PDA

View Full Version : AncestryDNA breakdown of SSA for African Americans & other Afrodescendants



Don Felipe
04-17-2014, 02:47 PM
Last year AncestryDNA updated their Ethnicity Estimate to include a very detailed breakdown of West African ancestry. Obviously 100% reliability cannot be guaranteed at this stage and there might still be some other imperfections as well (lacking reference populations, sample size etc.). But still I do find the results to be quite fascinating and it must be very exciting for African Americans and other Afrodescendants wanting to learn more about their African roots. Here's a link with more info

AncestryDNA Makes Scientific Breakthrough in West African Ethnicity (http://blogs.ancestry.com/techroots/ancestrydna-makes-scientific-breakthrough-in-West-African-ethnicity/)

I've been collecting results posted on the net, right now i have over 130 results for Aframs and many results for other Afrodescendants as well. For easier comparison I have recalculated everyone's original African %'s so that they add up to 100%. In other words only the African part of the results haven been taken in consideration and they have been scaled to 100%. For example, someone with a total African score of 75% and a Nigeria score of 25% would get an adjusted score of 25/75= 33,33% Nigeria out of his total African ancestry.

Here's a link to the spreadsheet which contains all the results. Besides a tab for the results and the statistics there's several other tabs on the bottom as well where i sorted the results according to USA origins and also for each African region from highest to lowest score.

Spreadsheet with results (https://docs.google.com/spreadsheet/ccc?key=0AnR-7UV1t1NWdFJyMGd6V3dqSHZPaHhPcjJ0OTJmNUE&usp=sharing#gid=0)

Really only the AA sample seems big enough to make any generalizing statements but the other ones are very interesting in themselves as well. I find that comparing these results between people of different backgrounds helps to make better sense of your own personal results but also gives more insight on grouplevel into the possible ancestral roots within Africa of various Afrodescended ethnicities like African Americans, Jamaicans, Haitians, Cape Verdeans and Dominicans. Learning which regional ancestry they might have in common and also where their overall profile looks more specific/typical for the ethnic group they belong to. Here's a screenshot of the stats sofar (see tab "Stats" in spreadsheet).

*******

EDIT: for an updated overview see this link: https://tracingafricanroots.wordpress.com/ancestrydna/

http://i989.photobucket.com/albums/af18/oditous2/StatsAA134_zps7ede8f93.jpg~original




Observations for African American (AA) results:



AncestryDNA results seem to deviate A LOT from what's predicted by GED-Match calculator Africa9. They are much more varied and shifting across various parts of West/Central Africa for each person, while the Africa9 results were rather consistent with Kongo, Bamoun and Fang (Cameroon and Congo) as the topmatches for most African Americans. Going by what 's been documented on slave trade patterns AncestryDNA seems much more reliable to me.




Although individual results vary a lot, on average the results seem to be roughly in line with what's known from slave trade records. Nigeria would be the biggest component (it also has the highest outliers) followed closely by Congo/Cameroon and Ghana/Ivory Coast, if you combine Senegal and Mali as 1 Upper Guinean category it would also be among the top 4 although perhaps less prominent than expected.




Biggest surprise is the high score for Benin/Togo (~15%), because that's an area where relatively few slaves were shipped from to the US according to the records. Below 5% according to most estimates. Taking a sceptical stand you could say this is the very first DNA test attempting to disentangle SSA regional roots and Ancestry.com also mentions themselves that their Ghanaian samples cluster very closely with their Benin ones. So possibly there's some misreading going on. On the other hand slave trade statistics don't tell the full story and future more refined DNA testing might debunk some myths about ethnic origins of AA's and other Afrodescendants. For one i think substantial Ewe slave exports via Ghana could partially be causing this outcome. According to Wikipedia 32% of population in Togo are Ewe and they are about 14% in Ghana. Partially it could also reflect some founding effect of Fon/Gbe people from Benin/Togo arriving via Barbados/Jamaica in the late 1600's early 1700's. Actually in the early slave trade period the Bight of Benin was a major stopping zone for English traders. Only later on Ghana and Biafra became more prominent.




Another surprise is the relatively high SE Bantu score (~7%), according to slavevoyages.org (http://slavevoyages.org/tast/database/search.faces) it would only have been 2%. But it's VERY likely that this category is actually reflecting genetical similarity being picked up for ancestry coming from Congo/Angola rather than Mozambique or Madagascar.




There's no separate category yet for Sierra Leone/Liberia/Guinea. But many AA's would have some partial ancestry coming from these parts. So i suppose right now this portion is being included in other regions, Ghana/Ivory Coast & Mali perhaps being the most likely candidates.




The Cameroon-Congo category is defined rather broadly. Given the widely reported Igbo contribution to the AA genepool it would be very useful if it gets split up in a future update in order to distinguish between ancestry coming from the Bight of Biafra and Congo proper. An actual Igbo person taking this ANcestryDNA test is likely to also get some susbtantial % for Cameroon/Congo in addition to just Nigeria.





Observations for Cape Verdean (CV) results




Their consistently high Senegal scores (above 50% for everyone) combined with significant Mali and North African scores seem to confirm predom. Senegambian/Guinean ancestry for CV's and also that AncestryDna analysis is not completely random :D




Intriguingly there's also other components showing up besides the expected Senegal/Mali/North Africa and even relatively high scores for Cameroon/Congo and SE Bantu. This is kind of surprising given Cape Verde's geographical location and also everything i've read about slave trade patterns between Cape Verde and the mainland. Which clearly describe the area in between Senegal and Sierra Leone as practically the only provenance zone for slaves brought to Cape Verde safe for some individuals who came on random slave voyages from further away. I suppose it might still be possible that slaves being brought from Angola or even Mozambique to CV were more frequent than recorded by history. Eventhough there's no independent genetical proof for this going by maternal haplogroups (all studies sofar suggest a Senegambian MT-DNA profile) and also not going by retention of any Central African cultural traditions in Cape Verde which has only Guinean and Senegambian influences in its Crioulo culture afaik. See also these papers:
[*=4]Mitochondrial portrait of the Cabo Verde archipelago: the Senegambian outpost of Atlantic slave trade (http://www3.uma.pt/abrehm/v1.1/docs/downloads/pdfs/Brehm_mtDNACaboVerdeCV_AHG2002.pdf)
[*=4]Dissecting the Within-Africa Ancestry of Populations of African Descent in the Americas (http://www.plosone.org/article/fetchObjectAttachment.action;jsessionid=D83AA0C839 05F9D6351B5073CDF96215.ambra01?uri=info%3Adoi%2F10 .1371%2Fjournal.pone.0014495&representation=PDF)





There might be interisland differences showing up. Although only a larger samplesize can reveal this more clearly. Right now most of them are from Santo Antão, but there's also 2 with (partial) origins from Brava and 1 from Fogo. Generally speaking CV islanders will be descended from the same founding populations. Fogo and Santiago being the islands with the earliest settlement. Other islands mostly being populated by people coming from these 2 islands later on in history. However because of bottleneck effects and recombination the regional DNA markers might have been preserved/inherited in diverging proportions for each island.




In any case the sample size is way too small to make any conclusions yet regarding these unexpected non-Senegambian/Guinean scores. Results might be different for other islanders because of some bottleneck effect on Santo Antão. I suppose another explanation could be that somehow the analysis by AncestryDna still isn't entirely waterproof although surely a big improvement because it does show a significant pull towards Senegal & to a lesser extent Mali for CV's as was to be expected. The only thing which could clear this case up would be to see the results of a fully Senegalese or Guinean guy and see if he might also have some %'s from Cameroon-Congo or even SE Bantu.




