry should be equally important because we receive half of our genes from the mother and half from the father. However, it seems that men have covered longer distances in a shorter time frame than women, i.e. when men are moving they tend to marry local women. In this way, the original genetic ancestry that is still visible in the yDNA marker nearly vanishes in the new environments further and further away. When people moved into new territories, e.g. at the end of the Ice Age, there may have been less admixture, if the new areas were uninhabited.
ReplyDeleteI agree with you that it is too simplicistic to reduce all variation to a tripartite division people A - hybrid zone – people B. All areas have their own particular history and hybridization and phenotype fixing must have happened many times in history in many different places. Moreover, on the basis of that paper I mentioned above it seems that, just as you said, Central Asians have their own phenotype that formed during the Neolithic and is not East-Asian or European.
I apologize for your blog problems. Although we do not always agree, I enjoy having these discussions with you.
"I apologize for your blog problems."
DeleteYou should not: it's not your fault. At least I don't think so.
"Central Asians have their own phenotype that formed during the Neolithic and is not East-Asian or European".
Phenotype or genotype? Neither or intermediate? Why Neolithic and not earlier. Why Central Asia and not West Siberia. I do have all those kind of questions open and that's why I tend to stick to mtDNA: it's simpler.
Good questions! I said phenotype because I wanted to underline that I am not talking about haplogroups, but if phenotype includes also the environmental factors, that is not what I wanted to say.
ReplyDeleteThat Gorny Altai paper uses the word "intermediate", but argues against simple hybridization. Recent hybrids should be very variable and heterogenous, but Neolithic Central Asians seem to have a fixed physiognomy that must have formed earlier. I said Neolithic because I thought that the population numbers grew fast at that time. During the Ice Age there must have been very few people. After the Ice Age the Western element may have come to Central Asia from Samara area and West-Asian element from the south from Tadjikista/Kirgistan. The Eastern element may be related to yDNA N, and Y-DNA Q may have persisted in small pockets all through the Ice Age. I think that all these layers are there and contributed to the fixing of physiognomy before the Iron Age.
As for this discussion on mtDNA C, you might find this interesting: ”The most amazing result concerns sample No. 6. This individual is characterized by Caucasoid cranial features and by a haplotype belonging to an Eastern Eurasian haplogroup C. A similar discrepancy between cranial and genetic data was observed in the Pazyryk people buried at Verkh-Kaldzhin, where the mtDNA haplotype was similar.”
This is also interesting: ”Haplotype 16223-16242-16290-16319 of haplogroup A (No. 5) has not been detected among either Eastern Central Europeans or Finno-Ugrians, but its frequency among the Selkups is 3.3 %.” … ”The presence of the ancestral haplotype of haplogroup A among the Pazyryk people might be due to a migration of one or several groups of early nomads to Gorny Altai from the southwest or south before the 4th or 3rd century BC. This idea is supported by the fact that the mtDNA haplotype of the individual buried at Ak-Alakha mound 5 (sample No. 5) belongs to a distinctly “Eastern” haplogroup A, and his cranial morphology is unambiguously Mongoloid.”
The haplotype of Gorny Altai sample 5 seems to be A8 and it belongs to the other main branch of haplogroup A than A4.
"This individual is characterized by Caucasoid cranial features and by a haplotype belonging to an Eastern Eurasian haplogroup C. A similar discrepancy between cranial and genetic data was observed in the Pazyryk people buried at Verkh-Kaldzhin, where the mtDNA haplotype was similar".
DeleteAlright. It seems to make sense, especially if mtDNA C had been in the West (Western Siberia, Eastern Europe) since Epipaleolithic. But of course, in few generations, admixture can totally erase the phenotype of either uniparental lineage (or even, in the right circumstances, both).
That is exactly what I think, i.e. MtDNA C has been in Eastern Europe very early, much earlier than most other East-Asian haplotypes.
ReplyDeleteBy the way, this C seems to be close to C4a1. HV1 mutations of standard C4a1 are 16129-16223-16298-16327. This Gorny Altai haplotype has an additional 16093. Today it is said to be found in Altaians, Mansi and Tuvinians. The standard C4a1 haplotype seems to be found in Baraba Late Krotovo and Andronovo specimens.
ReplyDeleteXiaohe C4 haplotype differs from Gorny Altai haplotype with 2 mutations. Four of Dneper Donetz haplotypes are also C4a but they are C4a3, C4a6 and C4a2'3'4.
Now this European-specific C4a types seem to have disappeared but instead we have C5 haplotypes. I tried to Google for C4b and it may be found in Tatars, Kurds. Please correct if know better!
I looked again that Derenko phylogenetic tree of haplogroup C. C4a has two branches: C4a1 and C4a2 (C4a2'3'4?). Both lineages have ended up in India. Unfortunately, I could not trace their location in India, but I suppose they are not found in Tibeto-Burmans who harbour C7. The age of the Indian branch is older than the age of the European branch (Mansi?) and is c. 8-11 kya. I now wonder if this haplotype went to India with yDNA Q or R1a. However, it seems that it must have come to India directly from Central Asia.
