May 19, 2014

The history of mtDNA haplogroup U6

BMC Evolutionary Biology 2014, 14:109 doi:10.1186/1471-2148-14-109

The history of the North African mitochondrial DNA haplogroup U6 gene flow into the African, Eurasian and American continents

Bernard Secher et al.

Abstract (provisional)

Background

Complete mitochondrial DNA (mtDNA) genome analyses have greatly improved the phylogeny and phylogeography of human mtDNA. Human mitochondrial DNA haplogroup U6 has been considered as a molecular signal of a Paleolithic return to North Africa of modern humans from southwestern Asia.

Results

Using 230 complete sequences we have refined the U6 phylogeny, and improved the phylogeographic information by the analysis of 761 partial sequences. This approach provides chronological limits for its arrival to Africa, followed by its spreads there according to climatic fluctuations, and its secondary prehistoric and historic migrations out of Africa colonizing Europe, the Canary Islands and the American Continent.

Conclusions

The U6 expansions and contractions inside Africa faithfully reflect the climatic fluctuations that occurred in this Continent affecting also the Canary Islands. Mediterranean contacts drove these lineages to Europe, at least since the Neolithic. In turn, the European colonization brought different U6 lineages throughout the American Continent leaving the specific sign of the colonizers origin.

Link

55 comments:

Dr. Clyde Winters said...

There is no archaeological support for the hypothesis that U6 in Africa is the result of a back migration. This hapologroup probably spread to Europe from Africa.

Secher et al, make a lot of unfounded speculations. First they talk about the possibility of carriers of L3(M,N) migrating back into Africa around 40kya via the Levant. This dating has to be wrong because 40kya the Levant was still occupied by Neanderthal.

The next great speculation is that the Aurignacian culture originated in Central Asia and expanded from their into the Levant. The present archaeological dates for the Aurignacian culture find the earliest dates for this culture in Iberia, and the latest dates in the Levant and Eastern Europe.

Vincent said...

The authors suggest that the paternal counterparts of the Eurasian back-migration which brought U6 and M1 in Africa might be R-V88 and T-M70, but they are obviously wrong. R-V88 is known to have arrived where it is most common not earlier than the Neolithic, and T-M70 is also probably Neolithic in Africa. When will the academic world wake up and consider the idea of E-M96 back-migrating from its Eurasian birthplace to Africa? The authors of the study also make a reference to the Hofmeyr skull, which I think was a carrier of Y-DNA E-M96* or E-M75* and mtDNA M1.

There is no better candidate than E-M96 to represent the Eurasian back-migration to Africa in UP times.

andrew said...

The paper is one of the most thoughtful in terms of integrating archaeological evidence with genetic evidence that I have seen.

@Vincent. They consider the possible fellow traveler connection with Y-DNA hgs, but they ultimately conclude as you do that the age differences between the Y-DNA clades and mtDNA clades disfavor that proposal. The authors of the study were not aware of the ancient DNA of the Hofmeyr skull but the mtDNA assignment (presumably because it was written earlier that any announcement of these result) that you cite confirms their conclusion.

@Clyde Winters

The case that U6 and M1 are back migrations to Africa are consensus views based upon the genetic evidence and have been for probably a decade at least for reasons summarized in the paper.

The Aurignacian culture was the culture of the first modern human Cro-Magnons in Europe and is not preceded by any other modern human cultures in Europe. An expansion into the Levant a few thousand years later than into Europe is consistent with the archaeological record and the authors position. And, a Central Asian origin explains parallel expansions of the Aurignacian culture into India. One can shift the boundaries a bit (e.g. to Iran in West Asia), but there is no other sensible central point for mtDNA haplogroup U that has a clade basal to U2 (India), U5 (Europe) and U6 (North Africa) in a common location consistent with the fact that Levantine sites are later than European ones.

You apparently seem to argue for a Aurignacian expansion from NW Africa to Spain then East across Europe and from there to India and the Levant. Articles such as the linked one published in Nature in 2004, look at the relevant date and suggest an East to West order of expansion. A good overview from 2011 makes clear that regional differences within Europe of the onset of the Aurignacian are smaller than the resolution of radiocarbon dating techniques and places the oldest such site in Bulgaria, several thousand years before the oldest Iberian site.

Genetic data tend to favor an East to West route in Europe since the genetics of pre-LGM Europe are not derived from haplogroups found in NW Africa from which they would have to be with an Iberian arrival of modern humans in Europe.

Slumbery said...

Dr. Clyde Winters

"This dating has to be wrong because 40kya the Levant was still occupied by Neanderthal."

Some Neanderthals still here does not mean that there were not modern human population as well. It is not like Neanderthals had an organised empire with border patrols.
Actually it is quite sure that modern humans were all around the Levant by 40k BP.

eurologist said...

It makes sense that U6 is part of the Levantine Aurignacian, and not the Ahmarian, and thus of more northern West Eurasian (if not E/C European) origins.

Obviously, the male counterpart may be of completely different, perhaps local origin. I find the timing (~ 45,000 - 35,000 ya) too late for a y-DNA haplogroup E back-migration. T-M70 would fit the more southern origin of the Ahmarian, while R-V88 appears to be not old enough.

Vincent said...

Eurologist,

I don't think that a range of 45-35 ky is too late for E-
M96's back-migration.
T-M70 is 26.5 ky old (range 19-34 ky), which is too recent, and certainly its clades into Africa are downstream and not ancestral. Plus, there are not other haplogroups compatible with an entrance 35-45 kya other than E-M96* (which is found at low frequency in the Maghreb). E-M33*, which also originated in the Levant I think, is a good candidate

astenb said...

@ Vincent reg: " The authors of the study also make a reference to the Hofmeyr skull, which I think was a carrier of Y-DNA E-M96* or E-M75* and mtDNA M1."

Can you supply a source for this information. I was not aware.
Thank you.

Vincent said...

astenb,

Read more carefully my statement. I wrote "I think", not "geneticists found out that"

Dr. Clyde Winters said...

We can only verify the origin dates for specific haplogroup based on archaeology and amtDNA (ancient DNA). The authors of this article make assumptions about historical events they can not support with the archaeology of western Eurasia, and the Levant in relation to amh belonging to the Aurignacian culture.

Slumbery said...


“Some Neanderthals still here does not mean that there were not modern human population as well. It is not like Neanderthals had an organised empire with border patrols.
Actually it is quite sure that modern humans were all around the Levant by 40k BP.”

This is your opinion. There is no archaeological evidence to support the presence of amh in Levant at this time. The best evidence for amh in the Levant comes from Ksar Akil, and later Natufian sites.

andrew said...