Observations for Puerto Rican (PR) results




Combined Senegal and Mali results are higher on average than for Aframs. High Upper Guinean scores (Senegal+ Mali) seem to confirm the founding effect of the early 16th/17th century slave imports into the Spanish Carribean which were predominantly coming from Senegambia/Guinea by way of Cape Verde. This outcome seems to corroborate earlier findings in this recent study:

Reconstructing the Population Genetic History of the Caribbean (http://livasperiklis.files.wordpress.com/2013/11/plos-genetics-reconstructing-the-population-genetic-history-of-the-caribbean.pdf)


We find evidence of two pulses of African migration. The first pulse—which today is reflected by shorter, older ancestry tracts—consists of a genetic component more similar to coastal West African regions involved in early stages of the trans-Atlantic slave trade. The second pulse—reflected by longer, younger tracts—is more similar to present-day West-Central African populations, supporting historical records of later transatlantic deportation.




Balanced results for all categories on average, only the relatively high North Africa score really stands out. Confirming Canarian origins i suppose. It's higher than the CV average but from what i've seen on 23andme it's the other way around with CV's usually scoring higher for North African. The sample is really small though and no CV's from Fogo or Brava yet. Also unlike on 23andme North African is included in total African which creates a skewed outcome if total African is minor as it is for PR's.




Because of their relatively low total African ancestry there might be a higher chance of regional outliers among PR's. Recombination leading to a disproportional % of inherited DNA markers for 1 particular region. Besides North African this also seems to be the case for 2 very high Mali scores sofar.




Observations for Dominican (DR) results




Their Senegal score is the highest among New Worlders sofar. Might be a clue that they share much Senegambian/Guinean ancestry with CV's dating from the 16th/17th centuries as was seen for the PR's as well. This founding effect is probably also going to show up for other Latin Americans.




Congo/Cameroon & SE Bantu seem to be consistently high sofar as well, in line with cultural retentions from Congo in the DR. Sample size isn't really big yet. But using a 3-way division of the African regions on AncestryDNA: Upper Guinea (= Senegal+Mali), Lower Guinea (Ghana/Ivory Coast + Benin/Togo + Nigeria) & Central/Southern Africa (Cameroon/Congo+ SE Bantu). I suppose we can say that AA's are mostly Lower Guinean with a slight shift to Central African. CV's are predom. Upper Guinean while PR's & DR's are almost evenly split but showing highest outliers for Upper Guinea and Central African.




Observations for Jamaican results




Surprisingly low Ghana scores sofar given the widely reported Akan connection for Jamaica! Important to underline there's only two Jamaican results to go by. But goes to show these results can be wideranging depending on someone's personal family tree and how many regional DNA markers were inherited through recombination.




Also rather high Mali & Benin/Togo. A similar explanation as used for the AA's should be valid for the Jamaicans as well regarding the unexpectedly high Benin/Togo. The Mali category has the lowest confidence range according to AncestryDNA. So i suppose it could also signal northern Ghanaian or Burkina Faso ancestry for Jamaicans.




Very low Senegal & SE Bantu scores are in line with slave trade records. Compared to African Americans Jamaicans are more likely to end up being predominantly Lower Guinean on average while the Central African and Upper Guinean parts will be insignificant for most. Which is not the case for AA's, Upper Guinean & Central African being minor compared to Lower Guinean but still very noticeable for most.




Observations for Haitian results




High Benin & Congo scores are perfectly in line with what's known about Haitian ethnic make-up from historical documentation as well as cultural retention. Regrettably only 1 result to go by though. I expect Haitian results will vary just like the ones seen among AA's but perhaps less so. Would be interesting to see which nationality shows the most or least consistent mix of African regions.

Don Felipe
04-17-2014, 04:13 PM
http://i989.photobucket.com/albums/af18/oditous2/West-African-ethnicity.jpg~original



http://i989.photobucket.com/albums/af18/oditous2/map-Africa_zpsf6d10da7.png~original

Some info on the categories being used by AncestryDNA. As always the labeling should not be taken too literally as it's dependent on the reference populations being used and also doesn't exclude the possibility of mutual overlap. I haven't done the AncestryDNA test myself yet but one thing i like about this company is that they provide much more details and clarification on how to interpret your results than 23andme for example...It's a more honest approach and avoids any possible misleading of people without enough background information.

Not sure how reliable the results might be, only test results of genuine Africans with certified ethnic/regional background can tell. But when describing a typical person from one of their 9 categories Ancestry.com mentions explicitly that because of widespread shared ancestry regionally speaking, a "typical" Malian or Nigerian will usually also get a breakdown into several categories and not fit 100% in their national category, in other words none of the categories are unique or discrete, because just like anywhere else in the world African ethnicities have been intermixing for ages and show overlapping ancestral markers. Genetics doesn't respect manmade borders. If you go back far enough in time you're always going to find some ancestral connections and therefore genetical similarity between populations even if they're geographically far apart. This complicates a breakdown into neat categories we all would like to see, but being aware of this fact does provide more insight for your results.

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


According to this link (http://dna-explained.com/2013/10/17/ancestrys-updated-v2-ethnicity-summary/) he following samples were used

Africa Southeastern Bantu 18
Africa North 26
Africa Southcentral Hunter Gatherers 35
Benin/Togo 60
Cameroon/Congo 115
Ivory/Ghana 99
Mali 16
Nigeria 67
Senegal 28

It also mentions that the socalled Human Genome Diversity Project data were utilized. In this database there's a sample of about 20-30 Mandenka individuals from Senegal. So it's very likely they're the reference population being used by AncestryDNA to define their Senegal category. Which would mean that the Senegal category is in fact also a proxy for a subgroup of Mande speakers. I used to think it might be a proxy for Atlantic speakers instead (Wolof and Guinea Bissau) while the Mali category could be a proxy for Mande speakers. This would seem logical given the CV results. Not sure yet which samples were used for the Mali category but might be either Dogon or Bambara. The samples being used for Cameroon-Congo are probably the same Bamoun, Fang and Kongo samples as utilized by the various GED Match calculators. South Africans were most likely used to define the SE Bantu category. Nigeria probably has Yoruba samples but not sure if it also includes Igbo ones. Also not sure about which ethnic groups were sampled for Ghana-Ivory Coast and Benin-Togo.

In addition they might also still be using the Sorenson database (SMGF (http://www.smgf.org/maps/collections.jspx)) like they did before the update.

The Benin/Togo and Ghana/Ivory Coast categories seem to cluster most closely according to this plot. Cameroon-Congo is quite condensed while Nigeria, Mali and Senegal are scattered. Senegal seems to have the most distance to the other clusters.

http://i989.photobucket.com/albums/af18/oditous2/African-Ancestry-2.jpg~original


Some quotes on how AncestryDNA describes the categories themselves (they provide even more detailed information on ethnic groups which i didn't copy)


Senegal

Primarily found in: Senegal, Gambia
May also be found in: Guinea-Bissau, Guinea, Mali, Mauritania

Typical Native 100%

"The Senegal region is less admixed than most other regions. That
means that when creating ethnicity estimates for people native to
this region, we sometimes see small amounts of DNA from other nearby
regions included. The typical person born in this region today has
about 100% of his DNA unique to this region."

Other regions commonly seen in people native to Senegal
From our reference collection of 28 people

Region % of natives that have this region
Mali 46%
Benin/Togo 7%






Mali

Primarily found in: Mali, Guinea
May also be found in: Ivory Coast, Burkina Faso, Senegal

Typical native 39%

The Mali region is more admixed than most other regions. That means
that when creating ethnicity estimates for people native to this
region, we often see DNA from other nearby regions included--in fact
the typical person born in this region today has only about 39% of
his DNA unique to this region."