ReplyDeleteC4a2 (C4a2'3'4?) has two Indian sub-branches. One of them is older than C4a1 (c. 14-20 kya) and the other is c. 8-16 kya. If C4a2'3'4 was in Dneper Donetz 7000 years ybp, these C4a2 branches may have entered India with the Indo-European languages, if we assume that the ages above are overestimations. C4b and C5 have not been found in India.
By contrast, on the West Coast of America a rare Native American C4c lineage has been found and it seems to be a recent arrival from Siberia.
"However, it seems that it must have come to India directly from Central Asia".
DeleteWhy? The most normal influx of oriental lineages in India is via the Brahmaputra corridor (itself a transition zone between South and East Asia), probably all them Neolithic arrivals. A local "age estimate" may mean absolutely nothing, not just because it is a mere guesstimate but also because it may reflect the relative "age" at the point of origin and not that particular destination, where the various lineages used for the measure may well have arrived already formed.
For example, if you measured the "age" of U5 in the USA, for example, it should not produce too different results than if measured in Europe. However in reality U5 in Europe is probably many tens of thousands years old and in the USA just a few centuries.
Yes, I know that, but I use the age as in indication of the haplotype's age compared to its sister haplotypes. That should be meaningful.
DeleteIt is true, now, when I look at the ages, C4a2b is the oldest. It seems to be too old to be linked with Indo-European languages. This could mean that R1a was there in the area between Kazakhstan and India before IE-languages and there were contacts between South and North, which led to the early arrival of C4a2b to India. The other C4a sub-branches are a better match.
I said "directly from Central Asia" because C4a1 was not found in Ukraine, and instead was found in Baraba and Gorny Altai and is today found in Altaians, Tuvinians, Mansi, Tibetans. Of course, this is not conclusive proof.
Interesting discussion. I am from south India and my mtdna is C4a1. No historical connections whatsoever with eastern India. Not to say that my mtDNA did not migrate from there(I don't know). But my sample is on Genbank: JN392434.
DeleteIt'd be nice to understand better the distribution of C4a in South Asia (and East Asia too). From what SB says, we could well be imagining the wrong scenario altogether, and therefore I'd be jumping to the wrong conclusions.
DeleteIf you look on this page: http://ianlogan.co.uk/sequences_by_group/c4a_genbank_sequences.htm
DeleteExcept for my sample(JN392434), all the other samples listed by Rao are from East India (Tibeto-Burman populations).
My sample on Genbank, sems to hae closest affinity with a Ukrainian sample, instead of NE Indian samples on Genbank: http://eng.molgen.org/download/file.php?id=250&mode=view
http://eng.molgen.org/viewtopic.php?p=9602#p9602
I thought I recalled some genetic-based maps which show that SE Asian genetic influence reached (at low levels) as far South as Tamil Nadu but I could not recall which author, so I did not comment yesterday. However later I found myself browsing Harappa Ancestry Project for another reason and found some such maps, all based on autosomal DNA, for example:
Deletehttp://www.harappadna.org/2011/03/south-asian-map/
So I guess it does make sense that some SE Asian lineages ended up in your area SB.
Notice also that when a Siberian component is considered, it's influence in South Asia is almost nil and concentrated towards the Pashtun area (Af-Pak political border). So I think that we can safely discard any meaningful Siberian influence in South Asia, especially far from Pakistan.
I belong to the TamilNadu Brahmin group. I doubt that Siberian or any other component on autosomal DNA should be linked with haplogroups considering that it is very easy to wash any trace of them away or inherit something that has nothing to do with mtDNA.
DeleteMaju, I do have trace Siberian and so does my dad from Geno 2.0 http://www.anthrogenica.com/showthread.php?719-Dad-s-Geno-2-results&p=4492&viewfull=1#post4492
DeleteMaju, I misspoke. On Harappa DNA project, I (HRP142) have 1.06% NE asian, 0.97% SE asian, and 0.23% Siberian. They may all be noise to begin with.
DeleteAlso, NE asians and siberiasn populations themselves have SE asian like componments in them. Hence it is impossible to use these breakdown in autosomal affinity to figure out where one inherited a component (other than in very special cases and populations whose history is quite well known).
Infact, I think the only way autosomal components can be used in general, is to assign an individual to a population based on general affinity, because the names of the various components are not representative to begin with (one could call them rabbit, tiger, lemon etc. and it still would make the same impact).
If it's an old lineage it does not have to be directly connected with any noticeable autosomal component. Assuming a single ancestor from that ethnic background, by the 8th generation (less than three centuries) it already becomes diluted under 1% values and by the 16th generation (some five centuries ago) it is just one of 32,768 ancestors, etc.
DeleteOf course, this is less likely to happen at population level, especially re. mtDNA (historically women usually traveled less and didn't go around invading others on their own, although they did tend to travel at local range because of patrilocality), but, being an exceptional linage in your area and population group (a so-called "erratic"), there's no particular reason to expect autosomal admixture from the same origin at noticeable level. As you say, it may well be just noise.
But the statistical fact is that in SE India there seems to be some minor SE Asian genetic influence, maybe from Austroasiatic gradual "genetic infiltration" from the Orissa area or maybe because of the historically important connections of South India with SE Asia (Viyajanagar and such) - or both.