“ You apparently seem to argue for a Aurignacian expansion from NW Africa to Spain then East across Europe and from there to India and the Levant. Articles such as the linked one published in Nature in 2004, look at the relevant date and suggest an East to West order of expansion. A good overview from 2011 makes clear that regional differences within Europe of the onset of the Aurignacian are smaller than the resolution of radiocarbon dating techniques and places the oldest such site in Bulgaria, several thousand years before the oldest Iberian site. “
This is false the earliest sites for Aurignacian are found in Spain. The radio carbon dates for Bugalria, i.e., the Kozarnika Cave, date back to 39-36kya. Earlier researchers claimed that the artifactual material found at the Bacho Kiro cave, dating to 46kya was Aurignacian, but the remains that consist of a pair of fragmented human jaws, is disputed and researchers don’t know whether these were early humans Homo sapiens or Neanderthals.
The first Aurignacians in the Levant date back to 36-34kya from Ksar Akil
The oldest Aurignacian remains come from Iberia/Spain. These sites vary in age from 41kya for the l'Arbreda Cave, and 43kya for Abric Romani, located in Catalonia, Spain.
The dates for the Aurignacian in Europe make it clear this culture spread from west to east. You can also recognize that Aurignacian appears not to have reached the Levant, until 11ky after it was established in Spain.
These dates for sites where amh were found in Western Europe make it impossible for claims of U6, M1 and etc., originating prior to 32kya in the Levant and entering Africa via a back migration 40kya.

For more information on the Aurignacian culture see:

Demidenko Y.E., Otte M. & Noiret P. (dir.) - Siuren i rock-shelter. From Late Middle Paleolithic and Early Upper Paleolithic to Epi-Paleolithic in Crimea. Liège, ERAUL 129, 2012, p. 343-357.

http://orbi.ulg.ac.be/bitstream/2268/135222/1/Chapter%2018%20Europe%20Aurignacian.pdf




astenb said...

@ Vincent. I saw that, but the way your sentence was written I didnt come to that conclusion. Also see the comment by Andrew who made a similar mistake as I. Also E-M33 is not Levantine. It is West African with diversity/concentration centers in the West African Sahel.

In any case. Looking at the distribution of E* lineages in Africa, DE* lineages in Africa and the diversity of and spread of E1a, E2a and E2b...It makes no sense to think of E* as the back-migrant to the region.

If anything the back-migrant theorist should be looking to E-M35. Most Africans that have E-M35 lineages also carry M1 and or U6 in tandem. It also is not too old (DE* and E* at nearly 65kya) nor to young (J1/V-88/T2a) to be a matching migrant of an incoming U6.

Vincent said...

astenb,

DE* almost certainly originated in Asia, and there is a high probability that this was also the case for E-M96* and several of its sub-clades. E-M96* is not an exclusive of Africa, it has been found in some parts of Europe (notably in the Balkans), Anatolia, the Levant, Saudi Arabia. Plus, the only real case of underived E-M215* was found in the region of Khorasan, North-East Iran. The previously identified cases in Ethiopia and Yemen were found to be in the E-M35 sibling clade, E-M281. Moreover, E-M2, the dominant sub-clade of E in sub-Saharan Africa is said by geneticists to have originated in North Africa, therefore in proximity of Eurasia. Geneticists also suggest that E-M75, aka E2 probably originated in East Africa, which, together with North Africa is in proximity of Eurasia. All of this suggests that most of the ancestral clades of E-M96, and E-M96 itself, probably originated in Asia, specifically in the Levant.

Regarding E-M33*, there are cases in Lebanon and Syria. I have analysed E-M33 haplotypes in Africa, Asia and Europe, and I found that diversity is higher in Asia than in Africa. For Asia I obtained a TMRCA of 44.6 ky with maximum diversity in Syria. For Africa I obtained 25.1 ky, and for Europe (represented uniquely by Portuguese haplotypes) I obtained 24.7 ky. To test the reliability of my method of analysis, I checked Gonçalves et al. (2005) to compare their TMRCA estimate for the same Portuguese haplotypes I used, and they obtained 22.9 +/- 7.2 ky. Here the quote from the authors: "The presence in Portugal of both the A and E1 haplogroups may be independent from the slave trade (otherwise E3a would be well represented since it comprises the majority of West African lineages). 

These findings either suggest a pre-neolithic migration from North Africa or a more recent origin from a founder population of small size that did not carry haplogroup E3a, which is a major component in North African populations today. 

TMRCA for Portuguese E1 lineages estimated as 22.9 +/- 7.2 ky favors the first scenario".

German Dziebel said...

@Vincent

"There is no better candidate than E-M96 to represent the Eurasian back-migration to Africa in UP times."

Another sane person around.

terryt said...

"DE* almost certainly originated in Asia, and there is a high probability that this was also the case for E-M96* and several of its sub-clades".

I'm not convinced one way or the other yet. However I'm inclined at this stage to agree with astenb,

"Looking at the distribution of E* lineages in Africa, DE* lineages in Africa and the diversity of and spread of E1a, E2a and E2b...It makes no sense to think of E* as the back-migrant to the region".

So far we have DE* in Nigeria (I think), E* in a single Bantu speaker from South Africa. Then we have two E haplogroups: E1-P147 and E2-M75. E2 is widespread in Sub-Saharan Africa but not particularly present in East Africa. Two branches of E1: E1a-M33 (basically a West African haplogroup) and E1b-P177 (widely distributed through Sub-Saharan Africa). We don't find specifically East African branches until we get to E1b1b-M215.

Vincent said...

@German

Thank you. I hope that other people will understand the obviousness of the things I exposed above.

eurologist said...

I agree that if there was an early back-migration, it would have been very high up the DE or E* tree. Conversely, most of NE Africa and SW Asia was basically uninhabitable during much of ~105,000 - ~50,000 ya. You have to go to the NE Mediterranean or the Southern costs of the Pontic or Caspian - or all the way to Pakistan to find inhabitable areas. Well, except for Pakistan it is pretty clear that those regions were then inhabited by Neanderthals and not by AMHs.

More importantly, it would be an incredible leap of faith to think that the UP developed for a second time and independently in SW Asia.

IMO all this means that 35,000 - 25,000 ya, or even 50,000 ya, there was no E left in SW Asia. Conversely, the current age estimates for T-M70 agree well with the time frame considered in the present study.

Rob said...

"Looking at the distribution of E* lineages in Africa, DE* lineages in Africa and the diversity of and spread of E1a, E2a and E2b...It makes no sense to think of E* as the back-migrant to the region"

But why do we automatically assume that maximal haplogroup diversity represents region of origin ?
Rather, it might only represent more recent founder effects and range expansions from an different source location

terryt said...

