PDA

View Full Version : AncestryDNA breakdown of SSA for Africans



Don Felipe
09-24-2014, 09:47 PM
Follow-up to this previous thread (http://www.anthrogenica.com/showthread.php?2419-AncestryDNA-breakdown-of-SSA-for-African-Americans-amp-other-Afrodescendants). But now i will be posting only results for Africans with confirmed ethnic background within Africa. Unlike for New Worlders these results can therefore be verified with known genealogy. Should be insightful on how reliable/predictive the various AncestryDNA categories can be and also how they might be interpreted.

Not many fully Africans yet on AncestryDNA, possibly less than on 23andme, but here's two to start with



IVORY COAST (7/8 Akan/Malinké & 1/8 Krio from Sierra Leone/Gambia)

Given that the Euro breakdown on Ancestry still gets some criticism i was very curious to know how the African breakdown would stand up for someone who can actually verify it. But they didn't do bad at all for this one. Of course it's not exact science (and perhaps never can be) but I'm just blown away by how the %'s seem to be pretty much in line even with this person's 1 non-Ivorian great grandparent who is Krio (http://en.wikipedia.org/wiki/Sierra_Leone_Creole_people), the descendants of ex-slaves from the US, marroons from Jamaica as well as liberated slaves from all parts of West/Central Africa being settled in Freetown. It 's known many of them were Yoruba so some minor Nigeria would fit well. The Trace regions might also be explained by the very divers Krio background but it could also just be noise, as the confidence level is lower than for the main regions.





http://i989.photobucket.com/albums/af18/oditous2/RangeTraceRegionsIvCoast_zps596d46e4.jpg





http://i989.photobucket.com/albums/af18/oditous2/IvCoast_zps99264001.jpg



Judging from this 1 single result the Ivory Coast-Ghana category seems to be VERY solid at least for someone of Ivorian descent :D. No one from the New Worlders got such a high score for a single category (relative to total African). They put down 83% but according to the range given it could be as high as 94%. According to this screenshot Ancestry performs 40 separate estimates and 83% would be the average of these 40 runs with 94% not even per se being the highest estimate. So that's a pretty tight range as well.




http://i989.photobucket.com/albums/af18/oditous2/Probablerange_zps756aef50.png~original





83% is also pretty much the same as the 86% scored by the "typical native" from Ivory Coast/Ghana according to AncestryDNA. This 86% being the median result of their 115 samples.


http://i989.photobucket.com/albums/af18/oditous2/IvCoastGhana_zpsf307345f.jpg~original




Perhaps for some Ghanaians there might be a pull towards Benin/Togo as Ancestry themselves has shown there's overlap between their samples for these 2 categories. But not so for this Ivorian. I suspect mostly Akan samples were used for the Ivory Coast/Ghana category. But i've not seen it confirmed yet. Interestingly this person's partial Malinké/Mandingo ancestry is not getting picked up in a separate category, like Senegal or the more weakly defined Mali category. So who knows they might actually also have used some northern Ivorians of Malinké background as samples. Still many of the Ivorian/Ghanaian samples used by Ancestry do score some noticeable Benin/Togo or Mali.



http://i989.photobucket.com/albums/af18/oditous2/IvCoastGhanaotherregions_zpsfd3ab712.jpg~original




In this screenshot below you can see how much these additional components would be on average. They should have picked less tricky colours lol, but from what i'm seeing the additional regions for Ivorians/Ghanaians would be about 10% Benin/Togo and somewhat less than 5% Mali while the other regions are at noise level.




http://i989.photobucket.com/albums/af18/oditous2/AncestryAdmixture_zpsd597dbbb.png~original

Don Felipe
09-24-2014, 10:29 PM
NIGERIA (fully Igbo afaik)

Very fascinating these results and also insightful given the prominent Igbo and otherwise Biafran (southeast Nigerian) derived ancestry for many AA's and Anglo Caribbeans according to historical sources. His substantial Cameroon/Congo % was to be expected. But his also considerable Benin/Togo % might put the scores of the New Worlders we've seen sofar in a different light :biggrin1:

The results might appear to be off at first look for those not accustomed to how AncestryDNA works. So i will also provide some background info to make more sense of the outcome for this individual. Most important thing to keep in mind is DON'T take the labeling of the categories too literally! They merely signal close genetic similarity to samples from the countries/regions after which the categories get named, they are NOT per se indicative of actual recent origins from those countries/regions!

Most African countries have dozens of ethnic groups and interethnic mixing has been going on since prehistorical times. So ancestral clusters won't respect national borders drawn by colonial powers! Personally I have a feeling they correlate much more with the big language subgroups within the Niger/Congo language family but even so people speaking the same or related languages might still have divergent DNA profiles because we all have unique genealogies which transcend ethnicity if you go back in (pre)history far enough and consider such things as language shifts, absorption of marginalized aboriginal populations by newcomers but also recent migration etc. etc.


Enough with the disclaimers for now, here's the results:


http://i989.photobucket.com/albums/af18/oditous2/IgboCM_zps42c3b86c.png~original






Here's the Nigeria category description by Ancestry showcasing the individual variation which includes scoring %'s of neigbouring regions and also how Nigeria is relatively less predictive than all the other categories except Mali. Not very surprising given that Nigeria is the most diverse and populous country of Africa. Notice that among the 67 samples used by Ancestry there's one Nigerian with the exact same "Nigerian" score as this Igbo. But a "typical native" of Nigeria would have 69% (=median of the sample group). It's too early to jump to any conclusions yet but I have a feeling that other Igbo's will also tend to score less "Nigeria" and more "Cameroon/Congo" than the average of Ancestry's samples which probably also includes some Yoruba's while the Benin/Togo might vary more wildly. From the last screenshot in my OP i'm seeing Nigeria's additional components on average being first Benin/Togo then Cameroon/Congo (yellow), the other regions seem minimal but third comes Mali (light beige) and fourth SE Bantu.





http://i989.photobucket.com/albums/af18/oditous2/Nigeria_zps73583a88.jpg~original


http://i989.photobucket.com/albums/af18/oditous2/Nigeriaotherregions_zps6c6158fe.jpg~original








I anticipated for Igbo's to score some additional Cameroon/Congo as i had already seen it on persons of half Igbo, half AA background. It makes sense given the geographical proximity and the Bantu expansion which is supposed to have started in the border area of Cameroon/Nigeria. Mali, Senegal and SE Bantu being tiny Trace regions at noise level is also to be expected. But the 17% Benin/Togo (and 8% Ivory Coast /Ghana) is more surprising given that the Igbo's are located in southeastern Nigeria, bordering Cameroon, while it is the Yoruba's from southwestern Nigeria who are bordering Benin.


