October 18, 2012

Intriguing North African Neanderthal admixture paper

This new study aims to quantify Neanderthal admixture in North Africans in order to evaluate the unlikely but popular-in-some-circles hypothesis that the signature of Neanderthal admixture in non-African Homo sapiens could be caused by structure in Africa prior to the migration into Asia and beyond ("ancient African structure model" hereafter).

It does seem to add evidence against the "ancient African structure model" and, incidentally, in favor of the mainstream interpretation that North Africans have large amounts of West Eurasian ancestry (questioned without clear base in some Africanist circles). However it also arrives to some odd results regarding East Asian Neanderthal admixture which are not addressed at all in the study.

Federico Sánchez Quinto et al., North African Populations Carry the Signature of Admixture with Neandertals. PLoS ONE, 2012. Open access ··> LINK [doi:10.1371/journal.pone.0047765]


One of the main findings derived from the analysis of the Neandertal genome was the evidence for admixture between Neandertals and non-African modern humans. An alternative scenario is that the ancestral population of non-Africans was closer to Neandertals than to Africans because of ancient population substructure. Thus, the study of North African populations is crucial for testing both hypotheses. We analyzed a total of 780,000 SNPs in 125 individuals representing seven different North African locations and searched for their ancestral/derived state in comparison to different human populations and Neandertals. We found that North African populations have a significant excess of derived alleles shared with Neandertals, when compared to sub-Saharan Africans. This excess is similar to that found in non-African humans, a fact that can be interpreted as a sign of Neandertal admixture. Furthermore, the Neandertal's genetic signal is higher in populations with a local, pre-Neolithic North African ancestry. Therefore, the detected ancient admixture is not due to recent Near Eastern or European migrations. Sub-Saharan populations are the only ones not affected by the admixture event with Neandertals.

Europeans and Africans

First of all they take a look at the structure of European and African populations regardless of Neanderthal ancestry, which is probably of interest on its own right:

Figure 1. Results of the ADMIXTURE analysis (k = 4) with North African populations

Sadly, the Tunisian Berbers who define the North African polarity, taken from Henn 2011 and 2012, are a recognized (by Henn herself) as a highly endogamous population that behave a bit weirdly. I do not think that in this specific context they are distorting the results significantly (compare with my own exercise earlier this year) but I do think that the sample should not be used anyhow.

Otherwise we can see that:
  • There are both West Asian and European components in North Africans (the red "Tunisian Berber" component is also probably of West Eurasian ultimate origin, from my previous work, where I carefully looked at FST distances among components - however some nuances, like a probable pre-OoA* residue in North Africans are not evident in this analysis, which is too shallow).
  • There is some African (mostly North African) ancestry in Iberia (more towards the West, as we know from other studies).
  • There is some minor West Asian ancestry in Europe, consistent with a less important penetration in Neolithic (and maybe also post-Neolithic) times.
  • There is Tropical African ancestry in North Africans as well (even if the Luhyas are probably not the best proxy).

Neanderthal blood in North Africans

The main goal of the paper is however to estimate Neanderthal ancestry in North Africans, with the following results:

Table 2. Estimates of Neandertal ancestry in North African populations, along with European, Asian and Sub-Saharan African groups. [Notice that GIH is mislabeled, it actually means: Gujarati Indians from Houston].

The results for Africa and West Eurasia are more or less within expectations, what is very anomalous however is the East Asian results of c. 200% Neanderthal ancestry (relative to CEU) but let's leave that for later.

We could well discuss however why, if North Africans are some 80% West Eurasian by ancestry their detected Neanderthal admixture is so low (c. 50-60%). It may well be a matter of diverse ancestral influences (not all West Eurasia is Europe), and would be consistent maybe with the relatively low amounts of Neanderthal admixture found in Indians (GIH, 84%). Sadly the Qatari sample was not tested for Neanderthal admixture, which would have been interesting on its own right and also in relation to North Africa

Whatever the case this is the authors' map of estimated North African Neanderthal ancestry relative to Euro-Americans (CEU, which have c. 2.4% Neanderthal ancestry):

Figure 3. Neandertal genetic introgression in North African populations as a fraction of that found in Europeans [colored bars represent K=4 apparent ancestry as above].

Maybe the most interesting population here is MOS (Southern Moroccans) which in my December 2011 Admixture analysis of North Africans appeared to show significant amounts (14.4%) of a very distinct component that I judge as Aterian remnant (in other populations not higher than 1.5%).

While the authors think that the findings are supportive but can't totally prove that the Neanderthal admixture appearance is real and original from Eurasia, I find the very low Neanderthal admixture signal in Southern Moroccans as very clear evidence that pre-OoA structure (which is preserved as minority component in this population apparently) that the Neanderthal admixture signal has nothing to do with any pre-OoA North African specificity. 