I'm not sure how that applies to the Brahmin case, because these seem to have (in most cases at least) much more Northern Indian ancestry than the bulk of the population. Would it be the case of your remote maternal ancestor, it's even possible that she descended from Tibeto-Burman groups of the Himalaya area. I can't say.
It is frustrating that many researchers have not bothered to resolve any subclades of haplogroup C in their samples of mtDNA. One of the most detailed analyses I have seen recently, from Fedorova et al. 2013, shows that haplogroup C4a is very common in the Turkic Yakuts (28/148 = 0.189 Northern/Arctic Yakuts, 25/111 = 0.225 Western/Vilyuy Yakuts, 39/164 = 0.238 Eastern/circum-Yakutsk Yakuts) and the Northern Tungusic Evenks (21/125 = 0.168) and Evens (19/105 = 0.181) of Yakutia in eastern Siberia. Every one of these populations contains representatives of both C4a1 and C4a2, without any significant cline (though the total of {C4a1+C4a2} tends to be slightly higher among the Turkic Yakuts than among the Tungusic Evenks or Evens).
ReplyDeleteC4a is also very common in the Turkic-speaking Dolgans of the Arctic coast in Taymyria and northwestern Yakutia (35/154 = 0.227), though this small population is dominated by the C4a1c subclade (33/154 = 0.214) with only minor representation of any other subclade of C4a (2/154 = 0.013 C4a2).
I think the high frequency of C4a in the Evenks and Evens of Yakutia is fairly good evidence for the antiquity of this haplogroup in eastern Siberia. It is difficult to imagine that these Northern Tungusic populations (and especially the Evens, who as recently as Soviet times were basically hunter-gatherers with the additional pastoralist habit of tending to herds of reindeer) might have assimilated such a significant number of females from a foreign population. If one really wanted to consider a substrate scenario, the only known population that conceivably might have served as a substrate to the Evenks and Evens in Yakutia is the Yukaghirs, but the tested remnants of the Yukaghirs from northeastern Yakutia do not exhibit haplogroup C4a mtDNA with such high frequency (only 1/22 = 0.045 C4a1c and 1/22 = 0.045 C4a2 for a total of 2/22 = 0.091 C4a).
It is frustrating that many researchers have not bothered to resolve any subclades of haplogroup C in their samples of mtDNA. One of the most detailed analyses I have seen recently, from Fedorova et al. 2013, shows that haplogroup C4a is very common in the Turkic Yakuts (28/148 = 0.189 Northern/Arctic Yakuts, 25/111 = 0.225 Western/Vilyuy Yakuts, 39/164 = 0.238 Eastern/circum-Yakutsk Yakuts) and the Northern Tungusic Evenks (21/125 = 0.168) and Evens (19/105 = 0.181) of Yakutia in eastern Siberia. Every one of these populations contains representatives of both C4a1 and C4a2, without any significant cline (though the total of {C4a1+C4a2} tends to be slightly higher among the Turkic Yakuts than among the Tungusic Evenks or Evens).
ReplyDeleteC4a is also very common in the Turkic-speaking Dolgans of the Arctic coast in Taymyria and northwestern Yakutia (35/154 = 0.227), though this small population is dominated by the C4a1c subclade (33/154 = 0.214) with only minor representation of any other subclade of C4a (2/154 = 0.013 C4a2).
I think the high frequency of C4a in the Evenks and Evens of Yakutia is fairly good evidence for the antiquity of this haplogroup in eastern Siberia. It is difficult to imagine that these Northern Tungusic populations (and especially the Evens, who as recently as Soviet times were basically hunter-gatherers with the additional pastoralist habit of tending to herds of reindeer) might have assimilated such a significant number of females from a foreign population. If one really wanted to consider a substrate scenario, the only known population that conceivably might have served as a substrate to the Evenks and Evens in Yakutia is the Yukaghirs, but the tested remnants of the Yukaghirs from northeastern Yakutia do not exhibit haplogroup C4a mtDNA with such high frequency (only 1/22 = 0.045 C4a1c and 1/22 = 0.045 C4a2 for a total of 2/22 = 0.091 C4a).
SB Interesting link!
ReplyDeleteI do not know if Derenko phylogenetic tree and Ian Logan haplotype codes are identical. According to Ian Logan page, you are C4a1 and that type is found in a Buryat and Teleut. Rao's Tibeto-Burman C4a1a cluster is under it. All other C4a1'5 clusters look very Siberian. Instead, C4a2'3'4 cluster looks very Indian and it is very intriguing if all of it is from Tibeto-Burmans.
The frequencies of C4a and C4b decrease in Amur and Beringia. According to one of my papers, the frequency of C4a is highest in Evenki (43.4%) and Tofalar (39.1%) and lowest in Ulchi, Udegey, Negidal, Koryak and non-existant in Nivkhi, Itelmen and Chukchi. The frequency of C4b is highest in Yukaghirs, Nganasan and Evenki and lowest in Ket, Ulchi, Negidal and Chukchi. I would say that C4a and C4b originated in Baikal area.
Please, do not think that I suggest that C4a arose in Europe. It is however intriguing that it was in Dneper Donets 6000-7000 ybp and seems to be linked with this Kurgan phenomenon.
Moreover, if both C4a1 and C4a2 come from Siberia and C4a1 and C4a2 are old in India and typical of TIbeto-Burmans, I would really like to know the yDNA that accompanied them.