It's tempting to correlate this finding with the commonly held belief that the Gbe speakers (http://en.wikipedia.org/wiki/Gbe_languages)(Fon, Ewe etc.) of Benin/Togo were migrants from somewhere in southern Nigeria around the same time (1200's-1400's) the expansion took place of the famous Bini/Edo empire (http://en.wikipedia.org/wiki/Benin_Empire)(confusingly also called Benin, but centered in southern Nigeria!). The exact extent of this empire is not known but i've read accounts that it stretched all the way west beyond Lagos into modernday Benin and it may also have caused population displacement in the east across the Niger delta into Igbo land. All this movement of peoples might possibly be the cause why an Igbo may have DNA markers in common with people from Benin/Togo and even Ghana (possibly Ewe mixed samples?). It would in fact be shared DNA that originated in Nigeria and still exists there but nowadays is more frequently seen in Benin/Togo.


An even more fascinating hypothesis but more speculative is that aside from historical migrations there might also have been more ancient prehistorical movements of people between Nigeria and Benin/Togo/Ghana which explain why an eastern Nigerian would have DNA markers in common with people from countries to the west of Nigeria. I'm talking about the flow of people that would have been involved with the spread of agriculture within West Africa and more specifically the Yam cultivation. This map shows it well. Mali & Nigeria can be considered the two major ancient centers of civilization within West African history, both of them located nearby the extremes of the Niger river. Both with distinct agricultural systems, different language groups and also separate influence/expansion zones.




http://i989.photobucket.com/albums/af18/oditous2/WesternAfrica-Spreadofagricultureintorainforestzone.jpg

Don Felipe
11-07-2014, 05:24 PM
SOUTH AFRICA (Coloured)

Very interesting because aside from genuine Khoisan and Pygmy's no other population will probably get such high "Hunter Gatherer" %'s as the South African Coloureds.

http://i989.photobucket.com/albums/af18/oditous2/ZAColoureda_zpsdf4f854c.jpg~original

The African breakdown is pretty much equally divided between SE Bantu and these socalled "South Central Hunter Gather's". It's actually one of the most straighforward ancestral clusters, strictly based on Khoisan and Pygmy samples.


The Africa South-Central Hunter-Gatherers region is made up of ancient hunter-gatherer and pastoral groups who, though small in number and physical height, are considered the wellspring of human populations around the world. Increasingly, southern African Khoe-San groups and Central Africa’s Mbuti and Baka (Pygmy) groups are drawing the attention of scholars and researchers for their genetic diversity, ancient origins and unique cultural traditions.

The samples they used were not completely "pure" though


Individuals from the Africa South-Central Hunter-Gatherers region are admixed, which means that when creating genetic ethnicity estimates for people native to this area, we frequently see similarities to DNA profiles from other nearby regions. We’ve found that approximately 86% of the typical South-Central Hunter-Gatherer’s DNA comes from this region.


For this region we see a substantial range: for most natives to this area, between 74% and 100% of their DNA looks similar to the profile. However, for others, as little as 62% of their DNA comes from this region. The other region most commonly found is the neighboring Cameroon/Congo region. About 54% of people from the Africa South-Central Hunter-Gatherers region have at least some DNA from Cameroon/Congo


This AncestryDNA category is also the equivalent of the socalled "Central & Southern African' category on 23andme, which is kind of a misnomer because it excludes Bantu speaking people from this area :doh:


Central & South African

Central Africa extends from the Central African Republic at its north to Angola at its south. Southern Africa encompasses Namibia, South Africa, Botswana and Zimbabwe. While the majority of its population is now composed of Bantu peoples, Central Africa is also home to many Pygmy populations. Southern Africa was first peopled by Pygmies, San and Khoisan. These hunter-gatherer populations still live in this region today.

Population Source Sample Size
Biaka Pygmies HGDP 32
Mbuti Pygmies HGDP 15

San HGDP 6


Most New Worlders will score only very small %'s for it, typically below 5%. From what i've seen Puerto Ricans and Brazilians score highest relatively speaking (as % of total African). In most cases it will be an ancestral component inherited from Bantu speakers brought over to the Americas who already carried these markers in their own genome. More specifically it will be an indication of Congolese/Angolan ancestry i suppose.

However obviously not so for this South African Coloured who are known to have very significant Khoisan ancestry B) The other half is made up of SE Bantu which is a very wideranging category as can be seen from the map below.


http://i989.photobucket.com/albums/af18/oditous2/SEBantumap_zps9901fd6b.jpg~original



It seems to be one of the less stable clusters aside from Mali & Nigeria. Some of their samples also scoring %'s for North Africa and the Middle East! In fact in absence of an Northeast African category it also seems to capture Cushitic/Nilotic ancestry. i've seen some Horner and also Coptic results whose SSA part is being shown as SE Bantu :biggrin1:


http://i989.photobucket.com/albums/af18/oditous2/SEBantucluster_zpsfc7041fe.png~original



For this South African Coloured person the SE Bantu would be genuine alright though. However it's most likely not South African Bantu (Xhosa, Zulu etc.) but rather Angolan, Mozambican or even from Madagascar, as these were the areas where the Dutch of the Cape Colony mainly got their slaves from (aside from Asia).