Otherwise South Moroccans should show a neutral signature and not this outstanding depression of Neanderthal admixture. There are surely other factors at play there (greater Tropical African admixture, lower European admixture) but this "Aterian" component should be high in Neanderthal signature if there was any pre-OoA African structure at all. 

Some people may want to cling to the proverbial burning nail, maybe arguing that the Neanderthal-like structure was in Egypt and not NW Africa. Not only that looks too much like the West Asian admixture episode that the mainstream theory argues for but it also ignores the cultural and anatomical similitudes between NW African Aterian peoples and those from Palestine in the context of the likely window for the archaeologically supported OoA: 125-90 thousand years ago. 

So, in my understanding the results of this paper are quite supportive of the mainstream (Asian admixture) model, rather dismissing the African structure counter-hypothesis, even more than the authors themselves admit to. 

European degrees of Neanderthalness?

I am sure that some readers will have noticed that there is an appearance of greater Neanderthal values among Basques (129% rel. to CEU) and Iberians (115-118%). However the differences are very small and not more marked than, for example North Tunisians (138%).

Considering the other issues that the results of this paper imply, I cannot really argue that these differences are meaningful at all. However they sparse data shown for other parts of West Eurasian would be consistent with greater Neanderthal admixture in Europe than in West Asia (and the corresponding gradation in Europe depending of Neolithic ancestry), with an apparent highest value in a population usually considered as a Paleolithic continuity genetic refuge: Basques.

But only further data and specific analysis can confirm or question this appearance. 

East Asian extremely high Neanderthal values

This is what is totally puzzling in this study and what can really bring to question all other conclusions. The authors appear to find that East Asians (CHB, CHD, JPT) have values of Neanderthal admixture that are double than those of West Eurasians. 

This clashes with all we know. In fact one of the findings of Green 2010 was that all Eurasian-derived populations have similar Neanderthal admixture values, c. 2.5%, with lesser population or individual variations only. This has only been confirmed once and again in successive studies, for example John Hawks earlier this year, who actually claimed that Europeans held slightly more Neanderthal admixture than East Asians.

Then why do the Sánchez Quinto team find such brutal East Asian Neanderthal admixture values? No idea. They don't even mention the matter in all the paper. And that is a big flaw, which can only cause readers to doubt the whole paper or even the methodology altogether. 

Any ideas?


*OoA: Out of Africa, it refers to the human migration episode and related founder effect (bottleneck) which populated Asia and its peripheral minor continents (Australasia, Europe, America) with our species, Homo sapiens. It is a critical episode in human prehistory and probably the time when most if not all Neanderthal admixture happened. But, in principle, already in Asia.


  1. Very good post, congratulations.

    First, I'd like to say: that's completely crazy! It just contradicts other "neandertal admixture estimations" we know so far.

    For example, the admixture percentages found in this study, in the first one (Green et al. 2010), in John Hawks' blog and finally in the "Denisovan" one (and others I've not mentioned) completely contradict each other.

    Which one is true or closer to the truth? Why do they say North Africans share more alleles with neandertals when East Asians show a much greater amount of them? How can Indians show much lower levels than other non-africans?

    Second: I just don't understand that, if northern Europeans are more similar to mesolithic hunter-gatherers (according some recent studies with ancient DNA) and show some "Asian" admixture as well, how can they be less neandertal than North Africans, who are admixed with tropical Africans?

    I find too many contradictions between these studies, and there are many populations missing there, they just picked up a dozen, in my opinion it's not enough to see patterns of such an ancient admixture (if there are any). I'm not going to believe any of them.

    Just another one to put in the list! Very interesting but frustrating at the same time. One wonders if scientists really read and document themselves about what's going on in their field before writing the conclusions.

    1. Re. the first pert of your comment, Neanderthalerin, implies an analysis of the methodology that, sincerely, I do not feel able to do at this point, possibly never. Someone with better grasp of the statistical-mathematical aspects implicated should do instead.

      Re. the second part, there is absolutely no evidence of your statement that reads "northern Europeans are more similar to mesolithic hunter-gatherers". That's mostly wishful thinking speculation by some people. I'll be glad to discuss the particulars if you can point me to specific contextualized claims, because there is probably very little facts and lots of over-interpretation in this. To this moment I remain persuaded of SW European specific autosomal components being at least as "aboriginal" as NE European ones (what is coherent with Neolithic mtDNA apparent replacement fists, and later semi-vanishing, advancing between them in the Balcano-Danabian wedge).

      However, considering the many doubts that the East Asian figures cause, I don't feel at all comfortable with the rest of the results, so I would not draw conclusions from them unless supported with other studies also.

      "One wonders if scientists really read and document themselves about what's going on in their field before writing the conclusions".

      I'd say so, up to a point, but still they manage to be incredibly sloppy or confusing sometimes.