"But why do we automatically assume that maximal haplogroup diversity represents region of origin ?"

Yes. But in this case I was dealing with 'basal diversity'. There are several explanations for 'diversity. Such as:

"Rather, it might only represent more recent founder effects and range expansions from an different source location"

In most cases I have looked at closely the latter explanation is usually sufficient. Such a solution seems to be the explanation for the diversity Vincent refers to above.

"I agree that if there was an early back-migration, it would have been very high up the DE or E* tree".

The back-migration Dienekes blogged about some time ago may pre-date the development of any modern haplogroups though.

" Conversely, most of NE Africa and SW Asia was basically uninhabitable during much of ~105,000 - ~50,000 ya. You have to go to the NE Mediterranean or the Southern costs of the Pontic or Caspian - or all the way to Pakistan to find inhabitable areas. Well, except for Pakistan it is pretty clear that those regions were then inhabited by Neanderthals and not by AMHs".

It depends to a considerable extent when the actual OoA occurred. Around 120,000 years ago the earth was warmer than at present, and presumably wetter. As far as I'm aware Neanderthals didn't move south into the Middle East until the general climate cooling and increased aridity that began developing around 70,000 years ago. That sequence would have allowed humans from Africa to expand considerably before being pushed southward.

"IMO all this means that 35,000 - 25,000 ya, or even 50,000 ya, there was no E left in SW Asia".

Yes, but obviously D and E (at least in the form DE) shared a considerable geographic range before the two became isolated in their separate regions. To me that separation must be associated with climate cooling and increased aridity. Their original geographic range must have stretched from North Africa to Central/East Asia.

"the current age estimates for T-M70 agree well with the time frame considered in the present study".

Yes.

Vincent said...

"...the current age estimates for T-M70 agree well with the time frame considered in the present study"

T-M70 originated 26.5 +/- 7.5 kya, and this is just out of the 45-35 ky range. Plus, I don't think that T-M70 in Africa is underived T-M70*.

It does not make sense that T-M70 or J or R-V88 back-migrated in Africa in UP times. Instead, it makes the most sense to consider E-M96 as the back-migrating Y-DNA haplogroup

terryt said...

"Instead, it makes the most sense to consider E-M96 as the back-migrating Y-DNA haplogroup"

Where are you proposing it originated then? I can easily accept that DE moved into Africa from somewhere outside the continent but to me E itself can only have evolved in Africa. You will find these maps relevant:

https://www.google.co.nz/search?q=y-chromosome+haplogroup+e&source=lnms&tbm=isch&sa=X&ei=DyKAU5HhL4SGkAX9mYHoBw&ved=0CAYQ_AUoAQ&biw=1582&bih=767#q=y-chromosome+haplogroup+e+africa&tbm=isch&facrc=_&imgdii=_&imgrc=a17fEAAMHQ-oTM%253A%3B2dbJFUjGfNQl0M%3Bhttp%253A%252F%252F3.bp.blogspot.com%252F_x6Y4ZgFsZdY%252FTSeTBdooZFI%252FAAAAAAAAAe4%252Fmy4e8LMw1vs%252Fs1600%252FY-DNA%252BE%252Bclades.png%3Bhttp%253A%252F%252Fwww.forumbiodiversity.com%252Fshowthread.php%252F29422-Language-of-the-Y-chromosome-hg-E-Eurafricans%3B641%3B795

The nomenclature hasn't changed much and so you should be readily able to see distributions.

eurologist said...

"It does not make sense that T-M70 or J or R-V88 back-migrated in Africa in UP times. Instead, it makes the most sense to consider E-M96 as the back-migrating Y-DNA haplogroup."

Vincent,

the main, oldest African U6a and U6a subgroups in this study date from ~18,000 to 26,000 ya and even 10,000 - 13,000 ya for U6[b,c,d]. The downstream U6a7 is an outlier at 29,000 ya. While U6 itself dates to ~35,000 ya, that does not mean that's the date it entered NE Africa. Instead, it appears that's the date it originated before it entered the Levant from the north. At any rate, as I said before, the timing matches T-M70 very well.

Vincent said...

@terryt,

E-M96 originated either in the Southern Anatolian/Northern Levantine region or further North.

@eurologist,

First, T-M70 originated in West Asia 26.5 kya, and this is not compatible with mtDNA U6's TMRCA. Second, African T-M70 is not underived/ancestral, it is T1a1a-L208, which is surely several millennia younger than T-M70*. In addiction, mtDNA U6 is not very common where T-L208 is found in great amount.

The authors of the study say that "the upper limit for the first U6 radiation within Africa, represented by the time to the MRCA of U6a is 26.2 (20.3-32.2) kya". This is in perfect agreement with the TMRCA I obtained for E-M33* in Africa, that is 25.1 ky. In Table 3 of the study, the age for the spread of the clade U6a1 in Europe (in this case Iberia, as explained in the Discussion section) is given as 18.6 kya, which is compatible both with the TMRCA of E-M33* in Portugal according to Gonçalves et al. 2005 (22.9 +/- 7.2kya) and to my results (24.7 kya)

Rob said...

It is too early to arrive at definite conclusions. However, it is possible that severe climactic fluctuations in Arabia and the Levant, might have erased past genetic patterns, with the rise of new dominant haplogroups (ie the J family). This is clearly the case, as there is a large geographic 'gap' between the Nth African dominant E on the one hand, and that of C and D in the far East.

This is further evidenced by E-V13 in Europe; with no evidence for a direct link between Nth Africa and SEE (unless one clutches at straws and draws simplistic conclusions on alleged 'upper palaelithic ' assemblage similarities). This suggests that E derived groups certainly existed, not necessarily implying they *originated*, in western Asia / Near East, but have since become extinct.

terryt said...

"E-M96 originated either in the Southern Anatolian/Northern Levantine region or further North".

On what evidence do you base that claim?

"This is in perfect agreement with the TMRCA I obtained for E-M33* in Africa, that is 25.1 ky."

But it conflicts considerably with earlier estimates of 57,703 years for the E1-P47 vs. E2-M75 split Dienekes linked to here:

http://dienekes.blogspot.co.nz/2012/07/estimating-age-of-y-chromosome-adam_30.html

"it is possible that severe climactic fluctuations in Arabia and the Levant, might have erased past genetic patterns"

I'd say it was extremely likely to have been the case. As you say, that explains the 'large geographic gap between the Nth African dominant E on the one hand, and that of C and D in the far East'.

