Tribes scores

[Noaide]

Profile matching (Omnipop, Tribes, ENFS, CSFS) is not getting-at the information people want. Admixture mapping with STR's is not terrific and many more markers would be needed to get anywhere with admixture. We need another approach.

So far ABDNA 2.5 has been the best to guess that I am actually a European, possibly that test is the most robust by my experience.

Noaide

[burto]

Hi,
Well I'm really confused since Mum had the update! For a start the wrong set of alleles was sent and then with correction they said that the results were correct first time even if the panel wasn't. So don't feel overally confident about that...are these actually her matches?!
She got:
World match:
Med 18.10
NW European 13.21
E European 7.90
Asia Minor 6.22
Mestizo 4.22
India 3.79
Arabian 3.42
Indian Tribal 0.47
North India 0.42
SSAF 0.31

Polish matches have gone, Buenos Aires is still top, loads of Indian and Italian matches now.
Known ancestry is half English back to at least 1700 and half unknown American, so now even the "Germanic" matches have gone
DNAP results 65% NW European 13% Med 22% SA and 11% EA which doesn't tally with Tribes.
Anyone else having problems comparing to known ancestry?

[burto]

So far ABDNA 2.5 has been the best to guess that I am actually a European, possibly that test is the most robust by my experience.

Noaide

Starting to think the same thing.

[burto]

Everyone on this thread, or any of the many autosomal threads running on this forum, ought to read that rootsweb thread.

Profile matching (Omnipop, Tribes, ENFS, CSFS) is not getting-at the information people want. Admixture mapping with STR's is not terrific and many more markers would be needed to get anywhere with admixture. We need another approach.

So would DNAP do things differently?

[tomcat]

There is some discussion at the Rootsweb DNA site about FTDNA's possible intent to have an interactive for the autosomal markers they have tested.
They seem to feel that OMIPOP and possibly ENSFI is not the way to go for population studies.

What is Cluster software ? Is that what is used at DNA-FP for the XSTR markers?

Didn't get any sense from the Rootsweb exchange that FTDNA was going to do anything in support of atSTR testing beyond offering the panels. They don't even want to post a hot link to Omnipop.

Only software mentioned was STRUCTURE a model-based clustering program that was successfully employed in study to sort individuals into ethnicities.(see Bamshad et al 2003 Am.J.Hum.Gen.)

[jaranta]

DNA Print?

Burto,
Has your mother taken the DNA Print test? If so, did you see any relationship between her DNA Print results and her Tribes results?
Judy

Hi,
Well I'm really confused since Mum had the update! For a start the wrong set of alleles was sent and then with correction they said that the results were correct first time even if the panel wasn't. So don't feel overally confident about that...are these actually her matches?!
She got:
World match:
Med 18.10
NW European 13.21
E European 7.90
Asia Minor 6.22
Mestizo 4.22
India 3.79
Arabian 3.42
Indian Tribal 0.47
North India 0.42
SSAF 0.31

Polish matches have gone, Buenos Aires is still top, loads of Indian and Italian matches now.
Known ancestry is half English back to at least 1700 and half unknown American, so now even the "Germanic" matches have gone
DNAP results 65% NW European 13% Med 22% SA and 11% EA which doesn't tally with Tribes.
Anyone else having problems comparing to known ancestry?

[Nagelfar]

Hi,
Anyone else having problems comparing to known ancestry?

Yes. A whole lot, There's me who gets Hispanic & African from OmniPop & ENFSI, and Australian Aborigine & East Indian from DNAtribes. (I'm German - surname; Theil.. Scottish - surnames; Rankin & McHenry (possibly English in origin).. English - surnames; Clark, Pagit, Carter, Fisher, Boone, Hammond, Whitaker.. French - surname; Sommer (possibly German in origin).. Irish - surname; Ferrell.. .... With family pictures of just about everybody)

There's also John who is nearly 100% Dutch, and whose closest match comes back as being Nepalese.

[burto]

Burto,
Has your mother taken the DNA Print test? If so, did you see any relationship between her DNA Print results and her Tribes results?
Judy

Hi Jaranta,
Yes she did she got on the overall one 89% European and 11% East Asian.
For the Euro, she got 65% NW European, 13% Med and 22% South Asian.
So there is a conflict in the Med scores from each test.
Tribes have said that there could be a small amount of Native American based on her top Buenos Aires score and Mestizo match which have been consistent throughout), together with a Brazil match,which I guess tally's with the East Asian, but Tribes say "nothing substantial" which suppose the 11% is too.
There are a lot of Indian matches and her world matches for that area fall above ordinary, but below very high and I guess that tally's with DNAP's SA, but the Med matches just threw me...in the last update she had loads of Polish matches which we thought were Germanic as she's half English and now they've gone!
Her English side has no Med links at all, back to 1700.
Her unknown US father's surname was Johnson, so even if his Mum was Spanish surely the Med score shouldn't be that high?!
I can understand the EA scores and the Mestizo/Latin American scores coming from her Dad, but I was expecting at least some British Isles matches!