Other regions commonly seen in people native to Mali
From our reference collection of 16 people

Region % of natives that have this region
Ivory Coast/Ghana 69%
Senegal 69%
Benin/Togo 25%
Nigeria 25%
Cameroon/Congo 19%





Ivory Coast/Ghana

Primarily found in: Ivory Coast, Ghana
May also be found in: Togo, Mali, Nigeria

Typical native 86-100%

Other regions commonly seen in people native to Ivory Coast/Ghana
From our reference collection of 115 people

Region % of natives that have this region
Benin/Togo 43%
Mali 32%
Nigeria 11%
Senegal 8%
Africa Southeastern Bantu 2%
Cameroon/Congo 1%







*Benin/Togo

Primarily found in: Benin, Togo
May also be found in: Ghana, Nigeria, Mali

*Benin/Togo

Typical native 82-100%

Other regions commonly seen in people native to Benin/Togo
From our reference collection of 60 people

Region % of natives that have this region
Ivory Coast/Ghana 43%
Nigeria 28%
Mali 25%
Cameroon/Congo 10%
Senegal 3%
Africa Southeastern Bantu 3%



Nigeria

Primarily found in: Nigeria
May also be found in: Niger, Benin, Cameroon

Typical native 69-95%

"The Nigeria region is the most admixed of our identified regions.
That means that when creating ethnicity estimates for people native
to this region, we almost always see DNA from other nearby regions
included--in fact the typical person born in this region today has
only about 69% of his DNA unique to this region."

Other regions commonly seen in people native to Nigeria
From our reference collection of 67 people

Region % of natives that have this region
Benin/Togo 45%
Cameroon/Congo 37%
Africa Southeastern Bantu 22%
Mali * 13%
Ivory Coast/Ghana 9%
Senegal 7%







Cameroon/Congo

Primarily located in: Cameroon, Gabon, Congo, Republic of Congo
Also found in: Angola, Chad


Typical native 92-100%

Other regions commonly seen in people native to Cameroon/Congo
From our reference collection of 115 people

Region % of natives that have this region
Nigeria 21%
Africa Southeastern Bantu 21%
Benin/Togo 10%
Mali 10%
Ivory Coast/Ghana 4%
Africa South-Central Hunter-Gatherers 2%
Senegal 1%





South-Central African Hunter-Gatherers

Primarily found in: South Africa, Namibia, Botswana, Congo
May also be found in: Angola, Zambia, Tanzania, Kenya

Typical native 86-100%

Other regions commonly seen in people native to Africa South-Central Hunter-Gatherers
From our reference collection of 35 people

Region % of natives that have this region
Cameroon/Congo 54%
Africa Southeastern Bantu 9%
Mali * 6%
Ivory Coast/Ghana 3%







Southeastern Bantu

Primarily found in: South Africa, Kenya, Namibia, Botswana,
Zimbabwe, Zambia, Angola, Tanzania, Mozambique, Uganda

May also be found in: Nigeria, Congo

Typical native 72-100%

"Like most of our identified regions, the Africa Southeastern Bantu
region is admixed, which means that when creating ethnicity
estimates for people native to this region, we frequently see DNA
from other nearby regions included. The typical person born in this
region today has about 72% of his DNA unique to this region."

Other regions commonly seen in people native to Africa Southeastern Bantu
From our reference collection of 18 people

Region % of natives that have this region
Cameroon/Congo * 44%
Africa South-Central Hunter-Gatherers 39%
Nigeria 17%
Near East 17%
Africa North 6%
Ivory Coast/Ghana 6%
Mali 6%
Senegal 6%

Don Felipe
04-17-2014, 04:39 PM
http://i989.photobucket.com/albums/af18/oditous2/EltisUSA99_zps95858018.jpg~original


http://i989.photobucket.com/albums/af18/oditous2/1_004Mw_zps500a48d0.jpg~original


It's known that there were distinct regional patterns in the direct African slave trade for the main ports of entry into the US. Therefore i separated some results for people who according to their own knowledge have deep roots within either Virginia/Maryland, Carolina's/Georgia or the Gulf states. I only got this info for a minority of the results though.

It seems however that few Aframs can trace their origins to only one state or region. Going back not just 3 generations but 200 or 300 years when their first Africanborn ancestors arrived in the US. This because of the high mobility of Aframs throughout US history and especially because of the socalled second Middle Passage to the Deep South, incl. the Gulf States. So basically I suspect that in many cases it could be quite complex to find much correlation because of all the mixing and migrations that took place. Very deeprooted rural and isolated communities might be the exception though.

Check these links for some excellent background info and maps

http://rootsrevealed.blogspot.nl/2013/10/the-african-americans-many-rivers-to_30.html
http://www.inmotionaame.org/migrations/landing.cfm?migration=3


Here goes the results i have, you can contrast the regional averages against the nationwide averages "AFRAM" (screenshots taken at an earlier moment than in the OP).

This is what stands out the most to me:




relatively high Upper Guinean (Senegal + Mali) score for the Carolina's/Georgia (SC&GA)), to be explained by the well known Sierra Leone/Gambia connection for this area? First two results are Gullah's.



relatively high Ghana & Nigeria for Virginia/Maryland (VA&MA), to be explained by the documented widespread presence of Igbo's as well as early Carribean imports via Jamaica & Barbados?



surprisingly low Upper Guinean (Senegal + Mali) results for the Gulf States (GS) and relatively high Nigeria average. Not really in line with what you'd expect going only by French slave imports into this area.



South Carolina/Georgia

http://i989.photobucket.com/albums/af18/oditous2/SCresults_zps71359a4c.jpg~original


Virginia

http://i989.photobucket.com/albums/af18/oditous2/VAresults_zps9f4076ce.jpg~original


Gulf States

http://i989.photobucket.com/albums/af18/oditous2/Gulfresults_zps7f8d2546.jpg~original




Remember the samplesize is very small so best not to jump to conclusions yet ;) I think the results from the Carolina's/Georgia might be most reliable because there's more who said they trace all their ancestry from just that area, some even with fully Gullah roots. The Gulf states i think might be a little misleading because i also included states like Texas and Florida, whose US regional origins might ultimately go back mostly to SC and VA (if you're able to trace back beyond 1870) and are not per se related to the slave trade conducted by the French in Louisiana/Mississippi. I've always been wondering how many people from Louisiana and neighbouring states can trace their roots to the French period. I always assumed this would be the case for a majority of them. But i'm not so sure now after reading about the domestic slave trade. It might be that Louisiana was rather thinly settled in the French time and therefore a majority of Aframs there descend from people brought in from the Upper South after the Louisiana purchase. The results sofar in my sheet for people from the Gulf states also seem to indicate so because they're out of line (relatively speaking) with the provenance zones known from French slave trade (esp. Senegal + Mali) and instead follow fairly closely the national averages. Traditionally Lousiana is seen as being the area that received the highest frequency of Senegambians (almost 40%!). However notice in the first chart above how the absolute numbers involved (8600) were very small compared with imports via South Carolina & Virginia.

Don Felipe
04-17-2014, 05:00 PM
Posting some AA results to show the range of their individual variation. Sofar Nigeria stands out as being most often at nr.1 and also having the biggest ratio to total African. Highest score sofar being 69%. Cameroon/Congo, Ghana/CIV and even Benin/Togo also very often come in at nr.1 or atleast the top 3. Senegal, Mali and SE Bantu are much less frequently seen in people's top 3 and almost never coming in at nr.1. The ranking can be VERY different though for each individual.

I sort of expected the top 3 regions to have been more consistent given that compared to other Afrodescendants in the New World the Aframs are descended from a relatively small founding population. The USA being one of the few colonies where slaves had a positive reproduction rate from the start which canceled the need for massive imports like in the Carribean and Brazil. Also the domestic slave trade whereby slaves from the Upper South (esp. Virginia) were transported to the Deep South (esp. after Lousiana was sold by the French) would have lead to Aframs having much shared ancestry. Intermixing within the Afram community steadily increasing because of 20th century migrations to the North and West. I suppose despite all that of course Aframs do also have unique family trees and individuals may have inherited disproportionally from specific lineages when compared with others. It could be that this is also reflected in the widely varying results sofar.


