Page 1 of 1

Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Fri Jun 28, 2013 5:42 pm
by Mehrdad
I am trying to understand the significance of ancient autosomal DNA vs recent. My admixture from 23andMe, is fairly well represented by what I know my family background to be.

South Asian - 40%
Middle East - 4%
European 1%
East Asian - 12%
Oceania - 35%
Un-Assigned - 8%

So I wonder how far back can autosomal results track our ancestry? I am assuming if I had an ancestor from Africa 1000 years ago, assuming the person's spouse was never African would I be able to catch that? I've seen this with my Spouse, she has very little African DNA showing in her autosomal around .3%, however her father is showing 1% and I am assuming her grandfather once we get his back will be a lot more significant. And the African ancestry is only from the 1700's so I am assuming anything past 300 ybp will not show up?

So how do we capture deep autosomals without counting on Y and mt DNA's, can chromosome comparisons be used instead? I am fairly new to this, so I am just trying to wrap my mind around it.

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Sat Jul 13, 2013 7:53 pm
by J-L858
i dont get it

my result in family finder shows that i have 8% indian varius subcontint

i canot track any indian hertige in my family so i canot know if it recent or ancient

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Sun Jul 14, 2013 3:41 pm
by Mehrdad
J-L858 wrote:i dont get it

my result in family finder shows that i have 8% indian varius subcontint

i canot track any indian hertige in my family so i canot know if it recent or ancient


That is very interesting, have you tried using the calculators on GedMatch to see what your percentages could be? At least authenticate any Indian heritage, also if its 8% I would say that's between the late 1800's - early 1900's, approx 100 years +- 20 years.

Also FTDNA doesn't do a good job in showing admixture on chromosomes like 23andMe does. It would be interesting to see if there's only one source in your Indian heritage.

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Mon Jul 15, 2013 1:43 am
by leonardo
Based upon an analysis that I had done by a respected person in the field, I believe 8% may be more like 200 years back. :)

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Mon Jul 15, 2013 1:59 am
by J-L858
Mehrdad wrote:
J-L858 wrote:i dont get it

my result in family finder shows that i have 8% indian varius subcontint

i canot track any indian hertige in my family so i canot know if it recent or ancient


That is very interesting, have you tried using the calculators on GedMatch to see what your percentages could be? At least authenticate any Indian heritage, also if its 8% I would say that's between the late 1800's - early 1900's, approx 100 years +- 20 years.

Also FTDNA doesn't do a good job in showing admixture on chromosomes like 23andMe does. It would be interesting to see if there's only one source in your Indian heritage.


I used every calculators its always from 3%to 8% so sure there is some

McDonald J Douglas 95.7% total Mideast (32.5% Mideast (all Bedouin)/63.3% Mideast (all Druze/Palestinian)) and 4.3% S. Asia (various subcontinents)

it is on chromosome 12 acording to McDonald J Douglas

Image


my 1st cousin orderd ff so i can see if it is from dad side

i am j1c3d and also j1b2 mtdna

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Thu Aug 22, 2013 3:35 pm
by ChrisR
Mehrdad wrote:So I wonder how far back can autosomal results track our ancestry? I am assuming if I had an ancestor from Africa 1000 years ago, assuming the person's spouse was never African would I be able to catch that?
Very improbable.
I haven't seen recent and paper publicated research, but this blog post is very good:
http://www.genetic-inference.co.uk/blog ... e-our-dna/
The number of genetic ancestors starts off growing exponentially, but eventually flattens out to around 125 (at 10 generations, 120 of your 1024 genealogical ancestors are genetic ancestors).

If Luke Jostins is correct with his calculations, your DNA holds only a little part of the ancient genealogical ancestry. Going back 7 Generations you already miss half of the ancestors in your DNA (given there is no implex) and shortly after your genetic heritage becomes a minority fraction of the real genealogical ancestry and diversity.

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Fri Aug 23, 2013 3:33 am
by Mehrdad
ChrisR wrote:
Mehrdad wrote:So I wonder how far back can autosomal results track our ancestry? I am assuming if I had an ancestor from Africa 1000 years ago, assuming the person's spouse was never African would I be able to catch that?
Very improbable.
I haven't seen recent and paper publicated research, but this blog post is very good:
http://www.genetic-inference.co.uk/blog ... e-our-dna/
The number of genetic ancestors starts off growing exponentially, but eventually flattens out to around 125 (at 10 generations, 120 of your 1024 genealogical ancestors are genetic ancestors).

If Luke Jostins is correct with his calculations, your DNA holds only a little part of the ancient genealogical ancestry. Going back 7 Generations you already miss half of the ancestors in your DNA (given there is no implex) and shortly after your genetic heritage becomes a minority fraction of the real genealogical ancestry and diversity.


Thanks ChrisR, I really appreciate this, btw - my wife's autosomal test from 23andMe showed that she is around 20% Scandinavian, and she has two grandparents from both sides that are from Sweden. But she is also showing some African, her father is over 1% of African. How long ago would this be? Colonial?! There's Native American from both sides and it's showing up on the DNA tests as this is reflected on their genealogy, but the African is not showing up anywhere on their genealogy so no one knows where it came from. But if its 1%, I am thinking that it probably is around the colonial period. Any thoughts?

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Fri Aug 23, 2013 9:34 am
by stoeni
in regards to ancient admixtures
Would being 100% european ( tested by doug and 23andme ) indicate late bronze-age start point in europe since both tests only go back for 2100 years?

Re: Autosomal DNA: Recent vs Ancient Admixtures

PostPosted: Fri Aug 23, 2013 4:25 pm
by ChrisR
Mehrdad wrote:But if its 1%, I am thinking that it probably is around the colonial period. Any thoughts?

DNA pieces have no timestamp. The size of the "unique" piece can tell about the age, when the "breaking" trough recombination did begin. This is a criteria for admixture calculators: if they consider only pieces beginning from a certain lenght, then they focus on ancestry of the predicted age for that DNA lenght. Changing that criteria you get different reference clusters. I confirm this topic is seldomly detailed in discussions, but it is a great difference if admixture calculators focus on ancestry 500 or 1000 or 2000 years or even older.

23andMe claims to focus on ancestry 500 years old. If this is correct and Luke Justins calculations are near the reality you can say about the 1% African: 1.25 ancestor/s did had that DNA piece 17 generations ago (500 years / 30 years per gen.) and was of the predicted reference cluster (population). N.B.: 17 generations before us we have 131,072 ancestors, but as desribed above we can analyze only the DNA pieces of ca. 125 of those ancestors, that is still present in us. By testing your parents you double this: 250 and by testing grandparents 4x: 500 genetic ancestors and their DNA pieces.
Mind that this are statistical averages and the real values can vary: already between men and women there is a difference, because a women has two X-chr. and therefore sligthly more DNA.