"This is further evidenced by E-V13 in Europe".

The E haplogroups in Europe are, without exception, downstream clades. They are therefore far from being 'evidence' for Paleolithic presence.

"with no evidence for a direct link between Nth Africa and SEE (unless one clutches at straws and draws simplistic conclusions on alleged 'upper palaelithic ' assemblage similarities)".

The connection need not be 'Paleolithic' at all. The Egyptians make great play of the connections between the 'Sea People' and the 'Libyans'. And boating had become widespread through the Mediterranean in the very early Neolithic.

eurologist said...

"First, T-M70 originated in West Asia 26.5 kya, and this is not compatible with mtDNA U6's TMRCA."

Vincent,

The point is, it does not have to be. It, or one of its subgroups, has to match the emergence of particular U6 subroups in Africa, most of which are dated much younger.

Secondly, I have no problem with E-subgroups matching a similar time line and perhaps even clustering with U6 subgroups. However, I would be more careful about the question whether they were already present in NE Africa at that time, with a back-migration of E farther upstream and much earlier.

Vincent said...

"On what evidence do you base that claim?"

I base that claim on the STR diversity I found for E-M96* haplotypes in that region and comparing them to the African E-M96* haplotypes.

"But it conflicts considerably with earlier estimates of 57,703 years for the E1-P47 vs. E2-M75 split Dienekes linked to here: 

http://dienekes.blogspot.co.nz/2012/07/estimating-age-of-y-chromosome-adam_30.html"

I don't see how this damages what I exposed. Look carefully and you will read this:

"UPDATE (August 1): The dates in this post have been superseded by the ones in Dates of major clades of the Y-chromosome phylogeny
."

In this updated article the P147-M75 divergence is 47 +/- 8 ky old, which is just compatible with the figure of E-M33 I obtained for the Levant, that is 44.6 ky.

Vincent said...

"It, or one of its subgroups, has to match the emergence of particular U6 subroups in Africa"

Can you explain why does it "have to"? Thanks.

Rob said...

Terry T- perhaps, however , firstly , the whole story about "sea peoples" is overplayed , and was largely a literary topos , not is there any reason to presume that they came from NE Africa , either wholly or predominantly

Secondly , you should not overstate the role of seafaring in the Neolithic . At best this was coastal hugging rather than transmieditteranean, and consisted of single merchants and their retinue. Not entire tribes , or even colonists as occurred much later (Eg Phoenecians, Greeks). What I was saying wrt V13 in Europe is that , yes whilst certainly downstream, it has no nearby 'sister Claude's' ; suggsesting they became extinct in the near east . The nearest subclades to v13 are in NE Africa. As I said above , no archaeological or historical event links nth Africa with Europe , in any time period, unless one makes broad sweeping speculations like "sea peoples" - an otherwise meaningless exonymic taxon whose role in history has been overplayed in outdated historical books about the "Bronze Age collapse"

A final point- in sure we remember the recent paper on North Africans which also used autosomal evidence whoosh suggests their recent origins - or significant westbasian input , incl Hg E
In fact, it's very similar to the case of R1b in Western Europe

terryt said...

"Secondly , you should not overstate the role of seafaring in the Neolithic . At best this was coastal hugging rather than transmieditteranean, and consisted of single merchants and their retinue. Not entire tribes , or even colonists"

That's not entirely true. Haplogroup E-V68 must have crossed open water to reach Sardinia. Any people capable of reaching that island would obviously be capable of reaching other regions.

"What I was saying wrt V13 in Europe is that , yes whilst certainly downstream, it has no nearby 'sister Claude's' ; suggsesting they became extinct in the near east".

Again that is not entirely true. Tha fact that 'The nearest subclades to v13 are in NE Africa' suggests V13 probably originated there. No need to postulate extinctions anywhere. After all NE Africa does include the River Nile.

"I base that claim on the STR diversity I found for E-M96* haplotypes in that region and comparing them to the African E-M96* haplotypes".

Have you a reference for that diversity?

"In this updated article the P147-M75 divergence is 47 +/- 8 ky old, which is just compatible with the figure of E-M33 I obtained for the Levant, that is 44.6 ky."

Perhaps so. But neither branch has a particularly noticeable representation in the Levant.

eurologist said...

Can you explain why does it "have to"? Thanks.

Vincent, what I meant was that within the context of T-M70 and its timing, it is not important how old U6 is, since it is the subgroups - mostly U6{b,c,d} and numerous subgroups of U6a that occur in Africa, "for which the timing has to match" if the connection to T-M70 is to hold up. Evidently, these subgroups appeared when U6 suddenly multiplied and occupied multiple different regions in North Africa and the Canary Islands.

And that apparently happened in the ~10,000 to ~26,000 ya time frame.

terryt said...

"Secondly , you should not overstate the role of seafaring in the Neolithic"

I believe that most haplogroups, in fact most genetic mutations, become concentrated in a particular region before expanding from that region, like ripples in a pond. The ripples are interrupted by various factors but all things being equal they expand in all directions.

You might care to look at the maps I linked to on May 24. Maps D and E show E1b1b-M35 and E1b1b1a1-M78 respectively. Notice the two centres for each haplogroup: one at either side of the Red Sea's mouth, the other at either side of the Mediterranean Sea's mouth. Members of both haplogroups have reached Mediterranean Europe, including most of the islands. You are prepared to tell us that this distribution has nothing to do with seafaring?

Vincent said...

"Perhaps so. But neither branch has a particularly noticeable representation in the Levant."

I don't understand what this means. Frequency is irrelevant, because it undergoes numberless variations in time. This means that some time in the past these lineages may have been at least statistically relevant. The time I am referring to is the Early Upper Paleolithic, when most of the dominant haplogroups of today's Levant (for example, J1 or J2 etc.) were either confined to relatively distant areas or non-existent in their derived form (for example I and J were still IJ, and IJ was in Central Asia when it originated). If this was the case, haplogroup E and its oldest downstream clades may have been the only candidates as markers of the Levantine (and probably African and European) UP populations. Think about this, it is not impossible that this may have actually taken place: haplogroup C was in Southeast Asia 50 kya, haplogroup F was in India together with its clades. IJK is not compatible as European or Levantine UP marker because it is too young and because it was confined to South Asia initially. The only serious candidates left are either DE or E. Besides STR diversity, the presence itself of DE* in the Levant and ancestral clades of E in Europe and the Levant (E-M96*, E-M33*) points at this conclusion, because in Europe one can hardly find ancestral C* or F* or IJ* haplotypes (these haplogroups are usually assumed to be the most likely candidates for the European UP). How comes then that we can find DE*, E-M96* and E-M33* but not C*, F*, G*, IJ*, K* or any other CF-derived ancestral haplogroup?