"Moreover, if both C4a1 and C4a2 come from Siberia and C4a1 and C4a2 are old in India and typical of TIbeto-Burmans, I would really like to know the yDNA that accompanied them".
DeleteGrosso modo, O2a corresponds to the Austroasiatic wave (more important and probably quite older) and O3 (O3-what?) to the Tibeto-Burman wave (concentrated in the Himalayan area). Ebizur or whoever, feel free to improve this info-bit.
There is no O2a nor O3 in south India and in my ethnic group. Almost everyone I share with on 23and me have the following groups: R1a close to 50%, L close to 15%, G2a3 10%, H1a 10%, C5 and others 5%.
DeleteKristiina, yes they all appear to be from that paper. Genbank confirms it. I think some form of C4a is common in Tibeto-Burman populations in Eastern India. But I don't think the C4a in the south or west of India are linked to them. This is purely a guess, and more data if it ever shows up will clarify the situation!
Delete"There is no O2a nor O3 in south India"...
DeleteI would not be so sure: http://blogs.discovermagazine.com/gnxp/2010/10/sons-of-the-conquerers-the-story-of-india/
Scroll down to the O2a maps, please: there seems to be slight but above zero presence of this lineage in SE India and Sri Lanka.
Whatever the case, I did not mean that East Asian maternal and paternal lineages traveled together to every corner of South Asia where they are found as rare "erratics", just the general pattern of spread into South Asia as a whole.
→ http://forwhattheywereweare.blogspot.com.es/2012/12/y-dna-from-tamils-and-south-indian.html
DeleteUnspecified O-M175 was found by Kumar 2012 in the following ethnic groups of Tamil Nadu:
· Thoda¹: 3.85%
· Parayar NTN²: 1.92%
· Parayar²: 4.17%
· Paravar²: 7.41%
· Yadhava³: 0.93%
· Vaniyar NTN³: 2.08%
· Nadar Cape³: 1.02%
· Maravar³: 1.25%
· Tamil Jains⁴: 2.11%
· Vadama⁵: 3.17%
Classes attributed in the paper to each micro-ethnicity:
¹ Tribal Hill Cremating
² Scheduled caste
³ Dry land farmers
⁴ Artisan-warriors
⁵ Brahmin (irrigation farmers)
The plausible oldest "classes": Hill tribe foragers and Hill tribe Kannada, totally lacked the lineage, but all the other classes had it to some extent (always minor). It is most likely O2a, at least most of it.
Yes, I agree it may definitely be from any source (in situ, from SE asia or from Central asia) but there is no way to figure out anymore.
DeleteHere is an interesting paper about haplogroup M and subgroups in India: http://ns1.ias.ac.in/jgenet/Vol88No1/127.pdf
The Kattunaiken (a tribe in Kerala in south India) have been found with mtDNA C, but it is unclear what subgroup they belong to.
I am highly sceptical of some of the typing in the Arunkumar paper because a lot of the STRs do not correspond with the haplotypes defined. But if we deduce that East asian mtDNA lineages go along with East asian Y-DNA, then this assumption is broken in the Kattunaikens (or kattynayakans) who are Hill tribes in the Kannada/Kerala area, because mtDNA C is already published in them.
DeleteHaplogroups are NOT defined by STRs, just by SNPs. STRs are not really too reliable and should only be used when no SNP data exist. If the STRs are very divergent, then they probably indicate a distinct branch under O or O2a but I would not dare to question ever SNP-based haplogroup assignation only on STR eyebrow-raising. SNPs (with extremely rare exceptions) only happened once in human history (and even in hominid history, etc.) The chances of SNP identification of haplogroup being wrong are almost zero.
Delete"Here is an interesting paper about haplogroup M"...
Thanks. I'm not sure to have read it earlier. However it does not seem too helpful re. mtDNA C4a (not even a generic mention to C or M8 for what I can see).
"The Kattunaiken (a tribe in Kerala in south India) have been found with mtDNA C, but it is unclear what subgroup they belong to".
DeleteWhere did you get that from? It does not seem from the paper you linked to. At least I can't find it.
That paper references this one:
Deletehttp://www.hygienecentral.org.uk/pdf/Cordaux%20%20European.pdf
look at table 3. "M-C" represents haplogroup C in tribal populations around India.
Alright, thanks. Quite interesting. Notably it seems like the main East Asian mtDNA lineages spotted in India (at least in Cordaux' paper) are A, F and C. However only C made it into the South (some mysterious and small founder effect, I guess).
DeleteIf you trust SNP typing on the Arun kumar paper, do you really think there is Haplogroup E-M96 in south Indian castes; J*-M304 and Q-M242 in south Indian tribes?? What could explain that?
DeleteI'd believe E should be from West Asia and arrived along with other such West Asian lineages in the Neolithic. E is an important haplogroup associated with Neolithic in Europe as well. E is only found in dry land farmers, what makes good sense with this explanation. There is no E in Tamil Nadu tribals, at least not per Kumar 2012.
DeleteJ* should be J1 and a plausible explanation is that it's related also with Neolithic flows from West Asia. However it is found in one tribe (as well as a scheduled caste) so maybe it represents pre-Neolithic flows of some sort (???), if so it should be J(xJ1, J2), I think.