Don Felipe
01-09-2015, 01:30 AM
NIGERIA (Igbo)


Don't have the screenshot but the results can be seen via this video by this Nigerian-American comedian named 'On1'. He says he's fully Igbo (at least going back 2 generations ;)) and also he states his AncestryDNA breakdown as follows:

77% Nigeria
13% Benin/Togo
5% Cameroon/Congo
4% Ghana/Ivory coast
1% Mali


https://www.youtube.com/watch?v=9eqWyf26naw




It's too early to jump to any conclusions yet but I have a feeling that other Igbo's will also tend to score less "Nigeria" and more "Cameroon/Congo" than the average of Ancestry's samples which probably also includes some Yoruba's while the Benin/Togo might vary more wildly.

Too early indeed :biggrin1: This guy actually has a higher than average "Nigeria" score while his Cameroon/Congo is quite minor compared with the first result i showed :doh: ^Perhaps i shouldn't be speculating anymore...but could it be that this guy is from a region within Igboland that's more closely related to other Nigerian ethnic groups than the previous one?

I have two new Yoruba results as well i will post later on. They both score lower "Nigeria" and (much) higher Benin/Togo than this Igbo.

Don Felipe
01-09-2015, 09:59 PM
NIGERIA (Yoruba)

Here's the results of a Nigerian Yoruba for contrast.

http://i989.photobucket.com/albums/af18/oditous2/Yoruba_zpsd21e78b8.jpg~original


Her Nigerian % is close to what a "typical native" would score according to the samples from Ancestry.com (69%). So that should pretty much be average. Also as expected the Benin/Togo % is quite pronounced, the Yoruba are afterall also to be found in Benin itself! Other ethnic groups from Benin, like the Fon, are also likely to be related genetically speaking as is usually the case for neighbouring ethnicities all over the world. I've actually seen the results of another Yoruba person and he scores even higher for "Benin/Togo": 52%! Again confirmation that these socalled "Benin/Togo" ancestral markers seen among New Worlders might just as well have been inherited via a Nigerian ancestor. When we get results in of Ghanaian persons they could also be scoring significant "Benin/Togo" %'s depending on their ethnic background.

Everything else is at minor Trace Region level so could as well be simply "noise" or just generic DNA that's difficult to classify. Still interesting that Cameroon/Congo is mentioned only as a very reduced %, the Cameroon/Congo category could be a tell-tale sign i suppose to distinguish the more archetypical Yoruba from the averaged out Igbo, however it's important to stress that even within ethnic groups you will find MUCH individual variation.


Some quotes & scans from the Ancestry website just to provide some more info on the "Benin/Togo" category that's been showing up as main region somewhat surprisingly for especially many African Americans and Jamaicans sofar. Again it might signify different ethnical roots depending on your nationality: for Haitians and Brazilians this category is very likely to be derived from Fon, Adja or Yoruba ancestry, for AA's and Anglo-Caribbeans it 's perhaps more likely indicative of Ewe or Ga ancestry from Ghana/Togo or else also partially southern Nigerian. However everything's possible in individual cases ;)



Benin’s largest ethnic group is the Fon (39%), followed by the Adja (15%), Yoruba (12%) and Bariba (9%). Togo’s largest ethnic groups are the Ewe (21%), Kabye (12%), Mina (3.2%) and Kotokoli (3.2%). Benin has more ethnic ties to its neighbor Nigeria; Togo has more links to Ghana. These ethnic ties are the result of long-standing kingdoms that flourished before European colonists created new borders.





Many people in Togo and Benin speak one of about 20 related Gbe languages. Linguistic evidence indicates that most of the Gbe people came from the east in several migrations between the 10th and 15th centuries. The Gbe were pushed westward during a series of wars with the Yoruba people of Nigeria, then settled in Tado on the Mono River (in present-day Togo).



http://i989.photobucket.com/albums/af18/oditous2/Benindescrip_zps4bdd0129.jpg~original




Even though it's also found in neighbouring areas it's still quite predictive for genuine Beninese ancestry. The thing is that their DNA happens to be shared by more people than just themselves :biggrin1: Still that doesn't mean these DNA markers originated in Benin ;)





http://i989.photobucket.com/albums/af18/oditous2/Beninsamples_zps1579391b.jpg~original

Don Felipe
01-09-2015, 10:46 PM
NIGERIA (Igbo)

Don't have the screenshot yet

Here it is

http://i989.photobucket.com/albums/af18/oditous2/On1_zps57890a78.jpg~original

Don Felipe
01-20-2015, 06:01 PM
NORTHEAST AFRICA/HORNER (Eritrea, Somali?)

Two Horner results, basically showing a combination of SE Bantu, North African and Middle East. No doubt SE Bantu became the default SSA category for Horners because in absence of a category based on Nilotic (Maasai) samples it will be the least distant SSA cluster for them on Ancestry. It won't be actual Bantu ancestry of course ;). The Trace regions have lower confidence level but still interesting what's being shown for the first result. Mali & Senegal are most likely the least distant SSA clusters after SE Bantu while the 3% Hunter-Gatherers might be hinting at the very ancient Horner connection with the Khoi-San??


Somali or Oromo?

http://i989.photobucket.com/albums/af18/oditous2/Som_zps480d64e2.jpg~original


Eritrean

http://i989.photobucket.com/albums/af18/oditous2/Eritrea_zps84309106.jpg~original



There's also North African being shown, which seems to be the most predictive/reliable African category on Ancestry right now together with Senegal, even though it's only based on 26 samples:

http://i989.photobucket.com/albums/af18/oditous2/NorthAfrica_zps69837338.jpg~original


Interestingly the North African scores are shown as being minor to Middle Eastern. On 23andme it's the opposite. With all Horner results i've seen only coming up as combinations of East & North African. Most likely having to do with the Gulf Arab and Levantine samples being used to define the North African cluster on 23andme :doh:


North African
Over the last several millennia the Sahara has been a significant geographic barrier for the people of Africa. Although trade routes between the northern and southern parts of the continent have long existed, northern Africa has been strongly influenced by people of the Middle East and southern Europe.