Rob said...

Terry - the 'nice distributions' are palimpsests and not necessarily the result of an initial sea-borne 'sweep'. I am not an expert in marine palaeo-archaeology, however, even as late as the Bronze Age, there appears to have been no trans-mediterranean voyages. Eg objects of Epyptian provenance appearing in Crete did so via down-the-line trade , going from port to port, rather than directly. Its all about prevailing currents and technological capabilities of the period in question. On the other hand, direct seafaring from Crete to Egypt *could theoretically* have been possible, but not vice-versa ! So we need to be careful with what we speculate,...

astenb said...

@ Vincent, did CT* DE* and E* back migrate from Eurasia before of after Eurasians interbred with Neanderthal, Denisovan and other yet to be named archaic humanoids who have left genetic signatures in Eurasians and not Africans?

Also what about E1b1* This ancestral clade is found in East Africa, West Africa as well as the Sahel. Did this backmigrate too?

Vincent said...

Point one: haplogroup E back-migrated to Africa. West Eurasian admixture is present virtually everywhere and in every ethnic group (farmers, pastoralists, hunter-gatherers). The only thing in common between all these ethnic groups is haplogroup E. Read this article and its author's opinion http://dienekes.blogspot.it/2012/03/effects-of-ascertainment-on-admixture.html?m=1

Point two: so far no E1b1-P2* has been found, as shown by Trombetta et al. (2011).

terryt said...

"the 'nice distributions' are palimpsests and not necessarily the result of an initial sea-borne 'sweep'".

I doubt they are 'palimpsests'. The fact that two haplogroups, E1b1b1-M35 and E1b1b1a1-M78, share the distribution makes your interpretation unlikely. The distribution could hardly be considered random, which is what we would expect if intermediate regions had been cleared of the haplogroups. In fact Spain, Italy, the Balkans and the Levant look to be intermediate regions of an original expansion, which includes the Mediterranean.

"even as late as the Bronze Age, there appears to have been no trans-mediterranean voyages. Eg objects of Epyptian provenance appearing in Crete did so via down-the-line trade"

Your idea fails to take the Sardinian E haplogroups into account. I agree that voyaging was probably coastal rather than direct but the haplogroup distributions show definitely a movement of populations around the Mediterranean at some time. It is unlikely that Sardinian E is anything other than a very early arrival there.

"The time I am referring to is the Early Upper Paleolithic, when most of the dominant haplogroups of today's Levant (for example, J1 or J2 etc.) were either confined to relatively distant areas or non-existent in their derived form (for example I and J were still IJ, and IJ was in Central Asia when it originated)".

Agreed. But:

"If this was the case, haplogroup E and its oldest downstream clades may have been the only candidates as markers of the Levantine (and probably African and European) UP populations".

I note you said, 'may have been'. Because:

"haplogroup F was in India together with its clades".

Not all 'its clades'. It is unlikely G or its ancestors were ever in India and the same may be true for H2-P96, a 'brother' to H1-M69. Both these haplogroups look native to somewhere in Southwest Asia.

"the presence itself of DE* in the Levant"

Has it been found there? What study discovered it there?

"ancestral clades of E in Europe and the Levant (E-M96*, E-M33*)"

Which studies? Was any effort made to find any downstream mutations to more closely define the particular haplogroups involved? As far as I'm aware the only E clades found in either the Middle East or Europe are subclades of E-V68 (still E1b1b1a?).

Vincent said...

"Has it been found there? What study discovered it there?"

The descendants of a past Syrian prime minister (Al-Bitar) have tested their DNA and found out to be DE-M1*. See here http://www.familytreedna.com/public/lebanon-syria-dna/default.aspx?section=yresults

"Which studies? Was any effort made to find any downstream mutations to more closely define the particular haplogroups involved? As far as I'm aware the only E clades found in either the Middle East or Europe are subclades of E-V68 (still E1b1b1a?)."

Karachanak et al. (2013) found an E-M96* in Varna, Bulgaria. They tested E-M35 and downstream sub-clades and E-M2. I compared the haplotype to E-M33* and E-M75 haplotypes in the literature and I didn't find a similarity. So that haplotype is probably a genuine E-M96*. Also, Ferri et al. (2010) found an E-M96* in their Gheg Albanian sample. They tested only for E-M35. I compared the haplotype to the same haplotypes as above and found no similarity. Badro et al. (2013) found 2 E-M96* in their Turkish sample, and Zalloua et al. (2008) found several Lebanese, Cypriots, Syrians and Palestinians to be E-M96*.

About E-M33*, it has been found in 5 Portuguese individuals by Gonçalves et al. (2005), in some Syrians by Zalloua et al. (2008), in North-Eastern Italy by Battaglia et al. (2008), in Costanta, Romania by Bosch et al. (2006), in both Italian- and Albanian-speaking individuals from Calabria, Italy by Semino et al. (2004). 2 E-M33* haplotypes were also found in Reutte, Austria by Erhart et al. (2012). They were labeled E-M96, but I compared them to other E-M33 haplotypes and found that they were similar. Niederstätter et al. (2012) also found an E-M33 in East Tyrol. The haplotype was identical to one of those found by Erhart et al.

So, it seems that there is hardly an exclusive presence of E-V68 and sub-clades in Eurasia (I think you mean E-M35, because besides DE-M1*, E-M96* and E-M33*, haplotypes of E-M35*, E-M123*, E-V257*, E-V68*, E-M78*, E-L618* etc. have all been found in Europe and Asia, too, and continue to be reported in studies dealing with European and Asian populations).

Apart from those two questions you asked, can't you also tell what Y-DNA haplogroups were responsible for the start of UP cultures in the Levant, Europe and Africa if not E-M96 and its sub-clades?

terryt said...

"so far no E1b1-P2* has been found, as shown by Trombetta et al. (2011)".

I was certainly under the impression it had been found in Ethiopia.

"The descendants of a past Syrian prime minister (Al-Bitar) have tested their DNA and found out to be DE-M1*. See here http://www.familytreedna.com/public/lebanon-syria-dna/default.aspx?section=yresults"

I looked. One example. Seems strange.

"About E-M33*, it has been found in 5 Portuguese individuals by Gonçalves et al. (2005)"

Unlikely downstream mutations were checked thoroughly at that time.

" it seems that there is hardly an exclusive presence of E-V68 and sub-clades in Eurasia (I think you mean E-M35, because besides DE-M1*, E-M96* and E-M33*, haplotypes of E-M35*, E-M123*, E-V257*, E-V68*, E-M78*, E-L618* etc. have all been found in Europe and Asia, too"

Very sporadically compared to their widespread presence in regions of Africa. As is the case with these, often single examples:

"Karachanak et al. (2013) found an E-M96* in Varna, Bulgaria. They tested E-M35 and downstream sub-clades and E-M2. I compared the haplotype to E-M33* and E-M75 haplotypes in the literature and I didn't find a similarity. So that haplotype is probably a genuine E-M96*. Also, Ferri et al. (2010) found an E-M96* in their Gheg Albanian sample. They tested only for E-M35. I compared the haplotype to the same haplotypes as above and found no similarity. Badro et al. (2013) found 2 E-M96* in their Turkish sample, and Zalloua et al. (2008) found several Lebanese, Cypriots, Syrians and Palestinians to be E-M96*".

I don't think you have yet made a convincing case for E having first developed anywhere but within Africa.

"can't you also tell what Y-DNA haplogroups were responsible for the start of UP cultures in the Levant, Europe and Africa if not E-M96 and its sub-clades?"

Have you checked out Dienekes' latest post concerning Y-DNA K:

http://dienekes.blogspot.co.nz/2014/06/refined-structure-in-haplogroup-k-m526.html

K's expansion is obviously pre Upper Paleolithic and so P and its descendants as well as C1 were quite probably responsible for the development and spread of the UP in Eurasia. And R certainly reached Africa.

Rob said...

@ Terry T

" It is unlikely that Sardinian E is anything other than a very early arrival there. "

-> Speculation does not make for proof. Until you show me aDNA from Sardinia, I'll be happy to point out that modern Sardinians are not so 'ancient' as we think, and are just as mixed and novel as other peoples - not to mention the very recent north African and Arab admixture in them, which might account for their apparent "Neolithicness".

You might be correct about Hg E and norther Africa, or not. Whatever the case, you cannot invent your own theories when no data exists to account for it. NO archaeological evidence exists for any direct contacts between north Africa and Europe, even southern / Aegean Europe. (and no, the Bell Beakers did not come from north Africa , either)

Vincent said...

"I was certainly under the impression it had been found in Ethiopia."

Your impression is wrong. There is currently no sample of E-P2*.

"Very sporadically compared to their widespread presence in regions of Africa."

Sporadic but definitely present, unlike ancestral haplogroups which you indicate as potential markers of UP populations in West Eurasia. West Eurasia simply lacks ancestral clades except E-M96*, and this means something in my opinion. Why do we not observe ancestral haplogroups compatible in age to the initial UP in West Eurasia other than E-M96*?

"I don't think you have yet made a convincing case for E having first developed anywhere but within Africa."

The same can be said for you and the African origin theory.

"Have you checked out Dienekes' latest post concerning Y-DNA K: 

http://dienekes.blogspot.co.nz/2014/06/refined-structure-in-haplogroup-k-m526.html

K's expansion is obviously pre Upper Paleolithic and so P and its descendants as well as C1 were quite probably responsible for the development and spread of the UP in Eurasia. And R certainly reached Africa."

K-M526 originated in Southeast Asia 40 kya (range: 35-45 ky). This is definitely out of the initial period of the UP in Europe (47 kya) and spacially too far removed from it, therefore K-M526 and all its sub-clades can be excluded as markers of the initial UP populations in West Eurasia.
Regarding Y-DNA C-V20 and upstream sub-clades of C, I think that an introduction in Europe in UP times is possible, but at the moment we don't have enough data to confirm it at least mathematically. However, I tend to view Y-DNA C as a newcomer from Central Asia in Eastern Europe 40-35 kya. Kostenki 14 carried C-V20 or other clades of C, in my opinion. Experts indicate a similarity to modern Melanesians in cranial traits for Kostenki 14, while the majority of contemporary specimen from the rest of Europe are relatively different from Kostenki 14 and resemble closely African specimens (Mechta-Afalou, Hofmeyr) which were the product of the back-migration of DE* and E* into Africa.

terryt said...

" Until you show me aDNA from Sardinia, I'll be happy to point out that modern Sardinians are not so 'ancient' as we think"

I don't follow your reasoning here. We all surely know that aDNA is a better indication of overall ancestry that is haploid DNA but here we are specifically discussing Y-DNA E. I have no doubt that Sardinia was settled some time during the Neolithic. We know that E-V68 is present on the island and is unlikely to have entered before the early Neolithic. We can also be sure it must have arrived by sea. Therefore we can be reasonably sure that a population containing that haplogroup must have been capable of sea voyaging, probably during the early Neolithic.

" you cannot invent your own theories when no data exists to account for it".

I have not 'invented a theory'. I have looked at the data and come to a conclusion, which is not what you appear to have done. You look to have invented a theory and have yet provided no data to support that theory. The overall distribution of E haplogroups makes it very difficult to maintain an origin for the haplogroup anywhere but within Africa.

"NO archaeological evidence exists for any direct contacts between north Africa and Europe"

The presence of Y-DNA E-V68 in Sardinia is alone evidence for such contact. I agree the contact may be very recent of course but I am yet to see any such claim. Besides which several other Y-DNA E haplogroups have been discovered in Europe, notably E-V13 in the Balkans. That too argues eloquently for some E haplogroups having voyaged around the Mediterranean.

terryt said...

"There is currently no sample of E-P2*".

That makes an East African origin unlikely and shifts E-V38's centre into West Africa.

"West Eurasia simply lacks ancestral clades except E-M96*, and this means something in my opinion".

Are you suggesting the UP originated in Africa?

"The same can be said for you and the African origin theory".

So we have to wait for more evidence one way or the other.

Rob said...

haplogroup K could still have arisen further west, ie India, then spread further east as part of an already partly differentiated population.

terryt said...

"haplogroup K could still have arisen further west, ie India, then spread further east as part of an already partly differentiated population".

Haplogroup K as a whole (including IJ and LT) almost ceetainly did arise further west. But it is extremely unlikely the haplogroups considered in this paper arose anywhere other than SE Asia. Each haplogroup considered in the paper (apart from NO and P) is confined to a very restricted geographic region. If they had moved east as an already partly differentiated population we would expect to find remnant haplogroups distributed along the route.

Vincent said...

"That makes an East African origin unlikely and shifts E-V38's centre into West Africa."

Not at all. E-V38's other major clade E-M329 is observed exclusively in East Africa. E-M215* has been observed once only in Khorasan, North-East Iran. So I think that E-P2 originated somewhere between Khorasan and East Africa, i.e. Levant.

"Are you suggesting the UP originated in Africa?"

Don't put in my mouth words I have never spoken. West Eurasian UP originated in the Levant, because datings to the East, West, North and South of the Levantine region are more recent than datings in the Levant. I suggest that the UP in West Eurasia (Aurignacian and Gravettian) and in Africa (Dabban and Iberomaurusian) had a common origin in the Levant and were the product of multiple migratory events from the Levant at different times. The populations who produced these lithic industries were either carriers of Y-DNA DE or E (and sub-clades).

I see you have completely ignored my paragraph on K-M526. You have failed to suggest other haplogroups as candidates for the spread of UP cultures in West Eurasia and you continue to ignore the evidence I have provided: ancestral E-M96* and E-M33* in the Levant and Europe, the impossibility of K-M526 as having had any role in West Eurasian UP until the Solutrean period (23 kya). Yeah, the Solutrean, because there is a lot of evidence for population shifting in the European EUP-to-LUP transition. Individuals living before the LGM were tropical-adapted and possessed long limbs and athletic body types, and were usually over 1.80 cm tall. Post-LGM individuals were instead short-statured (1.65 cm), cold-adapted with short limbs, and among them there was a definite brachycephalic element which was absent in pre-LGM individuals (who were virtually all dolichocephalic). Plus, we observe for the first time high crania, which also were absent in pre-LGM populations. All this means that, apart isolated cases of stand-out individuals in pre-LGM Europe (Grimaldi, probably Y-DNA A, and Kostenki 14, Y-DNA C) the pre-LGM population was homogeneous and probably monophyletic, while during and after the LGM we observe a relatively mixed population in which the pre-LGM element is virtually absent or at best very diluted. That's why we don't see so many traces of E-M96* in Europe today.

Rob said...

@ Terry.
By aDNA I mean ancient DNA (not autosomal DNA = atDN).
Sorry for any confusion

terryt said...

"E-V38's other major clade E-M329 is observed exclusively in East Africa"

I see in the latest phylogeny at ISOGG E-M329 has been downgraded to a separate branch within E1b1a-V38. That makes complete sense. We have separate West African (E1b1a1) and East African (E1b1a2) branches within E1b1-P2.

"E-M215* has been observed once only in Khorasan, North-East Iran".

But as a whole E1b1b-M215 primarily Northeast Africa. The presence of a lonely E-M215* in Khorasan is hardly convincing proof of origin there. We know that E1b1b-M215 became very widespread from eastern to western Mediterranean and inland from there.

"So I think that E-P2 originated somewhere between Khorasan and East Africa, i.e. Levant".

Thinking E1b1-P2 may have originated there is OK. But where has E1b2-P75 been found? That would provide a much better indication as to where the two of them originated. It's mentioned in this paper as being 'newly discovered' but the authors don't say where as far as I can see:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2336805/

"I see you have completely ignored my paragraph on K-M526".

I ignored it because I don't think haplogroup dating is very accurate yet, and probably never will be. But according to dates Rokus supplied at the relevant post K-M9 is 47,500 years old. Therefore haplogroup IJ must be older than that, and must have been present somewhere in SW Asia.

"You have failed to suggest other haplogroups as candidates for the spread of UP cultures in West Eurasia"

There you go: IJ. And G is presumably much the same age, or at least its ancestral version. G was probably the first F-derived haplogroup to split off.

"you continue to ignore the evidence I have provided: ancestral E-M96* and E-M33* in the Levant and Europe"

I haven't ignored it. I have always provided reasons why the very limited presence of those haplogroups outside Africa is hardly convincing proof for an origin outside Africa.

Vincent said...

"Thinking E1b1-P2 may have originated there is OK. But where has E1b2-P75 been found? That would provide a much better indication as to where the two of them originated."

E1b2-P75 also originated in the Levant and migrated into Europe. There is no study from which we have information on its distribution, but I have found this forum discussion: http://www.anthrogenica.com/archive/index.php/t-99.html

Four people there are P75+ and their locations are:

- Canary Islands
- Rafidya, Palestine
- Suffolk, England
- Nottingham, England

As the majority of them are Eurasians and we know that Canary Islands' male population was almost wiped out completely by the Spanish colonists, it would be not so crazy to assume that the individual is probably of Spanish descent.

E1b1-P2, E1b2-P75 and almost all major sub-clades of E-M96 (itself included) originated in the Levant or in other parts of Eurasia. Anyway, among other things, why do we not observe significant levels of African admixture in Eurasia as it would have been expected if E began to spread out of Africa merely 10.000 years ago?

Vincent said...

"I ignored it because I don't think haplogroup dating is very accurate yet, and probably never will be. But according to dates Rokus supplied at the relevant post K-M9 is 47,500 years old. Therefore haplogroup IJ must be older than that, and must have been present somewhere in SW Asia."

I have read his comment. Does he have evidence for what he says? Did he analyse STR diversity of K-M9 and the other haplogroups he has provided datings for? If haplogroup dating is not accurate, why do you favour Rokus' datings more than others' (which are more reliable)?
IJ* has only been found in Iran and most people agree with a dating of 35-40 ky.

terryt said...

Thanks for the link regarding E1b2-P75. From the link:

"E1b1-PN2 is very likely to have arisen in Eastern Africa (Trombetta et al. 2011 (http://www.plosone.org/article/info:doi/10.1371/journal.pone.0016073))".

I find that difficult to believe as E1b1a-V38 is almost entirely West African. Just the sub-branch E1b1b-M215 can be reasonably considered East African. Therefore it is difficult to draw any conclusion as to which was its region of origin. Especially when we consider another comment in the link:

"E1a is primarily West African (but present in Sudan as well)".

So that places E1a in West Africa as well as E1b1a.

Your comment:

"E1b2-P75 also originated in the Levant and migrated into Europe. ... Four people there are P75+ and their locations are ..."

That distribution is just as consistent with a West African origin for the haplogroup.

"E1b1-P2, E1b2-P75 and almost all major sub-clades of E-M96 (itself included) originated in the Levant or in other parts of Eurasia".

I cannot understand how you can be so confident when making that statement.

"Anyway, among other things, why do we not observe significant levels of African admixture in Eurasia as it would have been expected if E began to spread out of Africa merely 10.000 years ago?"

We know haplogroups are hardly rigidly defined by autosomal DNA. We could be dealing with a considerably admixed population that spread around the Mediterranean. The haplogroup distribution definitely tells us that some E haplogroups reached islands in the Mediterranean at some time. We also know that most Mediterranean islands were uninhabited until the Early Neolithic. Those are facts to which we must accommodate the haplogroup evidence.

Vincent said...