As for Q, which is indeed notoriously present among tribal groups, it should be a remnant of ancient (basal) Q, which originated somewhere between South, West and Central Asia (and only later expanded to Siberia and America via Altai). Not sure if you are aware but P, Q and R look original from South Asia, or at most nearby regions like Iran or Central Asia (Q?, surely R1). I'm talking here of a very remote past, of course, some 60 Ka ago (very roughly). This kind of data, after proper processing, may well add weight to the hypothesis of Q ultimately originating in South Asia.
Are Rao 2009 haplotypes on Ian Logan site these?:
ReplyDeleteKumar, S., R. R. Ravuri, P. Koneru, B. P. Urade, B. N. Sarkar, A. Chandrasekar, and V. R. Rao. 2009. "Reconstructing Indian-Australian phylogenetic link." BMC Evolutionary Biology Vol 9:173.
Accessed on the 15/10/2010, available at
http://www.biomedcentral.com/content/pdf/1471-2148-9-173.pdf
They say that they analyzed 966-mitochondrial genomes from 26 relic tribes of India and identified seven genomes, which share two synonymous polymorphisms with the M42 haplogroup, which is specific to Australian Aborigines.
Both M42 and its "sister" M74 are found in South Asia, however, outside of the subcontinent, M42 is only found in Australia and M74 only in East Asia. M42'74 therefore suggests an "Eurasian expansion" time-frame with origin in South Asia.
DeleteThe references are somewhere in this page: http://ourorigins.wikia.com/wiki/MtDNA_haplogroup_M
DeleteI made a review of C4a haplotypes on the basis of Ian Logan site and a big number of genetic studies. Age estimations are form Derenko paper.
ReplyDeleteC4a1 is found in Tibet (3.2%) with age estimation of 27.0, in Shungan of Tadjikistan, Mongolians, Koreans, Kalmyks, Bashkirs, Mansi, Nogays, Kurds, Druze, Adyge, Ukrainians, Georgians, Turks, Yakuts, Evens, Evenks
C4a1a is found in Siberia with the age estimation of 11.33
C4a1a is found in East-India in Lepcha, Lachungpa and Wanchoo
C4a1a is found in South-India
C4a1b'd is found in Mongolia
C4a1b is found in Inner Mongolia
C4a1b is found in India with the age estimation of 11.71
C4a1c is found in Tubalar and Altaian Kizhi, Dolgans
C4a1d is found in Buryat, Shor
C4a2'3'4 is found in ancient Dneper Donetz
C4a2a1 is found in Siberia in Shor and Evenks, Evens, Yakuts, Turkmen, Bashkirs, Nogays with the age estimation of 6.1
C4a2a2 is found in India and China (Tibet?) with age estimation of 12.85
C4a2a2a is found in India with the age estimation of 8.31
C4a2b is found in India with the age estimation of 14.49
C4a3 is found in ancient Dneper Donetz
C4a3 is found in TIbet
C4a3 is found in South India
C4a3a is found in South India
C4a4 is found in South India and in Siberia (Bargut), Shungan of Tadjikistan
C4a4a is found in South India
C4a6 is found in ancient Dneper Donetz
As age estimation for C4a1 is equal in India and Siberia, I would propose that the haplotype originated in Tibet before the Ice Age or during the Ice Age. Also the recent Qui et al. 2013 paper on Tibetan mtDNA and Ydna consider C4a one of the earliest haplogroups in Tibet (27 kya). According to that paper, only B4, D4, D5a2a, D6a, G2a and M62b have older age estimations in Tibetan groups. The oldest yDNA haplogroups are D1a-N1, NM231 (usually LLY22G), O3a-M324, O3a3c-M134). It is probable that C4a1 went south with yDNA O3 and D. In India, C4a1 is found in Sikkim (Lepcha, Lachungpa) near Nepal and Bhutan, which could be the place of entrance of C4a1 in India. In the North, C4a1 has been detected in Jinggouzi site (2500 ybp) in Inner Mongolia where all yDNA was C (xc3, C1).
In the Qui et al. paper, the age estimation for C4a2'3'4 in Tibet is 21 kya. It is also noticeable that the age of C4a2 is much older in India than in Siberia. I would say that C4a2 spread to India earlier than C4a1. After the Ice Age, C4a2 may have spread from India to Central Asia or from Tibet through Altai to Central Asia.
However, I would not connect C4a2 with yDNA N1c or N1b, because C4a2 is not typical of Uralic people. According to Bermisheva 2002, “only one variant, 16,223–16,298–16,327 mapping to the center of the phylogenetic tree, was found in Udmurts, Tatars, Bashkirs, Chuvash, and Komi-Permyaks.” The frequencies are: Chuvash 1.8%, Mordvinians 2.0%, Komi-Permyaks 8.1%, Mari 0.7%, Udmurts 3.0%. This ancestral haplotype was found in Baraba Steppe 4.0 kya and Bolshoy Karelia 3.5 kya pb. On the basis of this, should we suggest that ancestral C arose in Southern Siberia 30 kya bp?
If this is true, then C4a1 may have spread from Northern China to Tibet with yDNA N-LLY22G which is wide-spread and old in Tibet.