Population Source Sample Size
Palestinian HGDP 51
Bedouin HGDP 48
Mozabite HGDP 30
Egypt 23andMe 28
Palestine 23andMe 28
Morocco 23andMe 19
Algeria 23andMe 14
Saudi Arabia 23andMe 8
Tunisia 23andMe 7
Jordan 23andMe 5
Yemen 23andMe 5
Kuwait 23andMe 3
United Arab Emirates 23andMe 2
Bahrain 23andMe 1



Here's the samples they use for East Africa on 23andme. Ancestry should also have a similar category but they should call it Northeast Africa or Nilotic rather than just East African, which tends to cause confusion with the Bantu speaking countries of Kenya, Tanzania etc. Perhaps no surprise it peaks with Somali's :biggrin1:




East African
Probably the cradle of modern humans, Eastern Africa spans from Ethiopia to Mozambique and Madagascar. Eastern Africa harbors an incredible genetic diversity, with Ethiopia alone containing dozens of ethnic groups. Home of the famous Maasai people, the southern part of Eastern Africa was first inhabited by Khoisan hunter-gatherers, until they were replaced during the Bantu expansion.

Population Source Sample Size
Maasai HapMap3 87
Ethiopia 23andMe 14
Somalia 23andMe 12
Eritrea 23andMe 3


Somali


http://i989.photobucket.com/albums/af18/oditous2/Somali_zpsfd20258b.jpg~original


Ethiopian 1

http://i989.photobucket.com/albums/af18/oditous2/Ethio_zpsd5cdb02a.jpg~original


Ethiopian 2

http://i989.photobucket.com/albums/af18/oditous2/Ethio_zpsd5cdb02a.jpg~original

Eritrean

http://i989.photobucket.com/albums/af18/oditous2/Eri_zps5cdc901b.jpg~original

Don Felipe
01-20-2015, 07:22 PM
EAST AFRICA??? (Kenya, Tanzania, Ruanda/Burundi???)

I would like to emphasize from the start i do not have any certainty about the ethnic/regional origins behind these results for this biracial person. So posting these results in this thread is based on (informed) speculation on my part ;) However i think combining the unusually high SE Bantu score on Ancestry's test and the 23andme composition of the same person showing a minor but hefty amount of East African (hardly ever present among New Worlders except for miniscule amounts below 1%) it seems to be highly suggestive of recent East African background. Not Northeast African because there's also much West African showing up and no North African.

Here's the AncestryDNA breakdown, it's the highest absolute level of SE Bantu i've seen and also relative to total African it's among the highest, 55% of total African (27/49). Dominicans and Puerto Ricans can also score high for this category from what i've seen but their total SSA was lower than 35% in those cases. I highly suspect black or pardo Brazilians could also score comparable levels of SE Bantu, but i have yet to see their results. Otherwise it's only genuine Africans who've scored this high for SE Bantu. Already posted in this thread, the coloured person from South Africa and the Horners. So like the name suggests and also the map (http://i989.photobucket.com/albums/af18/oditous2/SEBantumap_zps9901fd6b.jpg~original) is showing the Southeast Bantu category can signal very wideranging ancestral origins for different people, much depending on their nationality i suppose.

If this person's African background is indeed from either Kenya, Tanzania or Ruanda/Burundi, it's interesting to note that the minor Cameroon/Congo and Nigeria being shown possibly might hint at shared ancestry between these countries having to with the Bantu migrations originating in southeast Nigeria and moving via Cameroon/Congo to East Africa.




http://i989.photobucket.com/albums/af18/oditous2/KenTZ_zps34d26637.jpg~original


23andme for the same person, notice 10,1% East African!

http://i989.photobucket.com/albums/af18/oditous2/Kentaz_zps8189426e.jpg~original



A couple of East African 23andme results to show how the East African category is coming up for them, noticeable for all but very varied in between about 10% and 90%, no doubt because of different ethnic backgrounds. The "West African" %'s can be explained again by the Bantu migration but also because of 23andme's choice to include Kenyan (Luhya) & Bantu samples for this category, despite them being East African ...:noidea: AncestryDNA has shown that subregional resolution, incl. a separate Bantu category is in fact very much possible.



West African
Expanding from Senegal to Nigeria, West Africa composes about a fifth of the African continent. West African have a long shared history, and was united by large empires such as the Ghana Empire, dating as far back as the eighth century AD. Other empires later succeeded, making Western Africa a region of strong common heritage.

Population Source Sample Size
Sierra Leone 23andMe 173
Luhya 1000 Genomes 97
Yoruba 1000 Genomes 88
Yoruba HGDP 24
Mandenka HGDP 24
Bantu HGDP 20
Nigeria 23andMe 11
Ghana 23andMe 11
Ivory Coast 23andMe 2
Cameroon 23andMe 1
Liberia 23andMe 1



Kenyan 1

http://i989.photobucket.com/albums/af18/oditous2/KenyaAO_zps394986dc.jpg~original


Kenyan 2

http://i989.photobucket.com/albums/af18/oditous2/KenyaRB_zpsb0947c32.jpg~original


Kenyan 3


http://i989.photobucket.com/albums/af18/oditous2/KenyaKTM_zps38732ee8.jpg~original


Tanzanian 1

http://i989.photobucket.com/albums/af18/oditous2/Tanzania2_zps03c4a94e.jpg~original


Tanzanian 2


http://i989.photobucket.com/albums/af18/oditous2/TanzaniaSM_zps00c64c7b.jpg~original


Tanzanian 3


http://i989.photobucket.com/albums/af18/oditous2/TZ1_zps386b222d.jpg~original


Ruandan 1

http://i989.photobucket.com/albums/af18/oditous2/Ruandas_zps3f9d6cc0.jpg~original


Ruandan 2


http://i989.photobucket.com/albums/af18/oditous2/Ruanda_zps98f72fcc.jpg~original


Ruandan 3

http://i989.photobucket.com/albums/af18/oditous2/Ruandai_zps52eba25e.jpg~original

RoseWater
07-14-2016, 11:48 PM
Very cool thread! Thanks for sharing.

rocky
08-01-2016, 10:27 PM
Very Interesting thread, thanks for sharing bro :)

Don Felipe
11-20-2016, 01:54 AM
the Cameroon/Congo category could be a tell-tale sign i suppose to distinguish the more archetypical Yoruba from the averaged out Igbo, however it's important to stress that even within ethnic groups you will find MUCH individual variation.