The quotes from the discussion are irrelevant to the topic we are discussing. The use of geographic labels applied to E sub-clades by the forum members is merely on the basis of distribution and frequency. Before 3-4 kya most of Africa had much lower frequency and a much more limited distribution than at present. And if you really think that frequency is important, would you even try to think of R1b as Cameroonian or Irish in origin? I don't think you would.
You can see that Canary Islanders less than 2000 years ago had 53% Y-DNA E (against modern Moroccans' ~80%; Moroccans are thought to stem from the same ancestral population who gave origin to the Islanders) and much higher frequency for rare lineages in today's Maghreb.
From the University of Khartoum report on Ancient Nubia we know that "Haplogroups A-M13 was found at high frequencies among Neolithic samples. Haplogroup F-M89 and YAP appeared to be more frequent among Meroitic, Post-Meroitic and Christian periods. Haplogroup B-M60 was not observed in the sample analyzed." Y-DNA E was therefore much lower (or absent? The report is unclear) in Neolithic Nubia (which began 10-6 kya according to sources) and rises in frequency at much later times (this also suggests A-M13 in the Neolithic was pushed south by incoming peoples from the North, i.e. haplogroup E sub-clades, IMO E-V12 and E-V22) and the very common B-M60 (in today's Sudan) was absent back then.
All this (together with the unusual Y-DNAs of recently-analysed Egyptian pharaohs) demonstrates that Y-DNA E becomes much less common in Africa the more back in time one goes.

"That distribution is just as consistent with a West African origin for the haplogroup."

Nothing is more wrong than what you have written. This distribution points exclusively at an Eurasian origin.

"I cannot understand how you can be so confident when making that statement."

I have gathered too much evidence to ignore it. Distribution of ancestral sub-clades and STR variance of the main sub-clades of haplogroup E, among others, are the main elements for which I suggest a Levantine/West Asian origin for E-M96 and most of its ancient sub-clades. One day, if you show me even one decisive proof for what you say (African origin theory), I will cease to defend my theory instantly. But until that day, I will always support what I (and the current evidence) say.

"We know haplogroups are hardly rigidly defined by autosomal DNA. We could be dealing with a considerably admixed population that spread around the Mediterranean."

I find that unlikely. If E-M35 was really African, we would see much more African autosomal ancestry in modern West Eurasians, whatever the degree of admixed ancestry (which is unlikely to have been so considerable as you suggest) E-M35 carried.

terryt said...

"Does he have evidence for what he says? Did he analyse STR diversity of K-M9 and the other haplogroups he has provided datings for?"

I think the dates are based on the findings of the recent paper on Y-DNA K. I also presume he has based his dates on reasonable grounds, although I am loath to take them as absolutely correct.

"If haplogroup dating is not accurate, why do you favour Rokus' datings more than others' (which are more reliable)?"

Rokus' dates are not too far removed from several others.

"The quotes from the discussion are irrelevant to the topic we are discussing".

I disagree. You are trying to understand what Y-DNA might have accompanied mt-DNA U6 into Africa. You have claimed Y-DNA E fits the bill whereas I believe it is an unlikely candidate.

"The use of geographic labels applied to E sub-clades by the forum members is merely on the basis of distribution and frequency".

Which gives a reasonable guide to prehistoric distribution unless you're postulating wholesale haplogroup replacement. Once you start using that as an explanation for haplogroup distribution it's possible to postulate almost any scenario. Such as German's out of America claim.

"And if you really think that frequency is important, would you even try to think of R1b as Cameroonian or Irish in origin? I don't think you would".

Of course I wouldn't think of it as Irish. It is a completely different clade. The fact its name starts with the same letter merely indicates ancient connection.

"You can see that Canary Islanders less than 2000 years ago had 53% Y-DNA E"

And they must have reached those islands by sea. They certainly didn't walk there.

"against modern Moroccans' ~80%; Moroccans are thought to stem from the same ancestral population who gave origin to the Islanders"

So, what's your point?

"All this (together with the unusual Y-DNAs of recently-analysed Egyptian pharaohs) demonstrates that Y-DNA E becomes much less common in Africa the more back in time one goes".

In the parts on mainly northern Africa so far studied. We have no information on the ancient Sahel population.

"This distribution points exclusively at an Eurasian origin".

No. It points to an (apparently) exclusively Eurasian modern distribution. Unfortunately for your claim its closest relations are not Eurasian but African.

"I have gathered too much evidence to ignore it".

I get the distinct impression you have started out hoping to find evidence of a Eurasian origin. And one usually finds evidence to support what one is looking for.

Vincent said...

"No. It points to an (apparently) exclusively Eurasian modern distribution. Unfortunately for your claim its closest relations are not Eurasian but African."

Wrong. E-P75 is Eurasian and its closest relations (his brother E-P2* and his nephew E-M215*) are in Eurasia. To understand what I mean, I address you to my latest comment on the article about the ancient Eurasian admixture in HOA populations.

"I get the distinct impression you have started out hoping to find evidence of a Eurasian origin. And one usually finds evidence to support what one is looking for."

Sorry, but I don't have to hope anything. The evidence just speaks by itself and is happily catering for me.

terryt said...

"its closest relations (his brother E-P2* and his nephew E-M215*) are in Eurasia".

I can't understand how you come to that conclusion although I agree branches of them are actually present 'in Eurasia'. But they could hardly be claimed as primarily Eurasian. Even E1b1b-M215 is far more widespread with more basal branches within Africa than are present in Eurasia. And E-M215's brother, E1b1a-V38, is even more predominately African.

"The evidence just speaks by itself and is happily catering for me".

The 'evidence' is obviously open to interpretation then.

Vincent said...

"Even E1b1b-M215 is far more widespread with more basal branches within Africa than are present in Eurasia."

None of the sub-clades of E-M215 (there are only two of them, actually) is absent from Eurasia. There's even ancestral E-M215* in Eurasia but not in Africa. I have not checked it properly, but looking again in my "database" it appears that there's even more E-M35* in Eurasia than in Africa, and possibly also E-M78* shows the same pattern, not to mention the two Sardinian V68* found in one study and the two Dutch V68* in the E-M35 Phylogeny Project, the Spaniards, Corsicans and Sardinians who are V257* and the M123* individuals who are widespread in Eurasia and found as far as Central Asia with highest world concentration in Northern Portugal and Galicia.

Unknown said...

Does anyone know what the frequencies are based on because this study shows much higher frequencies than any previous study I've seen. A frequency of 30-40% U6 in Ethiopia seems like too much, but it doesn't seem to say anywhere how many samples were used for each location. Has anyone else found this? Thanks.