Between the desert and the sea

By Razib Khan | January 13, 2012 12:06 am

Zinedine Zidane, a Kabyle

There is a new paper in PLoS Genetics out which purports to characterize the ancestry of the populations of northern Africa in greater detail. This is important. The HGDP data set does have a North African population, the Mozabites, but it’s not ideal to represent hundreds of millions of people with just one group. The first author on this new paper is Brenna Henn, who was also first author on another paper with a diverse African data set. Importantly the data was posted online. Unfortunately though most of the populations didn’t have too many markers. This isn’t an issue in an of itself, but it becomes a big deal when trying to combine it with other data sets. If you limit the markers to those which intersect across two data sets you start to thin them down a lot, to the point where they’re not useful. Though the the results of the paper are worth talking about, the authors claim that they’ll be putting the data online. This is important because they used a large number of markers, so the intersections will be nice (I can, for example, envisage exploring the relationship between the North Africans and the IBS Iberian sample in the near future).

As for the paper itself, Genomic Ancestry of North Africans Supports Back-to-Africa Migrations:

Proposed migrations between North Africa and neighboring regions have included Paleolithic gene flow from the Near East, an Arabic migration across the whole of North Africa 1,400 years ago (ya), and trans-Saharan transport of slaves from sub-Saharan Africa. Historical records, archaeology, and mitochondrial and Y-chromosome DNA have been marshaled in support of one theory or another, but there is little consensus regarding the overall genetic background of North African populations or their origin and expansion. We characterize the patterns of genetic variation in North Africa using ~730,000 single nucleotide polymorphisms from across the genome for seven populations. We observe two distinct, opposite gradients of ancestry: an east-to-west increase in likely autochthonous North African ancestry and an east-to-west decrease in likely Near Eastern Arabic ancestry. The indigenous North African ancestry may have been more common in Berber populations and appears most closely related to populations outside of Africa, but divergence between Maghrebi peoples and Near Eastern/Europeans likely precedes the Holocene (>12,000 ya). We also find significant signatures of sub-Saharan African ancestry that vary substantially among populations. These sub-Saharan ancestries appear to be a recent introduction into North African populations, dating to about 1,200 years ago in southern Morocco and about 750 years ago into Egypt, possibly reflecting the patterns of the trans-Saharan slave trade that occurred during this period.

The model outline here is straightforward:

– A population of West Eurasian provenance migrated across the fringe of the southern Mediterranean >10,000 years B.P. (Maghrebi)

– This was later overlain by a later West Asian migration (Near Eastern)

– A third major element here seems to be Sub-Saharan African admixture, which these authors claim is rather new (post-Roman)

Two of the methods used will be familiar to readers of this weblog. They used ADMIXTURE to generate barplots which fractionate putative ancestral components given K number of components. Second, they also use PCA to visualize the largest components genetic variation within the samples on a plane.

We cannot display this gallery

As you “move up” the K’s you note that Maghrebi populations “split” from the Near Eastern reference, the Qataris. This is supported by the PCA, which shows that there is a dimension of variation which separates Near Easterners & Europeans from Maghrebis. The authors note that this dimension is orthogonal to the Sub-Saharan African vs. Eurasian component. That suggests that the putative Maghrebi component is likely to be part of the set of “Out of Africa” populations, rather than an African population which simply experienced continuous gene flow with West Eurasians.

They also estimate a Fst, a statistic which partitions genetic variation within and between groups. The value between Sub-Saharan Africans and Europeans is ~0.15 using HGDP SNP data, and between Europeans and East Asians ~0.10.  Using the Tuscans and Qataris as European and West Asian references against the North African populations along their east-west cline they estimate Fsts from ~0.03 to ~0.06. The higher end values are from populations which are less admixed with Near Eastern elements, and the colored polygons illustrate the domain generated by ADMIXTURE Fsts across inferred ancestral components. You also see in the chart estimated time of divergence. I won’t get into the assumptions in the model, but the authors do note that ~12,000 years B.P. seems to be the low bound estimate for when the Maghbrebis diverged from other West Eurasians. This is important, because it predates agriculture.

The final set of methods outlined in this paper looked at ancestry on a more fine-grained genomic scale. To the left you see a plot where each horizontal bar represents an individual’s chromosome 1 (among a set of North Africans). Each color in that bar indicates a component of ancestry (except the black, which are centromeres). This sort of information is important, because saying someone is 50% X and 50% Y summarizes information to the point of eliding it. An individual who is a first generation product of a Chinese-European marriage is going to have the same ancestral proportions as someone who is a Uyghur for those respective populations. But a fine-scale mapping of the genomic ancestry would look very different, because the history of the admixture is very different.

There are many inferences in the paper which I won’t address. Rather, let me focus on this one assertion:

After accounting for putative recent admixture (Figure 1), the indigenous Maghrebi component (k-based) is estimated to have diverged from Near Eastern/Europeans between 18–38 Kya (Figure 3), under a range of Ne and k values. We hence suggest that the ancestral Maghrebi population separated from Near Eastern/Europeans prior to the Holocene, and that the Maghrebi populations do not represent a large-scale demic diffusion of agropastoralists from the Near East.

This is not implausible on the face of it. The component of ancestry modal in the Mozabite HGDP sample tends to have a relatively high Fst in relation to other West Eurasian groups. I had wondered if this was due to ancient Sub-Saharan African admixture which had produced a particular stabilized hybrid, but these results indicate that the component is no closer than other West Eurasians. What I’m confused and skeptical about are the range of divergence times which different papers are producing which seem somewhat implausible taken together.

