Final summary: North & East African AncestryDNA results 2013-2018

Without wanting to rehash things this blog post and following ones will provide a quick recap of my final AncestryDNA survey findings for continental Africans. These results were obtained during 2013-2018 but I had not fully processed all the data up till now. Continuing with my North & East African section. Which I first published on 20 November 2016 when I had only 58 North & East African AncestryDNA results available for my analysis (see this overview). As my survey has been ongoing I have managed to collect a sample group which has now more than doubled in sample size. Consisting of no less than 135 AncestryDNA results of  North & East African persons!  Follow the link below for detailed analysis & new screenshots:

Table 1 (click to enlarge)

Stats (n=135)


I discontinued this survey after Ancestry’s update in September 2018. Because in my opinion Ancestry’s new version of their ethnicity estimates regrettably has been a downgrade in regards to their African breakdown, overall speaking (see also this blogseries). Which is why I think these “old” results may still be useful and are not obsolete yet. I originally singled out two main implications for Afro-Diasporans which I still stand by.

  1. Careful follow-up research is required to substantiate any DNA results seemingly suggesting East African lineage.
  2. Careful follow-up research is required to substantiate any DNA results seemingly suggesting North African lineage.

Then again my intention is not only to look backwards but also forwards! In particular I believe that this dataset may provide a helpful baseline to compare with 23andme’s newly updated African breakdown (see this overview). Furthermore my findings may also serve as a benchmark for Ancestry’s upcoming update, which hopefully this time will lead to improvement! I find that currently 23andme is much better equipped to detect Northeast African & North African lineage. While also the distinction made for Southeastern African lineage on 23andme is particularly useful for Afro-Diasporans seeking to validate seemingly East African DNA test results.

From what I have seen (see this page) the new “Eastern Africa” category on Ancestry is much less predictive for actual East Africans. While interpretation for Afro-Diasporans is often tricky. When it comes to the new “Northern Africa” region on Ancestry the record is mixed. With interpretation again being tricky for Afro-descendants. More clarity may be obtained however when you critically assess your North and/or East African DNA matches, if you happen to have them. Taking into account shared segment size and the possibility of population matches and false positives, so that only genuine IBD matches are left over. Scroll down to sections 3 & 4 of this page for how I performed such a task for 50 Cape Verdeans and their 180 North African & 9 East African matches. 

Over all I would say that my African survey findings are pretty much reinforcing the main message I have been trying to get across from day 1: the labeling of ancestral categories in DNA testing is often not be taken too literally. However this does not mean you cannot derive a great deal of informational value from your socalled ethnicity estimates! Proper interpretation is a precondition. And it has been my experience that being aware of the DNA results of native Africans certainly helps in this regard. Even more so when also taking into account your African DNA matches, relevant historical context, (advanced) genetic genealogy and other means of follow-up research.

***(click to enlarge)

Compil NEA


9 thoughts on “Final summary: North & East African AncestryDNA results 2013-2018

  1. Fonte,

    I am very interested in the tests showing a very significant difference between the Hutu and Tutsi. Of course, I’d like to see more samples, but this along with other tests do seem to show the Hutu as being more Bantu than the Tutsi, genetically.

    Also, I’m surprised by the Egyptian results. It was always my understanding that the Egyptians, genetically, were much more indigenous to their region, so I expected them to show unusually high levels of “North Africa.” Your survey shows exactly the opposite. But maybe this is evidence that “North Africa” is measuring Berbers, who I don’t think were particularly genetically close with the ole Egyptians. Maybe it would be good for Ancestry and the rest to build up an Egyptian sample; I feel pretty confident they’d differ quite a bit from “West Asia” if the sample were bigger. Just kind of superficially, your average Egyptian looks noticeably different than your average Arab to me, even though the Egyptians obviously have had a significant genetic inflow from Arab invasions.

    Liked by 1 person

    • If you scroll down the main page where I have posted all the screenshots you will find several links about Tutsi & Hutu genetics. It is indeed a very interesting topic although also quite sensitive.

      About the high “Middle East” scores for Egyptians, they are not per se traceable to modernday ethnic Arabs. Or at least not completely so. In fact it could also be “indigenous” lineage when we’re talking about blood lines which have been around in the area for more than 2000 years. Again if you scroll down the main page you will also find two Coptic results and they score similar amounts of “Middle East” and “Africa North”. But interestingly much smaller amounts of other African regions.

      Same goes for arabized Sudanese whose “Middle East” scores might partially have been inherited by way of their Nubian ancestors who might have been genetically similar to Ethiopians and Somali (therefore also incl. a very ancient West Eurasian component). For Tunisians for example at least a part of their “Middle East” scores might also have been inherited by way of Phoenician ancestors. Also for other North Africans “Africa North” has proven not to be a perfect measure of their Berber lineage and their “Middle East” scores will not per se be from modernday Arab lineage. Although clearly it will be correlated. Even when a bit confusing also with the overlap with southern European regions I find that the gradient for “Africa North” (maximum in Morocco and least amount in Egypt) does make sense. Given that Morocco is known to have the largest self-identified and also still Berber speaking population in North Africa.


