Ancestry By DNA results

[belovedwisdom]

20% Native American Ancestry

I received my results of the Ancestry By DNA test last week. The results say that I am 80% Sub-Saharan African, 20% Native American. Since both percentages are greater than 10%, can I assume that this is fairly accurate? I had suspected for years that I had Native American ancestry (my maternal grandmother has told me stories about a great great grandmother whose ancestry might have been Native American), but I didn’t expect the percentage to be that high. My MtDNA results put me in the L0 haplogroup, which is not Native American. Perhaps there is Native American ancestry on my father’s side as well.

[burto]

I noticed on the Trace Genetics site that the ABDNA 2.5 test is marketed as showing your ancestry within the last 5 generations, so if that is the case then one of Mum's Great Grandparents must have been of East Asian (I use that loosely ) descent OR I guess two of them could have been mixed to get 11%.

[tomcat]

If your mother got 11% EA then her parents - your grandparents - had 22% EA and her grandparents - your great grandparents - had 44% EA.

However -

All those %'s are merely nominal. Any ancestor might have had the EA score while the other had none. Or any pair of ancestors may have shared EA heritage. And the score does not tell you when EA entered your family.

[burto]

So one parent can't be responsible for the whole 11% then? I know that Mum's Mum's history is firmly rooted in England so don't see how it can come from her. The ABDNA literature also says that it is not found in Western Europeans. So it must be from her Dad?

[tomcat]

The EA could have a single or multiple sources. Tribes blog archive has item stating they see evidence of an ancient EA signature in western Europe including the UK.

[burto]

So what exactly is the point of these tests then? I keep saying this because you think you've found something relevant then other people tell you it could mean A, B, C or D it's just not good enough in my opinion.
Tribes say this, ABDNA say something else. Who do we trust? I have heard that ABDNA's methods use more stable "ancient" markers whereas Tribes use ones that have only recently have come about and are more generic. If this is the case then I will take ABDNA as more valid. There are too many weird Tribes scores and even Omnipop and RCMP results are different to theirs. ENFSI is bizarre too....the British Isles are at the bottom of my list which is just inaccurate. The only pattern I can find with Omnipop and RCMP is white with approx. 15% Native Americn or Hispanic, but how accurate is that?
It has left my Mum and myself feeling totally in the air and confused, we are no better off.

[GregKiroKH2]

With admixed individuals, results seem to depend upon with which markers are used for the readings.

[burto]

With admixed individuals, results seem to depend upon with which markers are used for the readings.

How do you know which ethnic groups are dependent on which markers? Are there patterns with certain ones? Are there particular marker scores that are strongest in Europeans and Native Americans (in fact every ethnic group so others can look for patterns too)?

[GregKiroKH2]

How do you know which ethnic groups are dependent on which markers? Are there patterns with certain ones? Are there particular marker scores that are strongest in Europeans and Native Americans (in fact every ethnic group so others can look for patterns too)?

I am unsure of specific details.

My first assumption is to use the laws of genetic segregation to follow each allele to a population.

Since most people married within their population, there would be specific locus characteristics of a population.

Admixed people should have some locus of one population, some locus of another population, and some of both populations.

When testing a restricted set of alleles, admixed populations ought to produce results that do not match with another test of a different restricted set of alleles.

And this is what I see with my data, and the few other sets I have looked at.

Population 1 -AA (A includes group mutations)
Population 2 -BB (B includes group mutations)
Admixed Population
AP Group 1 - AA
AP Group 2 - BB
AP Group 3 - AB
AP Group 4 - BA
AP Group 5 - AM
AP Group 6 - MA
AP Group 7 - BM
AP Group 8 - MB
AP Group 9 - MM (M includes new mutations)