More sampling needs to be done however sofar this assumption seems to be valid. More discussion and background information for below findings (based on 15 Nigerian AncestryDNA results) via this link:




Nigerian AncestryDNA results (https://tracingafricanroots.wordpress.com/ancestrydna/african-results/nigerian-results/)




https://tracingafricanroots.files.wordpress.com/2016/08/naija-ethnic-diversity.jpg





***





https://tracingafricanroots.files.wordpress.com/2016/09/naija-igbo-edo-yoruba-n15.jpg

Don Felipe
11-20-2016, 02:06 AM
Two Horner results, basically showing a combination of SE Bantu, North African and Middle East. No doubt SE Bantu became the default SSA category for Horners because in absence of a category based on Nilotic (Maasai) samples it will be the least distant SSA cluster for them on Ancestry. It won't be actual Bantu ancestry of course ;).

This still rings true.





There's also North African being shown, which seems to be the most predictive/reliable African category on Ancestry right now together with Senegal, even though it's only based on 26 samples

However i'm not too sure anymore about the predictiveness of the socalled "Africa North" region on AncestryDNA. It seems to be heavily conflated with the Iberian region. When reported it could still be a valid finding but in many cases it might also just signal VERY ancient genetic similarity with North Africa.


More discussion and background information for below findings via this link:




North & East African AncestryDNA results (https://tracingafricanroots.wordpress.com/ancestrydna/african-results/north-east-african-results/)



https://tracingafricanroots.files.wordpress.com/2016/11/compilnea.jpg





***



https://tracingafricanroots.files.wordpress.com/2016/11/stats-ken3.jpg

Tsakhur
11-22-2016, 04:14 AM
Do you have any results of Senegambians, Malians, Guinea Bissau or Conakry such as those of Wolof, Serer, Jola with fully West African composition without any North African?

Tsakhur
11-22-2016, 04:15 AM
OMG double post again!

Don Felipe
11-22-2016, 09:46 PM
Do you have any results of Senegambians, Malians, Guinea Bissau or Conakry such as those of Wolof, Serer, Jola with fully West African composition without any North African?


i'm sharing with several Senegambian people on 23andme but most of them do in fact show some minor North AFrican or East African component, although often quite minimal. However 1 Mandinga from Guinea Bissau is practically 100% SSA.

Tsakhur
11-22-2016, 09:52 PM
i'm sharing with several Senegambian people on 23andme but most of them do in fact show some minor North AFrican or East African component, although often quite minimal. However 1 Mandinga from Guinea Bissau is practically 100% SSA.

Interesting. Do you know the ethnicity or tribal group of these Senegambians? How minimal are the minor components? Is their East African component something Nlotic/Nilo-Saharan or Horner type?

Can also you post the results of these Senegambians including the 1 Mandinka from Guinea Bissau who is 100% SSA please? Btw are there also gedmatch results of them?

Thank you and regards! :)

Don Felipe
11-22-2016, 10:48 PM
You can see some of them already on this page:

https://tracingafricanroots.wordpress.com/east-africa/

I believe the socalled "East African" reported for them represents some very ancient component, quite possibly Nilotic in origin, but other explanations might also be valid. I have seen two Senegambian AncestryDNA results sofar, 1 Fula and 1 predom. Wolof person from Gambia. I will post them later on this week, as well as the Guinea Bissau result ;)

Don Felipe
12-10-2016, 10:27 PM
Can also you post the results of these Senegambians including the 1 Mandinka from Guinea Bissau who is 100% SSA please? Btw are there also gedmatch results of them?

I do not have their Gedmatch, however here's some Senegambian results. Second one is mainly Wolof from Gambia but most likely also partially Krio or Aku (https://en.wikipedia.org/wiki/Aku_people)which would explain the minor Benin/Togo. The missing 1% for the Guinean is mostly S.European while it is Middle East for the Gambian.


Guiné Bissau

https://tracingafricanroots.files.wordpress.com/2016/12/gb.jpg



Gambia (Wolof/Aku?)

https://tracingafricanroots.files.wordpress.com/2016/12/gambia-abubbby.png


Fula

https://tracingafricanroots.files.wordpress.com/2015/03/sa-fula.jpg

Oleg (Rus)
12-11-2016, 12:00 AM
This Fula has a big non-SSA admixture. Does he/she come from a region close to Berber people?

Tsakhur
12-11-2016, 12:00 AM
I do not have their Gedmatch, however here's some Senegambian results. Second one is mainly Wolof from Gambia but most likely also partially Krio or Aku (https://en.wikipedia.org/wiki/Aku_people)which would explain the minor Benin/Togo. The missing 1% for the Guinean is mostly S.European while it is Middle East for the Gambian.


Guiné Bissau

https://tracingafricanroots.files.wordpress.com/2016/12/gb.jpg



Gambia (Wolof/Aku?)

https://tracingafricanroots.files.wordpress.com/2016/12/gambia-abubbby.png


Fula

https://tracingafricanroots.files.wordpress.com/2015/03/sa-fula.jpg

Thank you! Is the Guinean with mostly 1% South European, the Mandinka from Guinea Bissau? Because I initially thought you mention that the Mandinka sample is 100% SSA.
Is the Middle East in Gambian include in the North African category?

Oleg (Rus)
12-11-2016, 12:25 AM
Thank you! Is the Guinean with mostly 1% South European, the Mandinka from Guinea Bissau? Because I initially thought you mention that the Mandinka sample is 100% SSA.
Is the Middle East in Gambian include in the North African category?

Middle East in AncestryDNA is separated from North Africa.

Don Felipe
12-11-2016, 12:09 PM
This Fula has a big non-SSA admixture. Does he/she come from a region close to Berber people?

