Friday, September 30, 2016

Comments on the new Horn African and north Sudanese samples

Well, as you can see here, we have 13 new samples from southern Somalia, 10 from Ethiopia of -West-Central Oromo origins and about 17 Sudanese-Arab+Nubian samples from northern Sudan.

Some interesting things to note would be:

  • The Somalis seem rather uniform in terms of their "African" & "West Eurasian" ancestry levels once you set aside the 3 outliers who look to have 30-50% West-Central African-related admixture of probably recent origins from Somali Bantus. [note] This is interesting because these same ~10 remaining Somalis display an odd substructure in the PCA where 3 samples, alongside at least two Oromos, pull quite northwards as though they're somehow slightly distinct from the other Horn Africans present in the PCA, even though all of the Somalis, except for those three outliers, seem more or less the same in regards to their admixture levels.

  • The West-Central Oromos seem to be about as African & West Eurasian as the Somalis and, if I recall correctly, these types of Oromos tend to vary between being 40-50% West Eurasian from what I remember so the Somalis in question look to be relatively standard in terms of admixture levels for pulling about as eastwards toward West Eurasians as these Oromos. Some of these Oromos are more West Eurasian-shifted than the Somali samples which would make sense as well.

  • The Sudanese Arabs and Nubians seem mostly more West Eurasian-shifted than the Somali and Oromo samples which, I suppose, could mean most of them are at Habesha~Agaw levels of African and West Eurasian ancestry while some, or even a decent number, possibly also surpass those admixture levels. They also definitely display signs of notable heterogeneity in terms of admixture levels which correlates well with prior data owed to Dobon et al.

One thing the way some of those Somalis cluster reminds me of is this, actually:

That PCA above, as discussed here, basically shows us that some of Lazaridis et al. 2013-2014's Somalis from Garissa seem slightly "Ethiopian-shifted" in a manner that might imply admixture from people like Borana Oromos, they're also very slightly less West Eurasian (on average) than the Somalis we so far have from Ethiopia and Somalia.

Now, this new study's Somali samples, originally from southern Somalia, are classified as "Sedentarian" which looks to impart that they're sedentary farmers rather than semi-nomadic pastoralists. Perhaps they're of the Rahanweyn clan (though they don't have to be) which has long had sedentary farmers within its clan-structure in the south? Some might notice that I once entertained the notion that some Rahanweyn clan members might have notable Ari-like admixture the way various Ethiopian populations do (see here). [note]

Well, perhaps the reason the majority of these Somalis are clustering so close to the West-Central Oromos is because they carry admixture from Oromo-related/like people and the ~3 who don't cluster that way and pull very upwards actually lack such admixture? 

But then it doesn't explain why all of the samples seem to be roughly the same in terms of West Eurasian and African ancestry levels (they're basically homogeneous in this case). Also, why do at least two West-Central Oromos actually pull north as well? And why do these Somalis seem so close to West-Central Oromos in terms of admixture levels which implies all of them have relatively typical Somali admixture levels? [note]

These issues don't make much sense if some of them are part Oromo (or even Borana Oromo-like) while others aren't.

So I suppose it may not be worth speculating too much with this study's PCA and ADMIXTURE run until we have these samples out in the open. There are also other odd things about these results like the Oromos showing tiny hints of West-Central African-related admixture when the non-outlier Somalis show none and that honestly doesn't make much sense to me so I'll have to just see these samples for myself when the time comes. 

I wouldn't be sure of too much until we can run these samples through other analyses but these are my quick two cents for now



  1. Great find Awale!

    - I was really hoping you would comment on the fresh Daza and Kanembu results, their possible interactions with one another, with North Africans and what implications this has on Y-DNA R-V88. It's awesome to get some clarification on who the people are in northern Chad

    - Surprised you didn't mention the Samburu results that seems to indicate a strong relation of the primary ancestors with Cushitic speakers like the Somali and Rendille

    - Despite their Nilo-Saharan(?) language classification, the Songhai seem to be genetically similar to Burkinabe groups and West-Central NK Africans in general

    1. Oh there's another post coming for those other samples. This one was Horn and Sudanese-Arab & Nubian centric. Apologies. Those other new samples are pretty intriguing too. And yes, Samburus look very Horn African-shifted. One of the first things I noticed. They cluster real close to the Somalis and Oromos.

  2. There was unsourced information on the the various Toubou clans having high R-V88, E-M78 and T-M70. No study on their uniparental markers has been done AFAIK, but those results would be rather odd for a solidly Nilo-Saharan population.

    1. Oh, I just noticed this comment. Would've been an interesting comment to be shared here:

      T-M70 is definitely surprising/interesting given that it's found in Horners and Cushitic speaker admixed Southeast Africans. The R-V88 implies some Chadic speaking origins to me too. But E-M78... Don't know what to make of that. These guys' West Eurasian admixture looks strangely Northwest African mediated based on auDNA so that's a bit confusing given these Y-DNA markers.

  3. @pgbk87

    From the Chad Tribes FTDNA project I guess.

    If I'm not mistaken who is and isn't a Toubou, it looks like there's 7/18 R (presumably V88), 4/18 E-L19(xM81), 1/18 E-V12(xV32), 1/18 A-M13, 1/18 E-M2, 1/18 T1a-M70, and 3/18 unidentified E-M35 which look by the haplotype like maybe 1 E-L19 and 2 E-V12.

    So yeah, high V88 and M78 but little A. They're Saharans though, some Kanuri groups have high R1b.

  4. Awale I wanted to find out about West Central African(Niger Congo) admixture in Northen Sudanese Arabs. I have watched several ancestry dot com videos of Northern Sudanese Arabs and they had like 5-10% Niger Congo admixture. Sudan is a much more interesting place than I first thought because All of the language families in Africa are spoken in Sudan......

  5. Is the west euroasian in Sudenese Arabs ancient or recent?!

  6. can you look into eastern oromo dna from dire dawa hararghe. Some say they are assimilated somalis from dir tribe.