Nigeria #1


This person has the 2nd highest Nigeria score sofar (46/73=63%), usually it's more balanced. Senegal coming in at nr.2 is also a bit atypical.

http://i989.photobucket.com/albums/af18/oditous2/AANig1Sen2.jpg~original

This one has the highest Nigeria ratio right now, almost 70% of total Africa! (29/42=69%), he's probably half Afram.

http://i989.photobucket.com/albums/af18/oditous2/AA5_zps99ec5a30.jpg~original

Another very high Nigeria score, this time with Mali as nr. 2

http://i989.photobucket.com/albums/af18/oditous2/AANig1Mal2.png~original


Nigeria at nr.1 again but not as dominant, very evenly distributed. Interesting how SE Bantu & Cameroon/Congo are almost equal.


http://i989.photobucket.com/albums/af18/oditous2/BBGAfrican-Ancestry_zpsbc1e28a4.png~original



Cameroon/Congo#1

Sometimes Nigeria gets really low though, some people even score 0% for it like this one. Congo/Cameroon on the other hand is very strong, 50% of total African.

http://i989.photobucket.com/albums/af18/oditous2/AACo1.jpg~original


Other people with Cameroon/Congo at nr.1 but still very diverse and predom. West African.


http://i989.photobucket.com/albums/af18/oditous2/AA6_zps2f98a12c.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AA7_zps07b9706d.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AA2Congo1Ghana2.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AACo1.png~original



Benin/Togo #1

First 2 have about 50 % Benin/Togo out of total African, first one also has quite high SE Bantu.


http://i989.photobucket.com/albums/af18/oditous2/AA5Benin1SEB2.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AA3Benin1Ghana2.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AA1_zps9507dce6.jpg~original



Ghana/Ivory Coast #1


Some high Ghana/CIV for these ones. Second one is a child of first one, shows how recombination can mix up the ranking of the regions. But nr.1 spot usually stays the same. Third one shows very balanced results.

http://i989.photobucket.com/albums/af18/oditous2/AA6senegal3mum.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AA6daughter.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AA4Ghana1Congo2.png~original

http://i989.photobucket.com/albums/af18/oditous2/AA2_zps04835ab1.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/AA3_zpsbd13dabd.jpg~original

Don Felipe
04-17-2014, 05:37 PM
Some non-AA results for contrast


Cape Verde (from Fogo)

VERY high Senegal (45/61), taken as % of total African it's 73,8% the highest i've seen scored for any category by anyone sofar. Aframs highest single category score being 69% for Nigeria, for Dominicans it's 42% for SE Bantu and for PR's a 72% ratio out of total African for Mali. The PR ratio's more likely to be extreme because of their lower total African though.

http://i989.photobucket.com/albums/af18/oditous2/CV-1.png~original



Puerto Rico

Very high Mali score for first one, 72% (13/18) of total African. Below is a result of a halfsibling for this person. Notice there's no Mali but instead Senegal having the highest score!

http://i989.photobucket.com/albums/af18/oditous2/PR-BF_zps383b4e91.jpg~original



http://i989.photobucket.com/albums/af18/oditous2/PR-BR2_zps27fdfb45.jpg~original


Highest SE Bantu on a PR sofar: 34,4% (11/32). Trace regions are not being shown but most likely there would be either Senegal or Mali there as well.

http://i989.photobucket.com/albums/af18/oditous2/TR-PR_zpsabd5d7a8.jpg~original



This one is is half PR, half DR. High Benin/Togo score (19/51=37.3%). Still his combined Senegal and Mali is in line with other DR/PR results.

http://i989.photobucket.com/albums/af18/oditous2/DR-PR_zpsd531d3b4.jpg~original



Dominican Republic

Senegal coming in at either position 1 or 2. Many other DR's will probably show the same pattern.

http://i989.photobucket.com/albums/af18/oditous2/DR2_zps845e34c0.jpg~original



http://i989.photobucket.com/albums/af18/oditous2/DR1_zpsa2ee82a7.jpg~original



Mexico

VERY interesting because it's usually assumed that the minor African ancestry reported for Mexicans would be overwhelmingly Angolan. But this isn't showing up in these 3 results. Instead Upper Guinean seems to be most important for these individuals if you combine Mali with Senegal and contrast with combined scores for SE Bantu + Congo/Cameroon. Of course total African is minor so recombination could distort things.

It's indeed true that during the peak of slave imports into Mexico (1590-1640) Angola was undoubtedly predominant as supplier of slaves to Veracruz, perhaps even as much as 90%. After 1640 when the Portuguese assiento ended, slave imports started to decline fast but still continued up till the early 1700's, most of them probably came from Ghana-Benin. However in the period 1520-1590 it was mostly Senegambians/Guineans who entered Mexico by way of Cape Verde or as socalled Ladino's from Iberia/DR. It seems their legacy still lives on in the genes of some presentday Mexicans.


http://i989.photobucket.com/albums/af18/oditous2/MEX2x.jpg~original



http://i989.photobucket.com/albums/af18/oditous2/Mex3_zpsb4e37f85.jpg~original


Columbia

Again some founder effects from either Upper Guinea or Congo/Angola even though total African is very minor. Black Colombians might show some more diversity though also reflecting African ancestry coming from Lower Guinea from the 1700's/1800's.

http://i989.photobucket.com/albums/af18/oditous2/Colombia_zps3b7d65c2.jpg~original



Cuba

Same comment as for Colombia.

http://i989.photobucket.com/albums/af18/oditous2/Cuba_zps55737003.jpg~original

RCO
04-17-2014, 06:59 PM
Thanks for showing the new proportions !

Mehrdad
04-17-2014, 07:03 PM
Thank you for sharing this, its interesting to note that Cape Verdeans, Dominicans and Puerto Ricans share a higher concentration of North African admixtures than African Americans and other people who may have had Slave ancestry.

Don Felipe
04-17-2014, 07:20 PM
Thanks for showing the new proportions !

Do you know any Brazilians who took the AncestryDNA test? Should be very interesting to see their results if they have any African ancestry. Most likely for people from Rio and Minas Gerais it will be mostly SE Bantu & Congo-Cameroon, reflecting Angolan, Congolese and Mozambican ancestry. But for Bahians it might be mostly Benin/Togo & Nigeria reflecting the high Yoruba connection. And for people from Maranhão & Amazonas it might even be Senegal which comes on top because unlike other parts of Brazil most Africans arriving there came by way of Guiné Bissau.


Thank you for sharing this, its interesting to note that Cape Verdeans, Dominicans and Puerto Ricans share a higher concentration of North African admixtures than African Americans and other people who may have had Slave ancestry.

Indeed, i suspect for Dominicans and Puerto Ricans it might mostly be derived from their Canarian ancestry. For CapeVerdeans it could be either directly from North Africa as well because some Moriscos and Berber slaves might have been brought there in the early settlement period (1460-1560). But partially it might also be because of Sahellian ancestry, esp. Fula, who might also carry an ancient North African component.

See also this thread

23&me + GED-Match results of a Guinean Fula from Fouta Djallon (http://www.anthrogenica.com/showthread.php?681-23-amp-me-GED-Match-results-of-a-Guinean-Fula-from-Fouta-Djallon)

RCO
04-17-2014, 08:08 PM
Don Felipe, I don't know Brazilians who tested with AncestryDNA but this new methodology would be very useful as Brazil was the biggest importer of Africans in the New World with the biggest territory and for the longest time. I hope they can improve and refine that kind of genetic tools.

Mehrdad
04-17-2014, 10:17 PM
Interesting thing, on the admixturemap.paintmychromosomes (http://admixturemap.paintmychromosomes.com/) website the Mandinka people of West Africa come close to Moroccans at 7.7%.

Don Felipe
04-18-2014, 11:03 AM
Very cool survey done by Ancestry.com!


A Genetic Census of America

Using AncestryDNA results from over a quarter million people, the AncestryDNA science team set out to perform a “genetic census” of the United States: a survey of the U.S. using only DNA. Where did the ancestors of today’s Americans come from? Do Americans in the Midwest hail from similar places of the world as in the Northeast, or as in the South?

- See more at: http://blogs.ancestry.com/ancestry/2014/04/04/a-genetic-census-of-america/#sthash.rVYxosg5.dpuf

If you follow the link you can select each one of the 26 regions used by AncestryDNA and see the average score for each state. Interesting to compare with my own findings sofar which were of course based on a way smaller sample size :biggrin1:. Then again Aframs in this survey aren't separated from people with other ethnic backgrounds and they only considered birth place and not deep rooted ancestry from 1 single state/area.