By the way, I also noticed that there is a very divergent C in Hunza and a haplotype with one identical mutation is found in Mongols. As for the frequency of C4 in Southeast Asia, in a recent study on Laos MtDNA, only C7 was found and no other C haplotypes.
A problem with Tibetan origins, especially in the Ice Age, is that most of that area was not habitable (extremely cold). Let's not forget that the Tibetan Plateau is the highest (and largest) such formation of Earth and that, together with the Taklamakan it constitutes the massively desertic buffer zone between East Asia and Centra-West Eurasia. There is no evidence of inhabitation in this gigantic arid zone before Neolithic times or later. The only corridor of direct West-East interaction was further North, in Mongolia-Siberia.
DeleteHowever the ancestors of modern Tibetans may have lived in the Eastern outskirts of the formation, in Yunnan and Sichuan.
Another problem with a Tibetan origin is the weakness of the main Tibetan patrilineages of the D1 haplogroup in the areas affected by the C4 expansion. Instead it would seem to be associated with patrilineages of the NO family (N1, O2a, O3), right? D3, more common among NE Asians has a "moderate distribution" in Central Asia and D* is somewhat important in Altai but these hardly show any connection to Tibet as such.
"It is also noticeable that the age of C4a2 is much older in India than in Siberia. I would say that C4a2 spread to India earlier than C4a1".
I can't say but it's also possible that the "age" only reflects two (or more) different expansions of C4a2 sublineages in more recent times, being then only an approximation to the "age" at the original C4a2 area, probably in SE Asia.
It'd be also convenient to discern NE India from the rest of the subcontinent, because this region (Assam, Meghalaya, etc.) is very much transitional between the two continental regions and largely aligns with SE Asia rather than with South Asia. It can be therefore a source of confusion, very especially if the lineage arose in the Brahmaputra-Irrawadi area.
Kristiina,
DeleteI don't know if it matters but in your list based on Ianlogan's data from GenBank,
Only C4a1 is so far found in South India (my sample, and I have NO tibetan heritage).
All the other samples listed from Rao are from East India where the population is Tibeto-burman.
Hence your list should read:
C4a1 is found in Tibet (3.2%) with age estimation of 27.0, in Shungan of Tadjikistan, Mongolians, Koreans, Kalmyks, Bashkirs, Mansi, Nogays, Kurds, Druze, Adyge, Ukrainians, Georgians, Turks, Yakuts, Evens, Evenks, South India
C4a1a is found in Siberia with the age estimation of 11.33
C4a1a is found in East-India in Lepcha, Lachungpa and Wanchoo
C4a1a is found in East-India
C4a1b'd is found in Mongolia
C4a1b is found in Inner Mongolia
C4a1b is found in India with the age estimation of 11.71
C4a1c is found in Tubalar and Altaian Kizhi, Dolgans
C4a1d is found in Buryat, Shor
C4a2'3'4 is found in ancient Dneper Donetz
C4a2a1 is found in Siberia in Shor and Evenks, Evens, Yakuts, Turkmen, Bashkirs, Nogays with the age estimation of 6.1
C4a2a2 is found in India and China (Tibet?) with age estimation of 12.85
C4a2a2a is found in India with the age estimation of 8.31
C4a2b is found in India with the age estimation of 14.49
C4a3 is found in ancient Dneper Donetz
C4a3 is found in TIbet
C4a3 is found in East-India
C4a3a is found in East-India
C4a4 is found in East-India and in Siberia (Bargut), Shungan of Tadjikistan
C4a4a is found in East-India
C4a6 is found in ancient Dneper Donetz
One of the problems is that "India" includes NE India, which is very much akin genetically to SE Asia, so it's hard to judge without knowing where exactly in India these lineages have been found.
DeleteThis applies especially to NE Asia but also to other regional distributions. India is a continent-sized state so it's not at all the same saying "in India" than saying "in Cambodia", for example. It's a bit like saying "in Russia", where it's not at all the same a sample from ethnic Russians from Moscow than ethnic Evens from the Sakha Republic (Yakutia).
SB Thankyou for the correction! I thought that the relict populations were for the most part from the south. It was not said in the article. The fact that it is found almost exclusively in Eastern India makes the Tibetan origin all the more probable.
ReplyDelete“Another problem with a Tibetan origin is the weakness of the main Tibetan patrilineages of the D1 haplogroup in the areas affected by the C4 expansion. Instead it would seem to be associated with patrilineages of the NO family (N1, O2a, O3), right?”
The paucity of D1 in the north is not a problem, in particular, if N LLY22G spread to Tibet from the north.
It is, however, remarkable that C4a1, C4a2 and C4a3 are all well anchored in Tibet and India and the northern clade C4a1 is today found also in West Asia but not in Western Uralics.
I cannot really see a clear connection between C4 and NO. N and most O clades have not much to do with India. The frequency of C4a is 0.47% in northern China and it is not found in southern areas or in Indo-China (Tanaka et al.).
According to my paper, the highest frequencies of C4a are in Tofalar and Evenks and the highest frequencies ofC4b are in Nganasan, Evenks and Yukaghirs.