There are papers which posit that East Asians separated from Europeans ~25,000 years B.P. This is in the same range as the divergence between Maghrebis and West Eurasians, but the Maghrebi genetic distance (Fst) is about 1/2 as great. Also, these sets of results which generate a “bunching” together of the separation of many extant non-African lineages in the 20-40,000 year range imply very rapid differentiation after the “Out of Africa” event, if that event did occur ~50,000 years ago (at least for most Eurasians, even assuming a revised model whereby Australian Aboriginals derive from an earlier wave). One at a time any given divergence estimate may be broadly plausible, but the literature is just not particularly coherent on this matter, and it often seems archaeologically implausible.

Citation: Henn BM , Botigué LR , Gravel S , Wang W , Brisbin A , et al. 2012 Genomic Ancestry of North Africans Supports Back-to-Africa Migrations. PLoS Genet 8(1): e1002397. doi:10.1371/journal.pgen.1002397

Image Credit: Raphaël Labbé


Comments (9)

  1. Karl Zimmerman

    One thing that surprised me about this paper is how much time is given to explore the Sub-Saharan African admixture into North Africa, when there is just a half paragraph discussing the European admixture, which is far more genetically significant for all populations except the South Moroccan and Saharawi samples. Of course, given the history of the region, there could have been continuous gene flow between Southern Europe and North Africa, to a certain extent, at least as far back as the Romans, and possibly even earlier in Libya and Egypt due to Greek influence.

    It’s also a shame that more Berber populations were not included. I think including Riffis and Kabyles in particular would be helpful. We know both populations have lower recent Sub-Saharan admixture than the Arab-speaking populations in Morocco and Algeria respectively. If their European admixture was appreciably lower, then most of “Arab” north African genetic component from Europe was probably historically recent, from the Mediterranean slave trade in the middle ages. If, in contrast, they also display high European values, then the European admixture is probably from late antiquity. The two Berber populations I mentioned above would be the most ideal to examine this because they are coastal, and presumably their ancestors lived in the same regions, and were under Roman rule.

    Of course, as the paper says, it’s also possible Qatar wasn’t the best comparison. I wonder which would be best, however. Jordan? Palestine?

    Also, I find it interesting the Fula show a significant Maghrebi component. It’s been noted that many Fulani are light for Sub-Saharan Africans, looking somewhat similar (in terms of skin tone) to the average African-American, and often having a somewhat near-eastern cast to their features. Given the recorded history of the Fula is so bound up with Islam, it’s perhaps not surprising that they would have more of a North African genetic influence than their neighbors, although it’s possible they absorbed this component earlier (maybe through links with the Tuareg, who really should be in this study as well).

  2. Onur

    The Tunisian Berber population seems to have been extremely isolated during the recent centuries. This makes it unreliable as a Berber population. Any inferences regarding Berbers and North Africans in general based on this Tunisian Berber population should be regarded with caution. The ADMIXTURE component modal in the Tunisian Berber population must be nothing but an artificial result of the extreme recent isolation of that community. This is apparent from the fact that at K=2 the Tunisian Berber individuals are indistinguishable from non- or almost non-recent-Negroid-admixed North African individuals irrespective of language while they are all distinguished from Europeans and non- or almost non-recent-Negroid-admixed Qataris.

  3. Onur

    I think what essentially genetically distinguishes the Tunisian Berber population from North Africans in general is that the ancestors of the Tunisian Berber population became isolated during the Islamic era and as a result was not genetically affected by the Trans-Saharan slave trade, which is largely an Islamic era phaenomenon.

  4. One thing that surprised me about this paper is how much time is given to explore the Sub-Saharan African admixture into North Africa, when there is just a half paragraph discussing the European admixture, which is far more genetically significant for all populations except the South Moroccan and Saharawi samples.

    the ss-african component is clear and distinct. the “european” element is more confused, because you can’t always differentiate it from “european” in some near eastern populations. IOW, it might not be a legitimate signal of “european” ancestry as such (though i suspect it is, because you see north africa in spain and sicily).

  5. Andrei

    What about the arguably undisputed European phenotypic traits of some Berber groups (fair hair & blue eyes)? Is is paleo, or is it recent introgression (i.e., Late Antiquity; e.g., Roman and Vandalic refugees from the Arab invaders; too simplistic explanation? perhaps, if one also remembers the Guanches. When and why did some Berber groups become isolated? Was it pre-Islamic, caused by environmental restrictions, or post-Islamic socio-cultural feature [ingroup vs. outgroup]?)

  6. What about the arguably undisputed European phenotypic traits of some Berber groups (fair hair & blue eyes)?

    if it’s arguable, it’s not undisputed 😉 in any case, it could be introgression, admixture, or convergence. the genetic architecture of the traits isn’t ta complicated.

  7. Eze

    It would be interesting to see whether Qataris have increased South Asian admixture compared to Saudis and Yemenites.

  8. Brenna

    They Affymetrix 6.0 data for North African populations and the Basque are now available online at
    We apologize for the short delay in getting up on the website.

  9. Grey

    I hadn’t thought about it before the post on the Vandals but if you stuck an Afrika Korps hat on his head it wouldn’t actually look that out of place.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar