June 21, 2012

Ethiopian origins (Pagani et al. 2012)

The study attempts to answer four questions:
Our current study is motivated by four questions. First, where do the Ethiopians stand in the African genetic landscape? Second, what is the extent of recent gene flow from outside Africa into Ethiopia, when did it occur, and is there evidence of selection effects? Third, do genomic data support a route for out-of-Africa migration of modern humans across the mouth of the Red Sea? Fourth, assuming temporal stability of current populations, what are the estimated ages of Ethiopian populations relative to other African groups?
Link to press release. Link the supplemental data.

The authors reiterate that modern humans left Africa 50-70kya, a hypothesis that seems to me pretty much dead in the light of recent archaeological evidence.

The lack of antiquity in the Ethiopian population, even in only the African component thereof argues against that population being ancestral to modern humans. Note that if the Out-of-East Africa hypothesis is correct, then skulls like Omo I represent ancestral modern humans and they are followed much later by modern humans anywhere else. So, while anatomical modernity may have emerged in East Africa --or maybe not; let's not forget that we have early modern skulls from the region in part because of the excellent preservation conditions and excess of scholarly interest-- there is no evidence that they spread from there.

I have little doubt that my own theory about substantial back-migration of Eurasians into Africa will eventually win the day. Of course, I am not referring to the recent (in the last 3,000 years) admixture with West Eurasians that the Ethiopian population has undergone, but rather to the more ancient migration that was probably associated with Y-haplogroup DE-YAP.

The fact that the African component of diverse African populations is more closely related to West than to East Eurasians is one piece of evidence among many for that scenario. Hopefully, it can be tested soon using whole genome data which may have enough density to detect much older admixture events.

UPDATE I: Since the dates in the paper are based on ROLLOFF, a piece of software that is not publicly available more than a year after its announcement, and which contradicts other software released by the same authors, I will take the Queen of Sheba stories circulated in the media with a huge grain of salt.

The American Journal of Human Genetics, 21 June 2012 doi:10.1016/j.ajhg.2012.05.015

Ethiopian Genetic Diversity Reveals Linguistic Stratification and Complex Influences on the Ethiopian Gene Pool

Luca Pagani et al.

Humans and their ancestors have traversed the Ethiopian landscape for millions of years, and present-day Ethiopians show great cultural, linguistic, and historical diversity, which makes them essential for understanding African variability and human origins. We genotyped 235 individuals from ten Ethiopian and two neighboring (South Sudanese and Somali) populations on an Illumina Omni 1M chip. Genotypes were compared with published data from several African and non-African populations. Principal-component and STRUCTURE-like analyses confirmed substantial genetic diversity both within and between populations, and revealed a match between genetic data and linguistic affiliation. Using comparisons with African and non-African reference samples in 40-SNP genomic windows, we identified “African” and “non-African” haplotypic components for each Ethiopian individual. The non-African component, which includes the SLC24A5 allele associated with light skin pigmentation in Europeans, may represent gene flow into Africa, which we estimate to have occurred ∼3 thousand years ago (kya). The African component was found to be more similar to populations inhabiting the Levant rather than the Arabian Peninsula, but the principal route for the expansion out of Africa ∼60 kya remains unresolved. Linkage-disequilibrium decay with genomic distance was less rapid in both the whole genome and the African component than in southern African samples, suggesting a less ancient history for Ethiopian populations.



Etyopis said...

Haven't read the study but from the figures posted here, the biggest elephant in the room is where are the CENTRAL CUSHITIC SPEAKERS?. Central highland cushitic speakers, also known as agews in addition to the Northern Cushitic speakers like the Beja from Sudan and Eritrea are not sampled? You can not make the distinction such as “revealed a match between genetic data and linguistic affiliation” (here, I am presuming they are talking between semitic/cushitic speakers?) without sampling central highland cushitic speakers and the Beja, sorry but that is simply Ethio-Genetics 101, nevertheless it is more data than we have now and I'm certainly appreciative of that, although I have very little confidence in any kind of dating using the recombining portions of our genomes.

Another thing is labeling the 'Tigray' in the reference map as being located South West of the 'Amhara' label, this is just simply incorrect.

The Omotic Ari placement in the intra African PCA is exactly where I would expect them to be placed, in fact that plot closely resembles my intra African PC1 vs PC3 plot where the diagonal axis which I designate as the 'Afroasiatic nucleus' shows up on there plots too: http://ethiohelix.blogspot.com/2012/03/afrasans-in-genome-wide-context.html

One other point I noticed in the abstract, “The African component was found to be more similar to populations inhabiting the Levant rather than the Arabian Peninsula,” is this a typo? Or did they really mean the 'non-African' component.

Anonymous said...

Thanks for this paper.

Onur Dincer said...
This comment has been removed by the author.
Andrew Oh-Willeke said...

@ Etyopsis re: criticism of the lack of Central Cushitic speakers. More is better, but grant money is finite. I suspect that the issue is simply one of limited resources and convenience sampling that maximizes those resources, and not methodological preference. I'm thankful for every bit of new data we get and this study, with 242 new genomes is a big improvement over the status quo in an undersampled region.

Re: back migration.

There is no real dispute that there has been some back migration from Eurasia to Africa. Y-DNA hgs T and R1b-V88 and mtDNA hgs M1 and U6, as well as autosomal data such as this study, for example, make a very strong case for this. There is also good evidence to suggest that there was at least, a relatively recent back migration that gave rise to the Ethio-semitic languages, and a more ancient back migration (probably late Upper Paleolithic or early Neolithic) associated, for example, with some of those particular uniparental haplogroups.

The case for a back migration of Y-DNA haplogroup E, the African branch of Y-DNA DE, is not well supported, however. All of the subhaplogroups of Y-DNA hg E which are found in Eurasia are from phylogenies rooted in Africa and the SW European and SE European Y-DNA hg E subhaplogroups don't have a common European source. African has basal branches of Y-DNA hg E not found anywhere else (e.g. E2). Y-DNA hg E is predominant in most of Africa. Y-DNA D is a rare hg that has a sporadic Asian distribution with a deep rooted divide between D2 a Japanese subhaplogroup, and D*, D1 and D3 found in a network centered on the Andaman Islands and Tibet and seeping out at low frequencies from Tibet into Central Asia - it is absent from West Eurasia, the Americans, Melanesia and Australia. There are just eight known cases of individuals from paragroup DE*, two in Tibet, and six from West Africa, although they may be some in the Andaman Islands. The YAP mutation that defines DE and associated hinted at research from the 1000 Genomes project suggest that DE's divide is vary ancient, dating to roughly the Out of Africa era, or at the latest, early Upper Paloelithic. While mutation dating is problmeatic, it isn't so irrelevant that one cannot say that the DE divide is much more ancient than the M1/U6/Y-DNA T/Y-DNA R1b-V88 backmigrations or events giving rise to Ethiosemitic languages. The most basal lineages, paragroup E*, have been found in a single Bantu-speaking male from South Africa, amongst pygmies and Bantus from the Cameroon/Gabon region, and in two individuals from Saudi Arabia. The geography of D is consistent with long range maritime migration to successive uninhabited islands or unoccupied areas in the wake of migrations occupying most territory by CF clade members (and not of the areas were it is found had modern human populations prior to about 30kya), followed by subsequent secondary migrations inland; the distribution of E in African is continuous and terrestrial, and is common in three of the four major linguistic groups of Africa (Afro-Asiatic, Nilo-Saharan, Niger-Congo; but not Khoisan). Y-DNA hg E subhaplogroups not due to Bantu introgression are found in African pygmies. This too points to its African antiquity.

terryt said...

"The African component was found to be more similar to populations inhabiting the Levant rather than the Arabian Peninsula, but the principal route for the expansion out of Africa ∼60 kya remains unresolved".

Aren't the two statements more than a little contradictory? If the 'African component' (or even the 'non-African' component) is more similar to the Levant than to Arabia isn't that evidence enough that the route out of Africa was via the Levant? Especially when we consider that:

"Linkage-disequilibrium decay with genomic distance was less rapid in both the whole genome and the African component than in southern African samples, suggesting a less ancient history for Ethiopian populations".

That very much supports Dienekes suggestion:

"I have little doubt that my own theory about substantial back-migration of Eurasians into Africa will eventually win the day".

I'm sure it will win the day.

"the biggest elephant in the room is where are the CENTRAL CUSHITIC SPEAKERS?".

I agree that the mountains may preserve the more ancient elements of the Ethiopian population, however I also think it unlikely that such a highland-inhabiting population would have been the one that migrated out of Africa originally.

eurologist said...

If I understand this correctly, the light blue, represented by French, is a combination of Mediterranean, Western Asian, and Southwestern Asian. Is that a correct interpretation?

Also, the STRUCTURE results look like there indeed is a unique Ethiopian component, given the sources and K-levels as selected - namely the yellow SW Ethiopian (vs. the purple Nilotic). Interestingly, outside of Ethiopia and Somalisa this component almost plays no role, at all. Conversely, Ethiopians don't have the red West African component, at all, even though it shows up in North Africa, everywhere. Did North Africans receive that West African component in recent times, or did the the West African component at one point (pre-neolithic) extend much further North?

Also, the Sandawe are an eclectic bunch! I wonder if that would change at higher K-levels (i.e., there may be a Sandawe element in many of the other groups).

Lank said...

The "Afar" samples appear to be mislabeled Central Cushites. According to the supplemental data (Table S1), they are Xamtan speakers and "highland agriculturalists".

This would explain their great genetic similarity to the Amhara. The Cushitic influence in Ethiosemitic comes from Central Cushitic. Also, some groups of Central Cushites are known to have switched languages to Semitic recently.

Anonymous said...

I think the origins of Ethiopia is a very important subject, for many reasons that I rather not discus to stay on topic!

An important point is that the older populations of Ethiopia are not getting studied, focus is always on the:

# E1b1b V12 majority (Common amongst Cushitic speakers & a large protion of Cushites who switched to Semitic)

# J P58 out of the Semitic regions of Asia (largely confined to a large minority amongst the Semitic speakers)

This leads to fast conclusions (sometime comfortable conclusions to both natives & Afriphobes) that Ethiopians didn't have a population before the arrival of the Cushites & Semites.

E1b1b expansions out of Ethiopia
Y-DNA evidence shows that E1b1b xV12 is present in Ethiopia at a relatively lower frequency, but very high diversity. opposed to other regions f the world, this along with Ethiopia geographic location between E1b1b populations suggests that V12 Cushites were a Lithic back migration

Autosomal Ethiopian evidence

The Expansion of E1b1b from Ethiopia, can be traced by Autosomal impartial analysis, because most people hate to use Ethiopia as autosomal contributors (for both E1b1a West Africans & E1b1b East Meds), instead of assigning a double digit autosomal Ethiopian contribution to North Africans, they rather use a low single digit % by using West African populations! Who simply share the autosomes with Meds because of a common source in Ethiopia.

Finally African populations in Yemen are very diverse from multiple origins & in some regions are closer to subsaharans than the ones in Ethiopia, inland Gaezin tribes share linguistic & historic ties with some Tigrinya. Afar are closer to Mocha & Southern Tihama. Oromo & Somalis are closer to Tihami tribes.

The African Yemeni of Abyan have a higher percentage of Subsaharan autosomes & YDNA that are rare in Ethiopia due to Slavery. Hadramawt, Omani Africans are more mixed and show evidence of residual Paleolithic populations that can be associated with older out of Africa L mtDNA

Lebanese Autosomes are also important, once compared to Oromo Ethiopians (to avoid the % of Semitic autosomes amongst Semitic speakers), you will see a double digit value.