As for the yDNA of the groups that are high in C4a, Western Evenks have 70% of C3 and 27.5% of N1b and 2.5% R1a, Okhotsk Evenks have 62.5% of C3 and 37.5% of N1c. Evens have 9.1% of N*, N1b and N1c. Yukaghirs have 23.1% of C3c, 7.7% of C3, 30.8% of Q-P36 and 30.8% of N1c. Dolgans have 12% of N1b and 22% of N1c. Tofalars have 70% of N1b and N1c. Yakuts have 80% of a special type of N1c, but they are a very small ethnic group with a narrow yDNA ancestry and a short history. In more western areas C4a is most frequent in Kets (13.2%) and probably spread from Kets to Mansi (6.1%). Nganasan are 91% of N1b and they have the following mtDNA haplogroups: U4a 12.8%, U4c 7.7% (Nganasan specific line), C4a 10.3%, C4b 20.5%, C5 20.5%, Z1 2.6%, D3 17.9% (Nganasan specific line), D4j 2.6%D4o 2.6%. Their original mtDNA’s seem to be U4, C4b, C5 and D3.
In my opinion, you could just as well say that C4a is linked with C3 and C3c or even Q, as Yukaghir, Koryak and Chukchi are all high in Q and C4a. It is not irrelevant either that C4a1 was detected in Jinggouzi site (2500 ybp) in Inner Mongolia where all yDNA was C (xc3, C1).
I cannot really see a clear connection between C4 and NO. N and most O clades have not much to do with India.
DeleteO2a and O3 are clearly the main patrilineal vectors of East Asian genetic influence in the subcontinent, being strongly associated with the Austroasiatic and Tibeto-Burman expansions respectively.
"In my opinion, you could just as well say that C4a is linked with C3 and C3c or even Q, as Yukaghir, Koryak and Chukchi are all high in Q and C4a".
I do not mean that mtDNA C4a has an overall and exclusive global relation with Y-DNA NO, first of all because it'd be like comparing a small apple to a whole orange grove; NO is a macro-lineage of widespread distribution and dominant in many populations (mostly East Asians but also Finnic, Indian Austroasiatic, etc.), instead C4a is a relatively small sublineage in spite of its large geographical scatter.
Also it seems only logical that in each different geographical and ethno-historical context there should be particularities and exceptions to the broad relation outlined with my "claim". You should not take what as a rule that must be fulfilled in every case but rather as a tendency that fits reasonably well in most cases.
Finally I was talking specifically in the context of your proposed association to Tibetans, which I reject. If C4a is associated in some cases with Y-DNA Q or C3, that does not still make it any connection to Tibet.
I do not understand how you can reject the connection between C4a and Tibet, if C4a, C4a2a2 and C4a3 (C4a3b1 and A4e3a in Sherpa) are all found in Tibet and the frequency of C4a1 is according to Qui et al. 2013 4.65% in Tibet and according to Gayden 2012 3.2% (C4) and 3.2% (C4a1).
ReplyDeleteHere you have some haplotypes from Tibet:
Monba Himalayish Nyingchi, Tibet 223 298 311 327 C*
Tibetan Himalayish Nyingchi, Tibet 93 129 223 297 298 327 C4*
Tibetan Himalayish Chamdo, Tibet 223 298 327 354 357 C4
Tibetan Himalayish Shigatse, Tibet 93 129 223 224 298 327 C4a1
Tibetan Himalayish Yushu,Qinghai 167 171 223 298 327 344 357 C4a2
Tibetan Himalayish Ngari, Tibet 93 298 327 C4d
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&frm=1&source=web&cd=1&ved=0CCgQFjAA&url=http%3A%2F%2Fcomonca.org.cn%2Flh%2Fdoc%2FA58S3.xls&ei=zNpGUpLiAYaK4ASO3YGQBw&usg=AFQjCNF9kaEArNFyERENUvv89nDr2RnYzw&sig2=FmP1S5MRgS2mUSD7oP678g&bvm=bv.53217764,d.bGE
Tibetans also have their fair share of Y-DNA O3. What I reject anyhow is any obvious origin in Tibet. That kind of "connection" you claimed, not any sort of generic one.
DeleteI agree on that! The place of origin is not easy to pinpoint on a map. The Ice Age surely caused a lot of movement and I would guess that people pressed south ward. However, it seems that people were amazingly well equipped for cold climates and were willing to go north even during the Ice Age.
ReplyDeleteThe intriguing question to me is still how C4a2 ended up in Central Asia. I cannot see a northern route through the areas of Fenno-Ugric speakers. I would be very interested to hear suggestions on the route, time frame and YDNA that acompanied C4a2.
lthough I rejected the connection between haplogroup NO and C4a, I agree that, in general, yDNA N and mtDNA C are linked, and what they have in common is the northern distribution pattern, but things are always complex, for example C7 is found in Indo-China.
ReplyDeleteIn Bodner et al paper they say that haplogroup C, in contrast, showed an unexpectedly high frequency of 6%, but very limited diversity: all haplotypes belonged to haplogroup C7. Interestingly, 12 of the 13 samples derive from Northern provinces. Then I noticed that paragroup N-LLY22g* has been found in 1 of 28 Bit, and according to Wikipedia, Bit is a language spoken by around 1,500 people in northern Laos. There are thought to be about another 500 speakers over the border in Yunnan Province, China. N-LLY22g* has also been found in Hani in Northern Vietnam with a frequency of 4/28 and in Tujia with a frequency of 2/49. Tujia live in Wuling Mountains, straddling the common borders of Hunan, Hubei and Guizhou Provinces.
