African dna

[cokie]

On Dr. McDonald's test he found about 8-10 Mb of African dna which also shows up in several other tests such as Eurogenes and Dodecad etc., in very small quantities, usually less than 2%. It's always present in any of the tests I take except on FTDNA there was no mention of it, however I have 10% middle-eastern dna from FTDNA testing which very likely is from my paternal, Ukrainian side, and some matches with Armenians, Ashkenazi, and vague others throughout the middle-east also likely from my Ukrainian paternal ancestry.
It may be the African dna came from North America where my ancestors from my maternal side (NW Europe) have been both in New England and New France since the early 1600's.
My question is - is this African dna recent, in the last 6 generations or can it be very ancient. I haven't any matches on the family finder on this segment where Dr. McDonald found it -chromosome 20 from about 45-55.

[MFWare]

It would be much easier to answer your question if you were to share your GEDmatch.com kit number.

[cokie]

gedmatch = f209848

[mixedkid]

gedmatch = f209848

Cokie, that is a relatively large segment of African (actually, West African) DNA on chromosome #20. You also have smaller segments of various kinds of African DNA scattered throughout your other chromsomes, including that from East Africa. I used several calculators to check your DNA, especially that segment of chromosome #20. (I even used the J-test which picks up African DNA rather well, and chromosome #20's African segment shows up right away.)

Also using several calculators on that site, you have small segments of Native American (American or Amerindian) scattered throughout your chromosomes.

In answer to your question: My guess is that your African ancestry (at least where that large segment is concerned) is fairly recent, most likely post-colonial/post-American Revolution. French Colonial or British Colonial might help explain the smaller segments of both Native American and African. You appear to have a combined genetic ancestry that is largely Western/Northern European (colonial, probably from your known ancestry) + post-colonial European immigrant groups + some recent African (West) + distant Native American + distant African (both West and East) from colonial times.

[MFWare]

gedmatch = f209848

I have looked long and hard at various admix model runs of your heritage. You have no hint of African ancestry above noise level. One of your more appropriate models is the MDLP World-22. Of its 22 reference populations, 13 are nulled-out, three are no higher than noise level, leaving six as significant. This model includes three African reference populations. Two are nulled-out; and the third, below noise level.

There was two other models that find only five populations are necessary to represent your admixture. None have African reference populations above noise level.

Each model appears to say essentially the same thing. Your admixture is approximately 55% Northern European and 25-30% Mediterranean. I have looked at the chromosomal paints generated by several of the models. With due respect to mixedkid, none show African reference populations on Chromosome 20 above noise level, if that much. To the contrary, each is an additional representation of the admixture pie charts. The paint of Chromosome 20 shows Mediterranean reference populations with virtually no African contribution.

[mixedkid]

Cokie, I very much caution you not to pay too much attention to the overuse of the term "noise level" on this forum. I stand by my original analysis. You do have an African chromosome segment of quite unexpected size on chromsome #20.

I have said it once before and I will say it again: Oftentimes the so-called experts who analyze these things are ignorant of settlement history, especially the colonial settlement history of North America. "Noise" is s term they use for something they do not understand or, is often the case, to deny the true genetic makeup of Americans and Canadians; in other words, they have found a way to deny their own true ancestral backgrounds, no matter how distant.

That large segment on chromosome #20 is unusual. It is unexpected from what you tell us. My guess is that one of your more recent ancestors actually had partial African ancestry and did not even realize it. If you have some ancestral connection to the American south or to old Acadia or Quebec, that might help to explain it.

[cokie]

Thanks for your responses but now I'm really confused. Is it or isn't it?
In all the various tests I've done myself the African and Asian always appear in small amounts - if it's noise then why doesn't noise from other groups show up such as South American or Australia - both of which don't ever appear in my dna? So why just noise from a few groups and why always there?
I can see lots of potential for both African or Native American ancestry, for my Quebecois ancestors were living very close with Natives as well as the New Englanders with Africans - the Dutch Church of New Amsterdam for example - in the earliest records my ancestors are present alongside Africans who were also baptized in this church.
I would appreciate more analysis of this if possible, as it's of course very intriguing to me. If I could find match on any of these chromosomes with proven African or Native ancestors I suppsoe that would be very telling...