I made some screenshots to see where each African category had its highest score. All things being the same you would expect this to be the state with the highest % of Aframs. Which would be Mississippi.


By 2010 census results

Rank State African-American Alone
Population (2010)[1] % African-American
1 Mississippi 1,074,200 37.30%
2 Louisiana 1,452,396 31.98%
3 Georgia 2,950,435 30.02%
4 Maryland 1,700,298 29.44%
5 South Carolina 1,290,684 28.48%
6 Alabama 1,251,311 26.38%
7 North Carolina 2,048,628 21.60%
8 Delaware 191,814 20.95%
9 Virginia 1,551,399 19.91%
10 Tennessee 1,055,689 16.78%



This is however NOT the case for Mali, Senegal and SE Bantu. Despite having a lower % of Aframs than Mississippi South Carolina still scores more for it. So i suppose this confirms that the Carolina's have higher than average Senegambian ancestry and also Congolese/Angolan (which was already known from historical references).


http://i989.photobucket.com/albums/af18/oditous2/Senegal_zps1b8ca0e1.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/Mali2_zpsd59432a2.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/SEBantu_zps505cdad1.jpg~original


For the other categories it was indeed as expected Mississippi with the highest %'s.

http://i989.photobucket.com/albums/af18/oditous2/Cameroon-Congo_zpsce368eed.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/Ghana-IvoryCoast_zpsa6a8d2dd.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/Nigeria_zpscbb15d9a.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/Benin-Togo_zpsbde037c7.jpg~original





Interestingly Cameroon-Congo had a higher average than Nigeria, unlike my findings sofar. However if you look at Virginia they have higher Nigeria than Cameroon-Congo. Confirming the strong Igbo presence?

http://i989.photobucket.com/albums/af18/oditous2/NigeriaVA_zpsa2e0cc81.jpg~original

http://i989.photobucket.com/albums/af18/oditous2/Cameroon-CongoVA_zpsd01936cf.jpg~original

Don Felipe
05-20-2014, 10:33 AM
Results of an Egyptian, mostly West Asian with only minor SSA which comes out as SE Bantu but might as well be East African i suppose because there's no separate category for that. Also interestingly only minor North African, i guess it means the North African category on AncestryDNA is more strictly Maghrebi than on 23andme where even Saudi samples are being used to define their "North African" category...

http://i989.photobucket.com/albums/af18/oditous2/AncestryDNA-02Egypt.png~original




Horner results, most likely Somali

http://i989.photobucket.com/albums/af18/oditous2/Omar10251274_629747697106657_1319571827_n_zps203a4 326.jpg~original

SSA part is being labeled SE Bantu, again because of lack of more fitting categories. Also interesting that West Asian > North African. Before they did their update it used to be the opposite on 23andme for Horners if i'm not mistaken.

Dessalines
05-20-2014, 03:18 PM
Thank you for sharing this, its interesting to note that Cape Verdeans, Dominicans and Puerto Ricans share a higher concentration of North African admixtures than African Americans and other people who may have had Slave ancestry.

Perhaps the higher percentage of North African admixture results from the Moorish presence in Spain.

Don Felipe
08-09-2014, 11:47 AM
New stats for African Americans, over 200 results processed right now B)
I don't expect the averages to change drastically anymore to be honest, this seems to be pretty much the average regional breakdown for AA's nation wide according to AncestryDNA and solely focussed on their African side. But the averages are hiding much underlying variation so it's still interesting to have a look at the outliers, which are shown in "maximum" and "ranked #1"



http://i989.photobucket.com/albums/af18/oditous2/AA204_zps4decd8c2.jpg~original




When it comes to exceptionally high scores for just one category it's usually Nigeria, Cameroon/Congo or Ghana/Ivory Coast. I was a bit surprised about the small number of Senegal & Mali outliers, but perhaps if they had been combined into 1 "Upper Guinean" category there would have been more. As i already mentioned earlier many of the higher Senegal or Mali scores seem to occur for people with deep USA roots in South Carolina. Here's a cool map by Ancestry which confirms that finding.




http://i989.photobucket.com/albums/af18/oditous2/SenegalinSC_zps72b022ae.jpg~original





Posting some individual results that seem to be outliers, keep in mind most results will be more balanced and in line with the averages.


High Senegal

First 2 persons indeed from the Carolina's. Senegal rarely pops up at #1 or even the top 3 for the AA results i've seen sofar, only 2 other persons i've seen had Senegal at nr.1. His Mali is also quite high btw.



http://i989.photobucket.com/albums/af18/oditous2/Senegal1_zps7267b22e.jpg~original


This is his son who has the highest Senegal % i've seen eventhough it's not at nr.1


http://i989.photobucket.com/albums/af18/oditous2/Senegal2_zps39c75985.jpg~original




this one's prolly half Afram, from Pensylvania, but who knows perhaps going back some generations also with Carolina roots.


http://i989.photobucket.com/albums/af18/oditous2/Senegal1_zpscaf4c6a4.png~original



High Mali

Mali also rarely makes a top 1 spot but more often than Senegal and with higher %'s


http://i989.photobucket.com/albums/af18/oditous2/Mali1_zps2b0127bf.jpg~original




http://i989.photobucket.com/albums/af18/oditous2/Mali1_zps65167704.png~original


http://i989.photobucket.com/albums/af18/oditous2/AA8Mali2_zps7bb320ef.jpg~original


Also notice rather high SE Bantu aside from high Mali


http://i989.photobucket.com/albums/af18/oditous2/Mali2_zps7fb3330b.png~original




High Southeastern Bantu

Angolan/Congolese ancestry in AA's seems to be split up between this category and the Cameroon-Congo one. The SE Bantu being less significant and only rarely showing up in top 3 scores of AA results i've seen sofar. Here's a few though

http://i989.photobucket.com/albums/af18/oditous2/SEBantu1_zps5671e247.png~original



At number 2 but higher % than first one.


http://i989.photobucket.com/albums/af18/oditous2/SEBantu2_zps3ebb3555.jpg~original




High Nigeria

Nigeria gets the most #1 spots and also the highest outliers. This one still showing high Mali in addition.

http://i989.photobucket.com/albums/af18/oditous2/NiegriaampMali_zpsae28ad74.png~original


This one having Senegal in top 3

http://i989.photobucket.com/albums/af18/oditous2/Nigeria1_zps0b7d416c.jpg~original

Don Felipe
09-13-2014, 11:07 AM
Two atypical AA results, from what i've seen sofar very unusual but still part of AA variety.


This guy has the craziest results i've seen for an Afram up till now :biggrin1: . In fact i'm not sure if he's Afram all the way although he considers himself "black", he's also on 23andme so i suppose his CoA results might provide some solid clues about any possible foreign ancestry from several generations ago beyond family recollection.

There's several things that stand out:

A whopping 13% North African! That's even higher than i've seen on this 1 PR guy who had 10%, the highest i've seen on any New Worlder
SE Bantu coming in at first spot, only happens very rarely for Aframs from what i've seen, it's not overly dominant though
His maternal haplo is M32c! Which is mostly seen among Malagassy if i'm not mistaken, quite rare but i've seen it reported for several other Aframs.
Not shown on the screenshot but in addition to 5% Amerindian he also has :

Asia 4%
Trace Regions 4%
Asia South 2%
Asia Central 1%
Asia East 1%

Pacific Islander 2%
Trace Regions 2%
Melanesia 2%

West Asia 2%
Trace Regions 2%
Middle East 2%

Europe 11%

Trace Regions 11%
Europe West 3%
Iberian Peninsula 3%
Italy/Greece 2%
Great Britain 2%
Ireland 1%



http://i989.photobucket.com/albums/af18/oditous2/M32cadna_zps7ad84355.jpg~original




Just speculating but the combination of his mtDNA, SE Bantu, minor Asia and even Pacific Islands (!) makes it eerily possible he has quite some significant Malagassy ancestry going on, inherited to a far greater degree than what i've seen on other Aframs who had reason to believe (mostly because of their maternal haplo being SE Asian) they had 1 single Malagassy ancestor in the distant past but somehow it got heavily dilluted for them and not for him. See also this excellent blog post:

Got Roots in Madagascar? (http://rootsrevealed.blogspot.nl/2014/06/got-roots-in-madagascar.html)

The North African is more mysterious and there could be several scenario's i suppose. The North African category on Ancestry seems to be more solidly "Berber" / "Maghrebi" than on 23andme from what i've seen from the Egyptian and Somalian results posted earlier. I suspect his Euro breakdown, which looks atypical for an AA who will be mostly if not exclusively NW European, will be related to some of those scenarios. His minimal Senegal & Mali make it seem less likely that it's derived from a Fula ancestor, but who knows.