The Fula person is from Guinea Conakry, the socalled Futa Jallon region, there are no Berber people living nearby. However it is assumed that the Fula migrated southwards and then also eastwards from their original homeland in Futa Toro. Which is situated inbetween Senegal and Mauritania.

There is still some research needed to establish how exactly the Fula ethnogenesis evolved. Personally i believe the non-SSA part might partially be attributed to intermarriage with Mauritanian Berbers taking place in a historical time period, say the last 2000 years or so. Probably most of it already absorbed before the arrival of the Portuguese in Upper Guinea in the 1400's, as they describe the Fula as having a distinct phenotype then already.

However i think partially it might also suggest a much more ancient admixture event to be traced back to a Green Sahara timeperiod. I find it very fascinating that the Fula DNA results i have seen sofar show a combination of North Africa and southern Europe. That last component should obviously not be taken literally to imply any recent Iberian connection. In fact most North AFricans also tend to score big amounts of socalled southern Europe, both on 23andme and especially on AncestryDNA. For them a historical connection with exciled Andalucian Morisco's may very well apply. Howver i think also in their case, in addition, also a much more ancient, prehistorical overlap in genetics between North AFrica and Iberia may be suggested.





Thank you! Is the Guinean with mostly 1% South European, the Mandinka from Guinea Bissau? Because I initially thought you mention that the Mandinka sample is 100% SSA.
Is the Middle East in Gambian include in the North African category?


Yes it's him and i said " practically" ;) Middle East is classified separately under West Asia by AncestryDNA.

SWAHILLI_PRINCE16
12-11-2016, 07:28 PM
Gr8 work, are there any Swahili results you might like to share as it would be interesting to see the similarity between other Swahilis from Kenya, Tanzania and Zanzibar and me.

Tsakhur
12-11-2016, 07:34 PM
Yes it's him and i said " practically" ;) Middle East is classified separately under West Asia by AncestryDNA.

OK. Maybe is because the very minor West Asia score of the result is not seen in the post. It will avoid initial a bit confusion I have regarding the West Asia.

SWAHILLI_PRINCE16
12-11-2016, 07:36 PM
My result
13006

My dads
13007

My moms
13008

Don Felipe
12-11-2016, 08:23 PM
Gr8 work, are there any Swahili results you might like to share as it would be interesting to see the similarity between other Swahilis from Kenya, Tanzania and Zanzibar and me.

I have seen one possibly Tanzanian Swahili result, but i think she might just be half and other half unknown and probably not even Tanzanian. I can post later if you like.

How many DNA matches did you get on AncestryDNA and did you test your parents on AncestryDNA as well?

SWAHILLI_PRINCE16
12-12-2016, 05:15 PM
I have seen one possibly Tanzanian Swahili result, but i think she might just be half and other half unknown and probably not even Tanzanian. I can post later if you like.

How many DNA matches did you get on AncestryDNA and did you test your parents on AncestryDNA as well?

Yes feel free to share it, I did not test my parents at AncestyDNA. How do you find out how many matches you have on ancestrydna?

Don Felipe
12-12-2016, 10:27 PM
Here's the screenshot, i'm not 100% sure but from what i've read this would be a half Tanzanian half Nigerian person and also one (great?)grandfather from Oman. Most likely her Tanzanian/Omani half will be comprised of 24% SE Bantu + 13% West Asia + 1% "Jewish" + 12% Cameroon/Congo.


1/2 Tanzania & 1/2 Nigeria

https://tracingafricanroots.files.wordpress.com/2016/12/tannaija1.jpg"]https://tracingafricanroots.files.wordpress.com/2016/12/tannaija1.jpg

SWAHILLI_PRINCE16
12-13-2016, 08:59 PM
Here's the screenshot, i'm not 100% sure but from what i've read this would be a half Tanzanian half Nigerian person and also one (great?)grandfather from Oman. Most likely her Tanzanian/Omani half will be comprised of 24% SE Bantu + 13% West Asia + 1% "Jewish" + 12% Cameroon/Congo.


1/2 Tanzania & 1/2 Nigeria

https://tracingafricanroots.files.wordpress.com/2016/12/tannaija1.jpg"]https://tracingafricanroots.files.wordpress.com/2016/12/tannaija1.jpg

Thanks for sharing, these results are like that of a Swahili person IMO. The Nigerian is could be a Southern Nigerian Bantu component and he or she scoring South Eastern Bantu as well gives me a picture of the person being from Kenya or Tanzania. As a Swahili our maternal homeland is believed to be from Shungwaya in Southern Somalia where many Bantu tribes fled because of conflict with the Oromo.

The admixture Middle Eastern and Bantu is typical of a Swahili person. Note that the admixture between the Omani/Yemeni and Bantu/Mijikenda in specific could be a reflection of what was going on in the 16th century or even starting form the 6th century AD.

Just a quick comparison...

My results(Kenyan Swahili from Mombasa)

13045

Don Felipe
12-16-2016, 11:14 PM
Thanks for sharing, these results are like that of a Swahili person IMO. .

It does look nearly like a Swahili result however i think more typically a combination of only SE Bantu + Cameroon/Congo + some minor Middle East will be seen for fully Swahili's. The socalled "Nigeria" region might indeed also be reported for people outside of Nigeria, however i'm very doubtful it would be as prominent as in this case. Also the minor but still considerable (above tracelevel) amounts of Ivory Coast/Ghana + Benin togo already betray that this person is indeed only half Swahili.

What i find interesting is that this half Swahili person shows no socalled "Africa North" at all. While you do show 4% in addition to your 3% Middle East. I am inclined to think that you inherited this socalled North African component by way of your Kikiyu side, representing some ancient and widespread Cushitic component for them.

Don Felipe
12-18-2016, 08:45 PM
How do you find out how many matches you have on ancestrydna?

Go to your DNA results summary where you also find your ethnicity estimates on the left. On the right you will see a button "view all DNA matches". It will be interesting to know how many New Worlder matches you have.