    2. "Re. the second part, there is absolutely no evidence of your statement that reads "northern Europeans are more similar to mesolithic hunter-gatherers". That's mostly wishful thinking speculation by some people. I'll be glad to discuss the particulars if you can point me to specific contextualized claims,"

      It seems like (according to some ADMIXTURE experiments) hunter gatherers show a relatively high Altantic-baltic component which is slighty higher in northern Europe. Also, it seems too that some other components appeared in neolithic farmers and are more common among southern Europeans. It has been noted that northern Euros are somewhat more "Asian" than southerners, and mesolithics seem to be more 'Asian' too. Mesolithics from Iberia show high levels of the "northern European" component, even higher than present day people living in Iberia. I'm not sure, though, and more ancient DNA is needed to address this issue.

      "To this moment I remain persuaded of SW European specific autosomal components being at least as "aboriginal" as NE European ones (what is coherent with Neolithic mtDNA apparent replacement fists, and later semi-vanishing, advancing between them in the Balcano-Danabian wedge). "

      That's true too. SW European components have been noticed in La Braña specimens.

    3. The specific reference you had in mind is important because I can now (and only now) address the specifics. Dienekes found that La Braña composite (in principle pre-Neolithic) is:

      For his K7b set: 9.3% African and 90.7% Atlantic_Baltic.

      Which are hight in the Atlantic-Baltic component in his K7b zombie set? Essentially Western and Northern Europeans, from Basques to Finns (>75%) but also at lower levels Spaniards, French, etc. It's a very common dominant component today, so, judging on it alone, most modern European blood would be Paleolithic.

      For his K12b set: 45% Atlantic_Med, 41.6% North_European, 10.3% East_African, 1% Sub_Saharan

      Who are nearly 50-50 Atlantic-Med and North-European in his K12b zombie analysis? The same ones very roughly (but less so those from NE Europe, which are dominated in the past as today by the North-European component), for example Argyll, British, CEU, Cornwall, French (but not North Basque), etc. French are maybe the ones who compare best with La Braña, because they retain the slight dominance of the AM over the NE component.

      "and mesolithics seem to be more 'Asian' too"

      African actually but only La Braña, and that is probably a West Iberian specificity, whose remains we still find in that area precisely, all the way between Asturias to Western Andalusia (IMO a Solutrean founder effect in relation with the genesis of Oranian in North Africa).

      "That's true too. SW European components have been noticed in La Braña specimens".

      Most important.

      Of course it could still be that La Braña an their flexed burials are very early Neolithics (misclassification?) and that would change some things. But it would probably also be the case with any new aDNA research: there are too few studied specimens to know with any certainty as of now.

      Also let's not forget that the "Mesolithic" specimens from Gotland are in fact regressive Neolithic ones. They may have more Paleolithic blood but only because they come ultimately from the Eastern European Neolithic, which looks rooted in the local Paleolithic and not a foreign arrival in any way, but not because they represent direct Paleolithic Continuity in Sweden.

      Of all aDNA specimens considered by Dienekes in his comparisons, none is 100% sure Paleolithic but only La Braña has real chances of being truly Epipaleolithic. All the rest, including the Gotland samples, are fully Neolithic in chronology and roots, even if they lived more of fishing than farming.

  2. What is interesting to think about is that there were extensive river systems in the Sahara sporadically during past ice ages. The Mediterranean also has virtually dried up in the past, maybe several times. Both of these conditions would seem to make a path from sub-Sahara Africa to North Africa and from North Africa to Iberia (and the rest of Europa).

    The strange bit is how much Neanderthal is in East Asians. That is astounding.

    1. The Mediterranean Sea has not dried up in the Human Age. Gibraltar Strait was closed last time (Messinian salinity crisis) c. 6-5.3 million years ago, long before Homo habilis. People may have crossed Gibraltar Strait, which was quite narrow in the Ice Ages, but always on boat or raft.

      "The strange bit is how much Neanderthal is in East Asians".

      How much the study seems to find (not the same as "is"). It's so strange that I remain highly skeptic.

  3. Here is an interesting article about Neanderthals as ancient mariners. Perhaps this would impact some current theories.

    1. I did mention it here (same source, just a note). It is the first clear evidence of Neanderthals as boat-users ("mariners" sounds a bit too exaggerated) but we have no evidence of the crossing the sea elsewhere.

      The islands were rather close to the mainland and one of the three mentioned (Lefkada) was not even an island yet (it was a peninsula until a canal was dug in antiquity). But it clearly indicates that they had the capacity, like ourselves, to use rafts or boats for short distances.

      A very different thing is that they used that capacity at all in other cases. For example there is no evidence of crossing of Gibraltar Strait before the Solutrean period, except in times long before Neanderthals, when Homo erectus/ergaster probably crossed from Africa to Europe bringing Acheulean industry with them.


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

Preliminary comment moderation is... ON (your comment may take some time, maybe days or weeks to appear).