Having one exotic lineage from Madagascar, Sahellian West Africa or even North Africa from many, many generations ago is always a possibility of course but it tends to be VERY much dilluted from what i've seen showing up in minimal %'s, often near the margins of error. Really the thing that amazes me with this guy is just the random combination of it all (remember he also shows above average Amerindian for an Afram) and particularly the ELEVATED level of his North African. I mean having 13% North African is pretty much unique for a New Worlder unless they have recent ancestry from those parts, dating from the POST-colonial era.


I also have his 23andme. The guy was complaining (https://www.youtube.com/watch?v=VnYLWWfdX-4) his AncestryDNA didn't correspond with 23andme but he used standard setting. On speculative it's bound to look more similar, already it looks quite unusual though for an Afram


http://i989.photobucket.com/albums/af18/oditous2/M32c23andme_zpsfdc9f3b2.jpg~original



**************************************************





Second atypical result for an Afram, not so much the SSA breakdown but the whopping 28% Amerindian! First time ever i'm seeing an Afram scoring more Amerindian than Euro, also on 23andme, almost zamboish:D This person actually has a fullblooded Native American grandmother from Utah so that accounts for 25%. In addition there's probably also some minor (3-5%) older Amerindian ancestry via her Afram grandparents. Even though for most Aframs their Native American %'s have been rather subdued contradicting family stories about significant Amerindian ancestry in the last 3 or 4 generations, this case shows it's not always like that.

Interesting quote by Prof. Gates


Why most black people aren’t “part Indian,” (http://www.theroot.com/articles/history/2014/04/why_most_black_people_aren_t_part_indian.1.html)




Again, here are the statistics: Whereas virtually all African Americans have a considerable amount of European ancestry in their genomes, only 19 percent have at least 1 percent Native American ancestry, and only 5 percent of African American people carry more than 2 percent Native American ancestry.

Despite these averages, however, some African Americans do have significant amounts of Native American ancestry, though almost no black American person today has as much Native American ancestry as they do European ancestry, by quite a long shot. (This does not include black people of Hispanic origin, in that Hispanic Americans tend to have far more Native American ancestry than African Americans do.)







http://i989.photobucket.com/albums/af18/oditous2/CW_zpscc31c053.jpg~original

Don Felipe
09-24-2014, 07:22 PM
Anglo-Caribbean results


A few Anglo-Caribbean results compared with other Afro-mixed groups. Of course only the AA sample size is big enough to be considered representative in any way, but you can already see some patterns that will be sustained i think even when i'll get more results in for the other nationalities. Anglo-Caribbean has 7 Jamaicans, 3 Belizeans, 3 Guyanese, 1 Trini, 1 from Montserrat and 1 mixed West Indian (several islands), Franco Caribbean is only 2 Haitians but might still be pretty representative. Hispanic Caribbeans is mostly PR's right now and a few Dominicans, things will look different if i get results for persons with higher SSA. Same goes for South Americans, which is mostly Colombians right now. North African is distorting things a bit for persons with lower SSA. If it wasn't included in total African you'd see even more pronounced Upper Guinean (Senegal + Mali) and SE Bantu in the SSA breakdown for Hispanics.


http://i989.photobucket.com/albums/af18/oditous2/AFrodiasporeansAC16_zps6fa0f2fb.png


Compared with the others i'm seeing Anglo-Caribbeans and AA's being predom. Lower Guinean (Ghana/Benin/Nigeria). Biggest difference between them being Benin/Togo, which is ranked #1 for Anglo-Caribbeans but only # 4 for AA's although it's still showing up strongly for AA's as well. Somewhat surprisingly it's especially dominant for Jamaicans i've seen up till now, ranked nr.1 for 5 out of 7 results. Besides the very real possibility (IMO) that it's (partially) genuinely Benin/Togo, meaning Gbe speaking groups like the Ewe and Fon there's 2 major caveats:



the Benin/Togo and Ghana/Iv. Coast categories cluster the closest from all other SSA categories on AncestryDNA so there could be some incorrect assignment between the two (see this plot (http://i989.photobucket.com/albums/af18/oditous2/African-Ancestry-2.jpg))
Additionally also Nigerians (even Igbo's!) might have carried over some ancient "Benin/Togo" component (originating within Nigeria but nowadays seen more frequently among people from Benin/Togo) into AA's, Jamaicans & other Anglo-Caribbeans. I've seen an Igbo result i will post in a separate thread (http://www.anthrogenica.com/showthread.php?3222-AncestryDNA-breakdown-of-SSA-for-Africans&p=53299&viewfull=1#post53299) who scores 17% "Benin/Togo"!



Senegal & SE Bantu is lowest among Anglo-Caribbeans for many even at very minimal or non-inexistent level. But for Latin Americans its among the biggest components corresponding with the earliest arrivals of Africans from Upper Guinea and Angola in the Americas in the 1500's/1600's. Given the low SSA of my samples i'm pretty sure things might be different when i get 50%+ SSA Latino results.

Cameroon/Congo might partially also be derived from Igbo's or related Southeastern Nigerians for Anglo-Caribbeans and AA's. This fully Igbo (http://www.anthrogenica.com/showthread.php?3222-AncestryDNA-breakdown-of-SSA-for-Africans&p=53299&viewfull=1#post53299) i've seen scores 24% Cameroon/Congo which is about half of his "Nigeria" score. But it's quite telling how the average Cameroon/Congo % stilll is slightly higher for AA's (who according to historical sources would indeed have more Central African ancestry than West Indians) than it is for Anglo-Caribbeans and also how the 2 Haitians score the highest average. Again according to historical slave trade records genuine Congolese ancestry would indeed be among the highest for the Haitians.

Don Felipe
09-24-2014, 08:04 PM
Some screenshots

Jamaicans

These two being very similar in the ranking of their top 3 but they're not related for all i know.


Jamaican 1
http://i989.photobucket.com/albums/af18/oditous2/JAM2_zpsdd22d943.jpg~original



Jamaican 2
http://i989.photobucket.com/albums/af18/oditous2/JAM1_zpsdf66d5ac.jpg~original


Belizeans

Kriol

http://i989.photobucket.com/albums/af18/oditous2/BelizeKriol_zpsa9f956c6.png




Kriol/Garifuna

http://i989.photobucket.com/albums/af18/oditous2/Kriofuna_zpsc859aa41.png



Garifuna, also notice his elevated Amerindian and ZERO Euro %!

http://i989.photobucket.com/albums/af18/oditous2/Garifuna_zps6f1cce10.png


Guyanese

High Ghana/Iv.Coast for the first one, more than 50% of his total African ancestry.

http://i989.photobucket.com/albums/af18/oditous2/GUY2_zps2976da07.jpg



High Mali for this one, it's the least reliable SSA category on AncestryDNA, so i guess it might as well signal ancestry from Sierra Leone/Guinea or even northern Ghana (Gur speaking groups)

http://i989.photobucket.com/albums/af18/oditous2/GUY1_zps4a8f86e0.jpg


Montserrat

First one actually also having partial roots in Barbados & Trinidad

http://i989.photobucket.com/albums/af18/oditous2/MontserratTriniBarbados_zps7c5f1470.jpg


This one having minimal Euro input, such results probably more frequent among Anglo-Caribbeans than for AA's, although both groups can show wideranging levels of admixture and i 've seen several nearly 100% SSA results for AA's as well.


http://i989.photobucket.com/albums/af18/oditous2/Montserrat_zps817292a2.jpg

Erik
09-24-2014, 09:36 PM
I've seen many African Americans show up with Pacific Islander. Why is that?