SWAHILLI_PRINCE16
12-18-2016, 09:10 PM
It does look nearly like a Swahili result however i think more typically a combination of only SE Bantu + Cameroon/Congo + some minor Middle East will be seen for fully Swahili's. The socalled "Nigeria" region might indeed also be reported for people outside of Nigeria, however i'm very doubtful it would be as prominent as in this case. Also the minor but still considerable (above tracelevel) amounts of Ivory Coast/Ghana + Benin togo already betray that this person is indeed only half Swahili.

What i find interesting is that this half Swahili person shows no socalled "Africa North" at all. While you do show 4% in addition to your 3% Middle East. I am inclined to think that you inherited this socalled North African component by way of your Kikiyu side, representing some ancient and widespread Cushitic component for them.

I was told that I have Middle Eastern ancestry through both of my two GGG grandfathers on both sides I think that's why I score an exact 7%. I'm interested if I'm a descendant of Ibn Battuta's Caravan/crew who reached and maybe settled in Mombasa. I was told that Mombasa has a big Algerian community too by my mums friend whose Algerian but born in Mombasa.

Could the European Jewish that person scores be from the Portuguese Armada which settled in Mombasa, Lamu and Kilwa, I mean there were European Jews in Iberia right?

And yes SE Bantu + Cameroon/Congo + some minor Middle East , some Swahilis even have East Asian and Iberian or European for those who come from he Comoros. I was told this by a tour guide in my Pemba Island visit.

SWAHILLI_PRINCE16
12-18-2016, 09:12 PM
Go to your DNA results summary where you also find your ethnicity estimates on the left. On the right you will see a button "view all DNA matches". It will be interesting to know how many New Worlder matches you have.

Most of my DNA matches are New Worlders and East Africans how many I cant find out.

Don Felipe
01-27-2017, 11:21 PM
Most of my DNA matches are New Worlders and East Africans how many I cant find out.

Are your matches limited to just one page or do you have several pages of AncestryDNA matches? I have seen the number of matches for several Africans now. And it's very interesting how some nationalities get many more pages than others. Obviously given Ancestry's customer database being mostly geared to Americans this will correlate with the Trans Atlantic Slave Trade patterns. Sofar from what i've seen:

Liberia (Kru/Kpelle): 28 pages (each page has 50 matches i believe)
Liberia (Kpelle, Grebo, Gola): 24 pages
Nigeria (Igbo) : 14 pages
Nigeria/Sierra Leone (Yoruba/Krio): 8 pages
Uganda (Aringa/Kakwa): two single matches on 1 page

Don Felipe
01-27-2017, 11:24 PM
Gr8 work, are there any Swahili results you might like to share as it would be interesting to see the similarity between other Swahilis from Kenya, Tanzania and Zanzibar and me.

I have one new result from a half Tanzanian, half Russian person, i don't know for sure but it seems that she might be (partially ) Swahili judging from the minor Asian and West Asian scores. You can see the breakdown on this page:

North & East African AncestryDNA Results (https://tracingafricanroots.wordpress.com/ancestrydna/african-results/north-east-african-results/)

Don Felipe
02-17-2017, 10:27 PM
Central & Southern African AncestryDNA Results (https://tracingafricanroots.wordpress.com/ancestrydna/african-results/central-southern-african-results//)





https://tracingafricanroots.files.wordpress.com/2017/02/compilcsa2.jpg







https://tracingafricanroots.files.wordpress.com/2017/02/stats-za3.jpg

Angoliga
02-18-2017, 06:50 AM
https://tracingafricanroots.files.wordpress.com/2017/02/stats-za3.jpg

From the sample set of South Africans, it's odd to see Senegalese, Malian and North African affinities whereas those of the more geographically northern countries of Cameroon and Congo have lesser/none existent traces of these affinities - could one of the South African samples have recent admixture from an ancestor bordering the Sahel? - that Senegalese 4.9% max seems quite high.

Don Felipe
03-05-2017, 09:41 PM
From the sample set of South Africans, it's odd to see Senegalese, Malian and North African affinities whereas those of the more geographically northern countries of Cameroon and Congo have lesser/none existent traces of these affinities - could one of the South African samples have recent admixture from an ancestor bordering the Sahel? - that Senegalese 4.9% max seems quite high.

It is indeed all due to only one of my still very minimal number of South African samples. It is a Coloured person who also has non-African admixture so i had to scale his African breakdown to 100%. His original amount for Senegal was 2% out of 41% African, and only at trace level.

While i do not know anything further about his background i do know that the African origins of the Cape Coloured are quite diverse as the Dutch brought over slaves from not only southeastern Africa (Mozambique & Madagascar) but also at times from West & Central Africa. So that history might very well explain it. Coloured people are bound to show more diversity not only in their non-African amounts but also in their African breakdown.

History of slavery and early colonisation in South Africa (http://www.sahistory.org.za/article/early-cape-slave-trade)

I have since added 2 new South African results though and their SE Bantu amounts are among the highest i have seen safe yours :biggrin1: One of them has 94% Southeastern Bantu but i don't have a screenshot the other one has 78% and intriguingly also seems to have some diluted Portuguese lineage in addition to an also rather high Khoisan score of 11%.


https://tracingafricanroots.files.wordpress.com/2017/02/za3.jpg

drobbah
02-19-2018, 01:13 PM
My father's results

44% SE Bantu
37% Middle East
11% North African
7% Senegal
1% Melanesian

Don Felipe
03-04-2018, 09:46 PM
West African AncestryDNA results (part 1) (https://tracingafricanroots.wordpress.com/ancestrydna/west-african-results-part-1/)


Most important findings:

(https://tracingafricanroots.wordpress.com/2018/03/04/benin-togo-also-describes-dna-from-ghana-nigeria/)