[tomcat]

Thanks for your responses but now I'm really confused. Is it or isn't it? ...

Clearly, opinions differ as to reality of various readings from various available calculators.

You could try another test.

23&Me's v3 samples up to a million SNP's from across the genome (a more extensive test than FTDNA) and they are close to launching an improved Ancestry Painting, their admixture tool.

Although details are lacking, AncestryDNA seems to be on a par with FTDNA's coverage, but they preliminarily phase data through reference to population samples, to get an improved admixture estimate.

Genographic 2.0 offers the least coverage of any retail scan but their chip is programmed to respond to selected ancestry-informative SNP's, individual mutations that have a marked incidence in one or another population.

[Yaffa]

Cokie,

did mcDonald give you an actual % on African? if he did not report % it may be noise or back there beyond paper.

One of my family members McDonald gave 1.5 % African.. On paper it's a 2nd great grandfather to the person tested. This 2nd great grandfather was African/American Indian mixed though I cant tell you how far back that 1.5 % African goes because my paper stops at the 2nd great. The part African is coming form his unk father line according to paper.

[cokie]

Dr. McDonald sent me a chart of the chromosomes and the Afrcican dna was 50 mb on chr. 20. All the rest was European, no meditteranean or anything else though FTDNA has me as 10% meditteranean.

[Yaffa]

cokie, besides the charts usually McDonald will give you % break down in the email he sends you with his charts. I have things on my chart that is just noise or too far back for %. Check the email you got form him

[tomcat]

...
I can see lots of potential for both African or Native American ancestry, for my Quebecois ancestors were living very close with Natives as well as the New Englanders with Africans - the Dutch Church of New Amsterdam for example - in the earliest records my ancestors are present alongside Africans who were also baptized in this church.
...

Does this mean that you see potential, in the paper, for African ancestry on both parental lines?

If yes, there is the possibility, that the 50mb segment on Chr 20, is, in fact, two or more smaller segments, coming from both parental Chr 20's, that read as one because they are unphased.

[cokie]

The email I got from Dr. McDonald gave 100% Europe and a "tiny" bit of African which he said was "very old".
There's more potential on my maternal side (NW European) for African or NA but very little on my paternal, Ukrainian side given the history of the Ukrainian region. Although the 10% Middle-eastern from FTDNA is probably from my paternal side as I have Ashkenazi and Armenian matches, perhaps the African is from there?
This is what the population finder on FTDNA says for middle-eastern..
"Palestinian, Adygei, Bedouin, Bedouin South, Druze, Iranian, Jewish"

[Javelin]

Although the 10% Middle-eastern from FTDNA is probably from my paternal side as I have Ashkenazi and Armenian matches, perhaps the African is from there?

Not likely.

[Yaffa]

The email I got from Dr. McDonald gave 100% Europe and a "tiny" bit of African which he said was "very old".
There's more potential on my maternal side (NW European) for African or NA but very little on my paternal, Ukrainian side given the history of the Ukrainian region. Although the 10% Middle-eastern from FTDNA is probably from my paternal side as I have Ashkenazi and Armenian matches, perhaps the African is from there?
This is what the population finder on FTDNA says for middle-eastern..
"Palestinian, Adygei, Bedouin, Bedouin South, Druze, Iranian, Jewish"

if he said your African is very old and you have haven't found it on paper than it probably is old before paper.

I would not rely on population finder at the moment. They are redoing it because they are giving people Middle Eastern that don't have any. You could ave American Indian and It's not readable because they don't have samples. Population Finder is still in beta