"[A]lthough I rejected the connection between haplogroup NO and C4a, I agree that, in general, yDNA N and mtDNA C are linked, and what they have in common is the northern distribution pattern"...
DeleteNice that we are reaching an agreement.
... "but things are always complex, for example C7 is found in Indo-China".
And what is odd about that? Isn't the Indochina Peninsula part of East Asia? Isn't it dominated by Y-DNA O? I really do not understand why this may be a "but", sorry.
I think that the frequency of C7 is too small and its area too restricted to match with O in general. According to Ian Logan site, C7 is found in Siberia, China and India. Unfortunately, I do not know how much C7 (or C4a) there is in Indo-China. The frequency of C7 (on the basis of HVR1 mutations, it is C5 in Hmong Mien paper Wen et al 2004) seems to be higher in Hmong-Mien populations than in average Han. According to Wen paper, it is almost completely absent in the northern East Asian populations and seems to be the major branch of haplogroup C in the southern East Asians such as Dai, Zhuang, and Lahu.
ReplyDeleteBy the way, I also noticed that, according to Wikipedia, haplogroup N detected in Hmong Daw in Laos is N1c and similar to Yakut branch. In that Hmong paper (Cai 2011) they also detected Q in Hmong Daw.
I am NOT saying that (mt) C has any particular link to (y) O or NO, just that it does not seem to have any correlation with D1 and therefore with Tibet. Don't go crazy about that remark please. (mt) C is probably old enough to have experienced many different expansions, however the most remarkable one is the one of Northern Eurasia, which, at least towards the West, I'd associate with (y) N1 (but maybe I'm wrong, it's just indirect inference).
ReplyDeleteWe can't aim to establish a simple one-on-one patrilineage-matrilineage relation: because of the nature of sexual reproduction and interactive nature of humans, the correlations will necessarily change through time and space. We can only infer, in the best case, some correlations for certain specific processes and not for all Eurasia and for the 100,000 years of H. sapiens (pre-)history in this huge continent. You must make room for complexity and multiplicity and, as I said before neither (mt) C4, nor C nor even all M8 are comparable in numbers to (y) NO, what means that other (mt) lineages were involved in the NO expansion processes (not one but several) and even in some cases also other (y) ones. We are talking after all of many dozens of millennia here in the largest continental landmass of Earth.
In fact, I suggested that C4a was brought to Tibet by N-LLY22G, and from Tibet to India mainly by D and O3e.
ReplyDeleteAs you may know, at the moment I think that North Eurasian mtDNA C1 was connected to yDNA Q, but, of course, there were probably also N lines in these northern areas.
Kristiina wrote,
ReplyDelete"I think that the frequency of C7 is too small and its area too restricted to match with O in general. According to Ian Logan site, C7 is found in Siberia, China and India. Unfortunately, I do not know how much C7 (or C4a) there is in Indo-China. The frequency of C7 (on the basis of HVR1 mutations, it is C5 in Hmong Mien paper Wen et al 2004) seems to be higher in Hmong-Mien populations than in average Han. According to Wen paper, it is almost completely absent in the northern East Asian populations and seems to be the major branch of haplogroup C in the southern East Asians such as Dai, Zhuang, and Lahu."
I think this sort of attempt to correlate the Y-DNA and mtDNA phylogenies is futile in most cases. For example, a search for an mtDNA haplogroup that is frequent in all populations in which a derivative of Y-DNA haplogroup NO-M214 predominates is bound to end in failure; the mtDNA haplogroups that are frequent in populations of North Asia in which Y-DNA haplogroups N1c1-M46 and N1c2b-P43 are common are derivatives of C and D, neither of which is found throughout Southeast Asia, where derivatives of Y-DNA haplogroup O-M175 predominate. Certain subclades of mtDNA haplogroup C or D have been detected in certain populations of Southeast Asia, but none exhibits the universal, high-frequency distribution of O-M175 in that region. The reverse is true, too; none of the subclades of mtDNA haplogroups B, F, M7, or E that predominate in mainstream populations of modern Southeast Asia is found with the broad geographical spread and high frequency of subclades of N-M231 in North Asia. The mtDNA of populations in Europe that exhibit N1c1-M46 Y-DNA with high frequency is again very different overall, though some of these populations do exhibit mtDNA belonging to haplogroup C, Z, or D with low frequency. (Some other Eastern Eurasian haplogroups, including A, B, G, F, N9(xY), Y, M7b, and M10, also have been found in Europe, but mainly in Turkic, Caucasian, or Slavic populations that are not particularly associated with any subclade of Y-DNA haplogroup N-M231 or O-M175.)
Consider the indigenous Americans for instance. All their (original) Y-DNA appears to belong to one of two phylogenetically widely separated clades, Q1a2-M346 and C3-M217, but their mtDNA belongs to one of five distinct clades: A2, B2, C1, D1, or X2a. One would have to go all the way back to the level of N vs. M in order to find a bifurcation that might allow one to neatly reconcile the Y-DNA heritage and the mtDNA heritage of these populations, but that would clearly be mere imagination.