      • It just doesn’t make sense that Egypt, which had a much larger indigenous population that anywhere else in the Maghreb prior to the Arab migration/invasion is more “Middle East” than any of its western neighbors. What the regions are measuring needs to be flipped to make sense. Middle Easterners left a far larger genetic footprint on the areas to the west of Egypt than they did in Egypt.


        • Several DNA studies have been published in the last couple of years on ancient North African genetics. Which should shed more light on this issue. Like I said “Middle East” is most likely also including pre-Arab conquest genetic affiliations with West-Eurasia.

          It is a fascinating topic no doubt. But frankly my interest goes more towards more recent time periods. For example I find it very useful to see how minor “Senegal” scores peak among Moroccans while for Egyptians it is mainly “Southeastern Bantu” which acts as a secondary African region.


          • Not to belabor this too much, but AncestryDNA only goes back something like 500 years. The Arab conquest of Egypt happened in the 600’s when Egypt was under Roman rule. And in any case, like the most recent poster said, there seems to have been relatively minor – relative to the areas to the west with lower population densities – genetic incursion into the indigenuous population.

            So, again, I have no idea what Ancestry is measuring with “West Asia/Middle East,” because Egypt is not that. Egypt is the least “West Asian” of the Maghrebi nations.


            • Haha i know discussions about Egyptian genetics can get contentious 😉 But I am actually in agreement with both of you that ethnic Arab gene flow into the Egyptian genepool may have been limited. I have said as much also on the main page where I have posted all the screenshots. Which to repeat myself also includes two Coptic results who I suppose may be assumed to represent pre-Arab conquest Egyptian bloodlines the best (generally speaking). Again new research, (also utilizing ancient DNA samples) will be much better equipped to deal with this issue than AncestryDNA’s basic framework.

              The big question indeed is how to make the distinction between historical geneflow and genetic similarity due to (very) ancient shared origins from West Eurasia. This issue also concerns other North & East Africans! For Horners it is quite clear for example that their socalled “Middle East” scores are generally speaking not referring to any historical geneflow but rather to be traced back to very ancient admixture events. Same goes for Cushitic influenced Kenyans and Tutsi. However it gets more complex for Sudanese Arabs, Swahili and also in particular Tunisians i suppose, for whom multiple ancestral scenario’s might be valid all at the same time.

              So I ‘m not sure why you said that “AncestryDNA only goes back something like 500 years”. It is very apparent not only from my African survey results but in fact also others parts of the world that often also (very) ancient admixture/shared origins rather than any recent lineage is suggested by these AncestryDNA estimates. And in this way many of these regional scores might perhaps best be seen as genetic echoes from a (very) distant past. Beyond family tales, historical knowledge or even ethnic origin traditions.

              As i keep on saying it is best not to overfocus on the labeling of Ancestry’s regions. I highly suspect that socalled “Africa North” is based on Mozabite samples from Algeria. And therefore primarily acting as a (imperfect) proxy for detecting Berber (like) DNA. Which is why again i find it only natural that this region should peak among Moroccans. I have never really looked into the question which samples Ancestry used for so-called “Middle East” . However it seems obvious to me that it is supposed to act as some kind of proxy for detecting West Asian-like DNA, which again might be due to very ancient migrations or also more recent historical geneflow.

              What i personally find an intriguing research question is how to determine any genuine Arab connection. In the Youtube section I posted a fascinating video of a Sudanese Arab woman who talks about this dilemma of Arab identity and DNA test results. For many it will be tempting to criticize her stance. However I find that anyone should be entitled to research any lineage they are personally interested in. I highly suspect that her “Middle East” scores are not entirely to be traced back to recent Arab lineage, perhaps not even for the greater part. Still given her own knowledge of her family history (mentioned also in the video description) i think some degree of Arab lineage is very likely. Even when of course family lore might get embellished for prestige reasons and also given the very real possibility of non-parental events and unrecorded geneflow through female lines.

              Admixture analysis (as performed by commercial DNA testing companies) does not seem really equipped yet to disentangle historical geneflow from genetic similarity due to ancient shared origins. Although it would be interesting to see her updated 23andme results (see this page for a couple of Sudanese results). I imagine her best bet on validating her Arab lineage would be through DNA matches. But if she wants to do a thorough job she should be focusing on the bigger matches (>10 cM). To leave out the possibility of population matches and false positives. I also imagine that if she got one of her direct male relatives to test for his Y haplogroup she might also find some confirmation, at least along the direct male family line.


    • There is hardly any gene flow from Arabians to Egyptians , only difference between modern and ancient ones is increased African ancestry on modern Egyptians (i dont remember the scale but basically from 10 % to 20 % ) , Arabians and and Ancient Egyptians are both Afro Asiatic speakers so obviously they have a lot in common ,and Egypt is next door to Levant , So Berber speaking ones got mostly West African dna and also Iberian contribution due geography that Egyptians lack , Egyptian African is East African due geography , but they all share Afro Asiatic roots, the same thing is also shared North Sudanese and Horners but lesser extent , they got more African dna.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s