What I forgot to mention last time also, is she has a really low Finno-Ugrian score (0.22) and always has...what does this mean? I thought people from the British Isles would have an average score?

[tomcat]

Omnipop for Jews

Anyone looking into Jewish ancestry via Omnipop and the 9 CODIS STR sample would be well-advised to get a .pdf of the paper from which the Omnipop data is derived as Omnipop frequencies are from the Pooled Data. The individual frequencies for Ashkenazi, Oriental, North African and Sephardic Jews can be somewhat different from one another - enough to make a difference.

"Jewish population genetic data in 20 polymorphic loci"
Antonia Picornell, Carmen Tomas, Gema Jimenez, Jose? A. Castro,
M. Misericordia Ramon. For.Sci.Int. 2002

[Sonia]

Back to the drawing board

I beginning to agree here after a month or so of research with 3 siblings, nothing really looks conclusive and the bit about the poplation sizes skews everything.

Is it just me or do there seem to be an unusual amount of matches with populations in Spain? I've seen more than a few results from people and it's amazing how many Spain, Basque, Northern Spain matches there are.... If I had to quess, the population size is unusually large for DNATribes.

Cheers

I have noticed some clustering in the Spain / Portugal area. I have been following the Rootsweb thread as well. Also, my last method of looking at alleles individually and then weighing them, isn't giving me enough distinction between people with known differences in admixture (Gram vs Mom vs Me). Perhaps in the next 10 years we'll get somewhere with making sense of this for admixed people.

[tomcat]

Hi , Tomcat

I`m looking for any connection to Jewish Ancestry and how close am i genetic wise. I`m alittle slow , could you please explain to me in layman terms, what`s a .pdf ? I don`t own a office 2007 ,but soon. I only have office 2000, i don`t know if Omnipop works with this type of office i have ? So would you please do me a flavor , i`ll send my information the email . you could run it through Ominpop for me and see what`s comes up .

I believe you can run Omnipop on a Windows 2000 machine. You need Excel an application that is usually part of the Windows suite. And you need to enable macros.

A .pdf is a page description file. If you got a Tribes report it was a .pdf. Adobe Acrobat Reader is a free download that permits reading of .pdf's. You can find a .pdf of the Picornell paper online and download as a .pdf.

I am willing to run your numbers BUT the machine on which Omnipop resides in this household is presently on the blink!

[tomcat]

...
Also, my last method of looking at alleles individually and then weighing them, isn't giving me enough distinction between people with known differences in admixture (Gram vs Mom vs Me). Perhaps in the next 10 years we'll get somewhere with making sense of this for admixed people.

Admixture analysis is of compelling interest to geneticists. There are severals ways to compute it and several marker sets employed.

"Informativeness of the CODIS STR Loci for Admixture Analysis"
Jill S. Barnholtz-Sloan, 1 Ph.D.; Carrie L. Pfaff,2 Ph.D.; Ranajit Chakraborty,3 Ph.D.; and Jeffrey C. Long, 2 Ph.D.

There is even some software online.

'LEA' at www.rubic.rdg.ac.uk/~mab/software.html
'ADMIX 1.0' at https://www.web.unife.it/progetti/ge...rgio_soft.html
'ADMIX2.0' at https://www.web.unife.it/progetti/ge.../admix2_0.html
'ADMIX95' at www.genetica.fmed.edu.uy/software.htm
'ADMIXMAP' at www.Ishtm.ac.uk/eu/genetics/admix.html
'STRUCTURE' at www.pritch.bsd.chicago.edu/

It is all way over my head. But if someone can figure this out and devise a successful routine they would have a business.

[Noaide]

More...

"Examining Population Stratification via Individual Ancestry Estimates versus Self-Reported Race "

"Population stratification has the potential to affect the results of genetic marker studies. Estimating individual ancestry provides a continuous measure to assess population structure in case-control studies of complex disease, instead of using self-reported racial groups. We estimate individual ancestry using the Federal Bureau of Investigation CODIS Core short tandem repeat set of 13 loci using two different analysis methods in a case-control study of early-onset lung cancer. Individual ancestry proportions were estimated for "European" and "West African" groups using published allele frequencies"

Noaide

[Maria_W]

Thank you!

Thank you for the link. Will try to read it all later!
Maria

[tomcat]

Anyone looking into Jewish ancestry via Omnipop and the 9 CODIS STR sample would be well-advised to get a .pdf of the paper from which the Omnipop data is derived as Omnipop frequencies are from the Pooled Data. The individual frequencies for Ashkenazi, Oriental, North African and Sephardic Jews can be somewhat different from one another - enough to make a difference....

In defense of Burritt, he likely chose the Pooled Data because the individual sample sizes were too small, only by using pooled could he approach a sample size more equivalent to others in Omnipop.