Don Felipe
09-24-2014, 11:09 PM
I've seen many African Americans show up with Pacific Islander. Why is that?

I wouldn't say it shows up for many AA's, it's just a few of them i've seen myself (out of over 200 results). Also it's almost always being shown as Trace region and almost always as <1%, making it pretty much a noise level component.

The only time i've seen someone who had it above 1% and not as Trace region was the "atypical AA" result i posted earlier who has 2% Melanesian. It might be real for him indeed or else signalling some other type of related ancestry not captured by the present categories. I've seen "Pacific Islander" %'s also for SE Asians and even South Asians.

Don Felipe
11-23-2014, 07:15 PM
Some more Jamaican & Belizean results,


Jamaica

Finally Ghana/Ivory Coast showing up as main region instead of Benin/Togo for two of them. I already came up with several possible reasons why this category sofar seemed to show up so consistently for Jamaicans, i would like to add another (speculative) possibility, that is that aside from Ewe also other non-Akan speaking ethnic groups living in Ghana might be contributing to the ancestral markers being picked up as "Benin/Togo" by AncestryDNA. Again the labelling with country names should NOT be taken too literally! It's all context dependent.

Especially the Ga (http://en.wikipedia.org/wiki/Ga-Adangbe_people) seem to fit the bill as they also live in Togo and are quite numerous (1.8 million according to wiki). The Guang (http://en.wikipedia.org/wiki/Guang_people) are also interesting as they are supposed to be the oldest inhabitants of Ghana. I suppose this could have made them especially vulnerable to enslavement after they got displaced by more dominant groups like the Akan. Here's a useful breakdown of ethnic groups in Ghana, of course it might have been different in earlier centuries. Akan constitute almost half of the population, Ewe + Ga combined are about 20%, and there's also several northern groups, i think they were generally named "Champa" in slave records. It might very well be that the selftermed "Ghana/IvoryCoast" cluster is only significantly predictive of Akan ancestry and less so for the other ethnic groups, more DNA research and ethnic sampling is needed to be sure though.


http://mecometer.com/image/piechart-country-data/ghana/ethnic-groups.png



First Jamaican has very cool results not showing ANY non-African admixture, 100% from the motherland B). There must be many like him still in Jamaica otherwise it would be very difficult to maintain such a "pure" lineage. Afterall having offspring with a 90% SSA Jamaican would already get in the way of preserving this 100% score.

http://i989.photobucket.com/albums/af18/oditous2/JAM3_zps7438a1bb.jpg~original



This one also has a quite high SSA level and again showing Ghana/Ivory Coast as main region, compared with the previous Jamaican results showing Benin/Togo i suppose we could say they might all still be pointing to ancestral connections from Ghana but an Akan ancestor is MUCH more likely for this result, while for the others it might be non-Akan ethnicities from Ghana (aside from other possible scenarios).


http://i989.photobucket.com/albums/af18/oditous2/JAM2_zps6401e028.jpg~original




These two Jamaicans showing Benin/Togo as main region. First one VERY convincingly even, it's almost 80% (32/41) of total African ancestry! The second one is actually from Panama but with mostly Jamaican born grandparents.


http://i989.photobucket.com/albums/af18/oditous2/JAM1_zpsf8adf932.png~original



http://i989.photobucket.com/albums/af18/oditous2/JAM1_zps75894c20.jpg~original





Belize

In their African breakdown they follow the Jamaicans quite closely sofar. Focus being on Lower Guinea (Ghana-Nigeria) and also Cameroon/Congo. Senegal being almost absent and Mali & SE Bantu quite low. This was to be expected as most Africans in Belize arrived via Jamaica. The main distinction is more so in the non-SSA admixture, especially the Amerindian scores.


http://i989.photobucket.com/albums/af18/oditous2/BELIZE2_zpse0f88e94.png~original


Very high Ghana/Ivory Coast for this one, over 65% (46/68) of total African ancestry


http://i989.photobucket.com/albums/af18/oditous2/BELIZE3_zpsea92fc80.png~original


http://i989.photobucket.com/albums/af18/oditous2/BELIZE1_zps05bf53af.png~original

krand
11-24-2014, 04:00 PM
This is very interesting to say the least. My late great grandfather was from Jamaica but said to have come from Africa - a Nago (Yoruba). I could only imagine what his dna results would've been. I hope that Ancestry makes a more consorted effort to increase its African based reference panel. This would tighten up the possibility to make even more precise predictions relative ethnic connections.

Don Felipe
11-25-2014, 10:59 AM
My late great grandfather was from Jamaica but said to have come from Africa - a Nago (Yoruba). I could only imagine what his dna results would've been. I hope that Ancestry makes a more consorted effort to increase its African based reference panel. This would tighten up the possibility to make even more precise predictions relative ethnic connections.

Yes that would be very intriguing to know about your great grandfather's results. It's interesting you yourself show only very little Nigeria while your main region is Benin/Togo. I have a strong feeling many Yoruba's taking this same AncestryDNA test will in fact also score significant "Benin/Togo". Afterall Benin itself has a Yoruba minority. Sofar i've only seen the results of an Igbo Nigerian (see this post (http://www.anthrogenica.com/showthread.php?3222-AncestryDNA-breakdown-of-SSA-for-Africans&p=53299&viewfull=1#post53299)), he also scored quite a lot of Benin/Togo aside from Cameroon/Congo, Nigeria was still his main region but only barely so with 49%. I'm guessing in his case "Benin/Togo" (perhaps better thought of as component X) is picking up on ancient southern Nigerian ancestral markers that nowadays are more frequently seen in Benin/Togo because of prehistorical migrations.

Taking these category names too literally can be misleading therefore, ancestral clusters don't follow artificial borders and also each African ethnicity itself will be mixed and showing different ancestral components as there's no such thing as a "pure" bloodline if you go back far enough in time. This kind of regional resolution for Africa is very much in its infancy stage and will undoubtedly be improved in the future. Still given widespread ethnic intermixing and genetic similarities across borders even then it will be inherently limited in assigning ethnic ancestry within Africa with 100% accuracy. If you keep this in mind and lower your expectations you can get more out of your personal results IMO. Right now it might be too soon to zoom in on a more detailed ethnic/localized level but by consulting historical evidence of slave trade and making comparisons with other Afrodescended nationalities you can get a more plausible interpretation.


For example your "Benin/Togo" score might very well be signalling a different type of ethnic connection than it would be doing for a Haitian. I already discussed on the previous page how this category is somewhat overlapping with Ghana/Ivory Coast and might possibly also be picking up on non-Akan ethnic ancestry from Ghana. Given slave trade statistics and cultural retention it would be much more likely that for a Haitian instead it will indeed be genuinely from Benin and ethnically it could be either Fon or Aja. Then again several scenarios can be possible and also would not be mutually exclusive. We all have unique familytrees after all and there's always individual variation.


Here's a new result of a Haitian


Haiti


http://i989.photobucket.com/albums/af18/oditous2/HT_zpse26e251d.jpg~original

krand
11-25-2014, 03:26 PM
This is very interesting to say the least. My late great grandfather was from Jamaica but said to have come from Africa - a Nago (Yoruba). I could only imagine what his dna results would've been. I hope that Ancestry makes a more consorted effort to increase its African based reference panel. This would tighten up the possibility to make even more precise predictions relative ethnic connections.

