Interpreting autosomal results correctly?

[Noaide]

All who have tested ABDNA 2.5 should read the above DNAWitness page, its far more informative than the consumer pages.

Noaide

[vineviz]

If an autosomal SNP mutation happend just after two small populations departet and prospered totally in one of the populations, and the two populations where totally isolated from each other, it should be possible theoretically to identify based on the single autosomal SNP what population each individual came from.

The problem is that gene frequencies render this theoretical possibility to be extremely unlikely except in limited cases.

Take a look at the plot of EuroDNA results for the sample populations used to construct the test.

http://ancestrybydna.com/welcome/ima...roDNA_0015.jpg

There are people in this sample from Italy whose test results are nearly identical to people from England, Finland, Germany, and Poland.

Imagine a person whose ancestors all came from the same part of Europe but who doesn't know which part of Europe that is. An autosomal test would probably be helpful in this case.

But the more common scenarios are ones in which people already know where most of their ancestors came from, and want the test to help them figure out th heritage of a single grandparent. Except a the continental level (European vs. African vs. Asian etc), testing probably cannot help there except in a recreational sense.

[tomcat]

...
As I understand SNP markers, you have the mutation or you dont, so to get for example a reliable native american autosomal SNP test you must find SNP mutations that occured after the Beringia passing and did not back migrate in any way, If this is the way ABDNA is made, ABDNA must have wrongly identified some NA autosomal SNP as exclusivly belonging to NA, they also occur in Asia, similar with the problem of Europeans getting "East-Asian" scores, the ABDNA product may have wrongly categorized actually common Euroasian autosomal SNP as exclusive to East-Asia. The same categorization problem also occur in the Euro 1.0 test.
...

That fundamental fault seems to derive from insufficient field data collection and the 'purifying' of data by exclusion - samples that did not conform to population norms were excluded as representing admixture. From an historical or genealogical perspective one might be fully 'European' but because examples of European admixture have been excluded one may get less than 100% European score combined with African, Native American and East Asian scores. In other words, the admixture reflected in one's test results might be 'old' but fully within range for a modern day European.

[burto]

I personally think after reading all the info from AbDNA that the high Asian scores Mum got do reflect something other than an ancient signature.
I agree that the test doesn't really tell us any more than we have Asian and European DNA which is annoying because the Asian could be from an actual East Asian ancestor, or a Native American ancestor and/or a South Asian ancestor.
What I am curious about is the fact that no Middle Eastern was recorded on her Euro 1.0 test. I would think the Roma would have migrated and mingled with locals on their route into Europe? Maybe not enough to influence their gene pool?

[Arch Yeomans]

Given Up

Congrats if your autosomal results match perfectly to your genealogical records. I've about given up on autosomal DNA testing and analysis. It's really frustrating that FTDNA offers the tests with NO explanations or any interpretation of the test results and yet DNA Tribes does. In my opinion, really this should have been an added feature, though OmniPop may have useful features for the forensic scientist, it does nothing for the poor sap researching his or her genetic line unless all the intricate and complicated features of OmniPop are understood (which most do not understand). You would think we would have more useful data to work with from FTDNA, but really all I have seen is a bunch of speculation and guessing of autosomal ancestry (it's all we have for now).

[Noaide]

... fundamental fault seems to derive from insufficient field data collection and the 'purifying' of data by exclusion - samples that did not conform to population norms ...

There is another problem with autosomal STR markers like CODIS, if your specific profile do not conform to the population norm, even if your ancestors all originated from a well tested european population may end up as been classified as East Asian, Indian or Nubian, this problem do not seem to appear in the autosomal SNP testing where the main ancestry is correct with usually only minor components related to the problem we just discussed.

Noaide

[litlnemo]

It's really frustrating that FTDNA offers the tests with NO explanations or any interpretation of the test results and yet DNA Tribes does.

Well, amen to that. Explanations are a weak point in all of FTDNA's services, but especially for autosomal results.

[vineviz]

It's really frustrating that FTDNA offers the tests with NO explanations or any interpretation of the test results and yet DNA Tribes does.

Perhaps FTDNA has decided that a lack of explanation is better than a wrong explanation?

[marvallen]

Is there really any origanization that can give even a quasi-correct interpertation of autosomal results as related to genealogy?

[nas]

Hello,
"Autosomal" isn't Ftdna's business,but Family Projects.
As far as I understand it..."Autosomal"is the future in Genealogy research(admixed people,adoptees,and NPE's).
I ordered the AncestryByDNA 2.5-test,to start with... .
I think it's all a question of personal interpretation

Thanks.
Nas

[vineviz]

Is there really any organization that can give even a quasi-correct interpertation of autosomal results as related to genealogy?

If so, I have not seen it.

Most companies to date seem to have focused on the "ethnic background" angle, which is inherently problematic except at the very broadest level.

There is real potential in the research into haplotype blocks, but that field is barely nascent.