“Benin/Togo” also describes DNA from Ghana & Nigeria (https://tracingafricanroots.wordpress.com/2018/03/04/benin-togo-also-describes-dna-from-ghana-nigeria/)
“Ivory Coast/Ghana” also describes Liberian DNA (https://tracingafricanroots.wordpress.com/2018/02/24/ivory-coast-ghana-also-describes-liberian-dna/)
“South-Central Hunter-Gatherers” suggestive of remnant West African Pygmy DNA? (https://tracingafricanroots.wordpress.com/2018/02/21/south-central-hunter-gatherers-suggestive-of-remnant-west-african-pygmy-dna/)



https://tracingafricanroots.files.wordpress.com/2018/01/stats-gh22.jpg


***


https://tracingafricanroots.files.wordpress.com/2018/01/wa-lga.jpg

Don Felipe
09-22-2018, 01:45 PM
/thread

Did Ancestry kill their African breakdown? (part 1) (https://tracingafricanroots.wordpress.com/2018/09/22/did-ancestry-kill-their-african-breakdown-part-1/)





https://tracingafricanroots.files.wordpress.com/2018/09/stats-n1141.png

Tsakhur
02-26-2019, 04:24 AM
Yes it's him and i said " practically" ;) Middle East is classified separately under West Asia by AncestryDNA.

Sorry for the very late reply, I stumble on this thread again three years later and read your reply to me again. But I don't understand what does "practically" is supposed to mean in this case? Can you explain it? Sometimes I get confused trying to understand English sentences. Also I the word "said or I said" bothers me for some reason. I have a pet peeve against it.

Don Felipe
10-20-2019, 06:46 PM
West African Results (Upper Guinea) (2013-2018) (https://tracingafricanroots.com/west-african-results-ii-2013-2018/)

Summary of my Upper Guinean AncestryDNA survey findings (2013-2018). A fitting conclusion of my African AncestryDNA research as I am myself of Cape Verdean descent. Looks like the upcoming update on Ancestry will restore much of the former "Senegal" scores and correct for the usually inflated "Mali" scores right now. So this overview might be quite reflective of how things will be yet again. But either way useful already to compare relative positions of various sample groups. For more details see also:


Fula, Wolof or Temne? Upper Guinean AncestryDNA results 2013-2018 (https://tracingafricanroots.com/2019/09/19/fula-wolof-or-temne-upper-guinean-ancestrydna-results-2013-2018/)


https://tracingafricanroots.files.wordpress.com/2019/09/stats-sen-n34.png



***

https://tracingafricanroots.files.wordpress.com/2019/10/stats-fula.png

Don Felipe
11-28-2019, 01:54 PM
Looks like the upcoming update on Ancestry will restore much of the former "Senegal" scores and correct for the usually inflated "Mali" scores right now.

Indeed! Also detection of Nigerian DNA has been greatly improved. "Mali" now also covers Sierra Leone and Liberia. The new "Ghana" region remains a sore spot though. See also:

Ancestry’s 2019 Update: Back on Track Again? (https://tracingafricanroots.com/2019/11/27/ancestrys-2019-update-back-on-track-again/)



https://tracingafricanroots.files.wordpress.com/2019/11/stats-afro.png

Alfa
12-01-2019, 01:17 AM
Indeed! Also detection of Nigerian DNA has been greatly improved. "Mali" now also covers Sierra Leone and Liberia. The new "Ghana" region remains a sore spot though. See also:

Ancestry’s 2019 Update: Back on Track Again? (https://tracingafricanroots.com/2019/11/27/ancestrys-2019-update-back-on-track-again/)



https://tracingafricanroots.files.wordpress.com/2019/11/stats-afro.png

Thanks for sharing. I like this ancestrydna updates better than 23andme, but I am curious about Congo+South Africa Bantoue vs Eastern Africa Bantoue. The Eastern
Bantoue, nikotic groups like Masai and Kalenjin have it in the range of 22%, perhaps this Eastern Bantoue cluster has some nilotic in it?

Credit to @Waringari, he/she posted the nilotic_cushitic admixed:

Maasai:
2017 update:
Africa South Eastern Bantu - 73%
Africa North - 6%
Middle East - 17%
Senegal - 2%

2018 update:
Eastern Africa - 47%
Cameroon, Congo, Southern Bantu - 5%
Africa South Central Hunter Gatherers - 44%
Middle East - 4%

2019 update:
Ethiopia & Eritrea - 25%
S & E Africa HG - 25%
Eastern Bantu Peoples - 24%
Somalia - 6%
Cameroon, Congo & Bantu - 4%


Kalenjin:
2018 update:
Eastern Africa - 66%
Cameroon, Congo, Southern Bantu - 6%
Africa South Central Hunter Gatherers - 27%
Middle East - 1%

2019 update:
Eastern Bantu Peoples - 47%
S & E Africa HG - 25%
Somalia - 11%
Ethiopia & Eritrea - 9%
Cameroon, Congo & Bantu - 8%



I would greatly appreciate if you share your opinions about this. Thanks

Don Felipe
12-16-2019, 10:35 PM
I am curious about Congo+South Africa Bantoue vs Eastern Africa Bantoue. The Eastern
Bantoue, nikotic groups like Masai and Kalenjin have it in the range of 22%, perhaps this Eastern Bantoue cluster has some nilotic in it?


I would greatly appreciate if you share your opinions about this. Thanks

Interesting results from the Masaai & Kalenjin! I really cannot say that much about it as Ancestry has not disclosed what type of samples are being used for "Eastern Bantu". Aside that most likely Luhya samples were included (as this is Ancestry's code name of the region). And from what I have read (https://en.wikipedia.org/wiki/Luhya_people) these people, although Bantu speaking, might also have intermixed with neighbouring Masaai & Kalenjin. So it's probably a hodgepodge cluster and some Nilotic could be in there as well.

Within my survey the 3 highest scores for "Eastern Bantu" were for two Tanzanians (Jita/Kuria & Kuria) and 1 Ugandan (northern). All three in between 66-67% "Eastern Bantu". While in the 2013-2018 version they had in between 94%-99% for "Southeastern Bantu".

You can see the screenshots of their complete results if you scroll down towards the bottom of this page:

https://tracingafricanroots.com/2019/11/27/ancestrys-2019-update-back-on-track-again