At first glance, i initially thought similarly - Benin/Togo but lacking Nigeria??? I had to remember though that (1) the Fon referred to "Yorubas" as Anagos (or Nagos). And more importantly, (2) the Fon kingdom was a long standing rival/tributary/vassal to the powerful Oyo Kingdom. Once the situation turned, the Fon didn't hast to capture and sell hundreds of thousands of Nagos (Yorubas) into slavery. Even according to Fon oral history, they (Fon) actually splintered from the group in Ile Ife (spiritual center of traditional Yoruba) and migrated Southwest and founded Allada. This would, i suspect, explain why the Fon (as well as the Ewe) and Yoruba possess near identical divination systems - Ifa (Yoruba) and Fa (Fon) and Afa (Ewe). Plus they both have direct and/or share corresponding deities (Orishas & Lowas). The Yoruba are known in Brazil as Anago; in Cuba as Anago; in Hayti as Nago; in Jamaica as Nago (there was also a town in Westmoreland named as such); and in Louisiana and South Carolina as Nago. So the Dahomean (Fon) designation of Yoruba (as Nago) was very wide-spread. It should be also noted that one of the last slave ships to arrive to the United States (Mobile, Alabama in 1858), the Clotide, debarked from Allada with many enslaved Yorubas (Nagos); they later established AfricaTown in Plateau, Alabama - just North of Mobile. Many of their descendants still live there to this day! Another important factor is that among the Gullah/Geechee of the Carolinas, they retain hundreds of Yoruba names (especially of deities/Orishas) and words as late as the 1940s. I conjecture that might be attributed to the thousands that were smuggled in (like the Clotide) after the prohibition of the International Slave Trade (1808). This is why, I believe, such records are absent in works like Phillip Curtin and the Transatlantic Slave Trade Database, for instance. I say this because there was also SUBSTANTIAL indigenous born African population in South Carolina in 1822. During the Denmark Vessey plot, many African were organized according to ethnic groups - Igbos, Mande, and etc. The Yorubas (Nagos) were said to have been populous on Johns Island (in South Carolina).

Don Felipe
11-26-2014, 11:13 AM
Yes i agree there's many similarities to be found between the Fon and the Yoruba, not only culturally but undoubtedly also genetically to some degree. This does complicate the exact matching with just one ethnicity as most of us would like to see it. To be honest i'm not sure if this will ever be possible even in the future. The same problem of genetically delineating ethnicities has also occurred for Europeans and it seems to be inherent to this kind of BGA ancestry testing. Finding DNA cousins from a certain ethnic group with the shared DNA segment size being significantly large might be more telling to confirm ethnic connections i suspect.

About the relative importance of Yoruba's in the USA, i do agree there's indeed documented evidence of their presence. But compared to other destinations in the Americas they seem to have been very minor in the US. Especially compared with imports from the Bight of Biafra (bringing in especially Igbo and related people) the direct slave trade between the US and the Bight of Benin was much less important. According to most estimates it would be below 5%, so that's why i've been suspicious of the relatively high Benin/Togo averages for AA's seen from the start.

I do agree that the slave trade records might not be 100% complete. But for example the Clotilda voyage can be found in the slavevoyages database and is included in their overall estimates. If you do a search for all voyages to the USA and sort according to date you should find 24 slave voyages made after 1808, three of them departing from Bight of Benin. At any rate it is known that the US volume of slave trade was highest in the 1700's because unlike Brazil and Cuba the American slave population had a positive growth rate so there was less need for additional imports when the slave trade was made illegal.

Here's a chart with estimated Yoruba's being brought to various parts of the Americas. It's in Portuguese, but you can see the numbers for the US in the first column and how they compare with other destinations. Out of an estimated total of 968.000 enslaved Yoruba's for the whole period less than 5000 are said to have arrived in the US.

http://i989.photobucket.com/albums/af18/oditous2/YorubasinAmericasnon-hispano_zpsbec2cbad.jpg~original


http://i989.photobucket.com/albums/af18/oditous2/YorubasinAmericashispano_zps0b2910d3.jpg~original

awodwaRSA
12-11-2014, 11:49 AM
Southern West Indies/Windward Islands(Barbados,T&T,Grenada etc..) was dominantly Bight of Biafra and Gold Coast/Windward coast. While western West Indies (Cuba,Haiti,Domicnican Rep etc. excluding Jamaica) was dominantly Angola-Congo and Senegambia.
3183

AJL
12-11-2014, 03:21 PM
awowdaRSA, I have moderated most of your posts for a number of violations of forum policy including




3.10 Certain standards of quality-control will be enforced to ensure a productive forum atmosphere. Breaches in basic forum etiquette include but are not limited to cross-posting different threads, consecutive posting in existing threads, or reviving old threads whose course of discussion has long since expired ("necro-bumping"). If a user feels they may have made such a mistakes, they are expected to contact a moderator for assistance.

3.11 Invectives and posts devoid of substance (e.g. threads or replies consisting solely of inflammatory content or triviality) will be considered junk postings and deleted. Anthrogenica encourages its members to engage in debate with each other, but discussions that degenerate into little more than flaming and inanity will be considered spam (see Rule 3.9).

3.12 Anthrogenica is a forum for free and honest discussion. It is not a pulpit to proselytize or to promote certain viewpoints that would be better-suited to a personal blog. Such behaviour is prohibited. What constitutes the aforementioned, and whether infraction delivery is warranted, will be determined at the administration's discretion.

[...]

4.1 Anthrogenica only permits the posting of open-source/freely-accessible publications that have been properly attributed to their authors.

4.2 To avoid future disputes in ownership members of Anthrogenica are advised to personalise or sign off any original pieces they may create, including but not limited to text, diagrams, maps, pictures and charts. Otherwise, in the event of a dispute over ownership, Anthrogenica is not obligated to support the claim made by its' user-owners on unsigned materials.

4.3 Illegal file sharing is strictly prohibited on Anthrogenica and any illegally shared files will immediately be removed as per the conditions agreed by Anthrogenica's website host and our continued support for current online legislation(s).

Further violations will result in the actions set out in Section 6 of the guidelines I linked to above. Please read and familiarize yourself with the Terms of Service.

JFWinstone
09-05-2016, 09:16 PM
Very interesting thread :D My mum is 1/2 Mauritian and 1/2 British and we have a few matches with people with known links to either Mauritius or Madagascar

11415
Her ethnicity estimate

11416
Mum's ancestryDNA map

bryan
06-07-2017, 11:16 AM
The ethnic breakdown would make more sense if they took reference populations such as Yoruba, Igbo, Fulani so we could see where in exactly our roots came from, because back in the trans Atlantic slave trade places as Nigeria didn't exist yet, but various tribes lived in the mentioned region and the genetic difference between a Fulani Nigerian and a Fulani living in Chad would be far less than between a a Fulani Nigerian and a Yoruba.

Georg
09-10-2017, 09:05 AM
Does anyone know what Paleo-African means? Does it mean someone, let's say from Norway during the paleolithic age, who happened to stumble into Africa by taking a wrong detour?

Don Felipe
06-16-2018, 11:41 PM
The ethnic breakdown would make more sense if they took reference populations such as Yoruba, Igbo, Fulani so we could see where in exactly our roots came from

This is something many people wish for, but i'm afraid that it's not as easy as it sounds. Due to shared origins from ancient times as well as more recent migrations and increasing inter-ethnic unions in modern times there will always be a great deal of genetic similarity for neighbouring ethnic groups and at times also for ethnic groups who are living in geographically distant countries. I am not aware of any DNA testing company which is able to adequately tackle this issue yet.

Frankly I prefer a sketchy regional breakdown above an overspecified breakdown which gives false hope about pinpointing specific lineage.

Don Felipe
06-16-2018, 11:44 PM
Final findings of the survey I have been doing on AncestryDNA results of various Afro-descended nationalities. Read blog posts below for more details:

Update: Afro-Diasporan AncestryDNA Survey (part 1) (https://tracingafricanroots.wordpress.com/2018/06/11/update-afro-diaporan-ancestrydna-survey-part-1/)
Update: Afro-Diasporan AncestryDNA Survey (part 2) (Update: Afro-Diaporan AncestryDNA Survey (part 2))



https://tracingafricanroots.files.wordpress.com/2018/06/africans-diasporans1.png