*SUCCESS* Autosomal DNA Test - my experience

[AngelaR]

Given my experience so far, Dudley, it could just be a couple of weeks! Hope so, anyway

[Redroger]

This is even more tricky when you get out into the Pre-Industrial Revolution countryside where the frequent marriage of country cousins means that people shared bits of DNA with most other people in the village!

Autosomal DNA is very very iffy...

Not just in the Pre industrial revolution era, well into the 20th century in the village of Lode Cambs
cousin marriage was the norm rather than the exception. I am my own 4th cousin as a result.

[DudleyWinchurch]

Just to reiterate,

I have a particular problem which is, that documentary evidence suggests, but does not confirm, that my great-grandfather may have had two siblings who moved to England at the same time but later all or part of their families moved on to the USA.

I am in touch with a descendant of one of them who, if we are correct in believing that the great-grandparents were siblings, would be my third cousin.

We are hoping that the autosomal test will either confirm the relationship or show that this is unlikely.

Most scientific tests only operate within a limited range of parameters.  If they are used to speculate beyond those ranges, then the results can only be speculation.

From what I have read of Autosomal DNA testing, this question should fall within a reasonable degree of confidence for this particular test. 

While I am happy to see and try to interpret any other results that I might get, against documentary evidence, I am not looking for vague hypotheses about where my family lines may have arisen much further back or fishing for hitherto unknown connections.

Some of my lines do go several generations further back, but this test does not offer any confidence in relationships identified beyond even one further generation, so I will not be counting on the test as evidence. 

Just trying to see if this DNA test can be of use in finding out more of my family history.

[DevonCruwys]

Dudley

There is a 90% chance that two third cousins will share enough DNA to show up as a match so the odds are very much in your favour. See this FTDNA FAQ:

http://www.familytreedna.com/faq/answers.aspx?id=17#628

However, this does mean that one in 10 documented third cousins will not show up as a match. This is not a deficiency of the test but simply a limitation of the genetics and a consequence of the random way in which we inherit our DNA.

What this means is that if you are a match you will be able to prove the relationship (assuming that there is no possibility that you are related on any of your other ancestral lines). However, the lack of a match will not prove that you are not related. If there isn't a match then you would need to test other relatives instead.

See this blog post from Judy Russell. She didn't match a known third cousin but a number of her other relatives did:

http://www.legalgenealogist.com/blog/2013/04/07/widen-the-net/

You might also like to read some of the articles linked here:

http://www.isogg.org/wiki/Autosomal_DNA_statistics

[DudleyWinchurch]

Hi Devon,

some very clear and useful explanations there.  I am used to working with statistics so aware of the possibility of anomalies like that.

If we don't get a match first time, I have siblings (and/or cousins) who may also take the test.  I'd be interested for at least one to do so, to find out how much our matches have in common, anyway.  It is also possible that the third family involved may have someone who could do so too, but I have not gone down that route as yet, as I think my contact in that family is at least one, if not two, generations later than us.

Unfortunately, on my line for this family, there are no survivors of an earlier generation, that I am aware of, to make the test more powerful.

thanks

[DudleyWinchurch]

AngelaR,

You were right.  Just logged on and founds results, some five weeks before expected.

Rather a lot of results - 28 pages in all, so nearly 280 "matches".  As most of them are 4th or 5th cousin to distant, I am guessing that they are spurious matches.  However of the 26 or so that suggest 2nd-4th or similar, I only recognise the one that comes up as closer, 2-3rd cousin.  That is someone with whom I am usually in regular contact but did not know he had taken the test, and is definitely a second cousin.

Looking through the rest, I can find no obvious connections although some have Mayo connections and others Pennsylvania and Irish connections so could be connected to the two families that we are trying to match up.

Most seem to have a significant number of German connections, not expected, so it would be interesting to see if they match each other in this respect, and might be a spurious connection to one extended string match.

I am now thinking that getting my brother to do the test too may be a way of at least partially separating those which are coincidental from those actually likely to be related.

[AngelaR]

Dudley

I'm pleased the results are through although, like mine, they seem to be rather mixed. My nearest were 2nd-4th cousins and there were only 2 of them, so I think my relatives must be rather shy of DNA tests!

As a matter of interest - did you look at the Population Finder under 'Family Finder'?  My results seem to say that I'm 1/16th Middle Eastern, which to me means that one of my Somerset/Wiltshire ag lab great grandparents had an exotic spouse.....

Good luck with the hunt

[DudleyWinchurch]

Hi Angela,

Yes, the results are interesting but I suspect most are not significant.  I've tried using the various tools and all but one, except for my second cousin, only seem to have one extended piece of DNA in common.  The odd one out has two bits on the same chromosome, but quite extensive and matching both me and my second cousin so, although there is nothing on the record to suggest that we are indeed related, I have emailed to ask if that shared family name appears in the other person's tree too.

I had the one 2nd-3rd cousin match and another seven 2nd-4th cousin matches but I haven't yet found any obvious connection among them.

Yes, I did look at the origins, but it seems that the database they are matching to is not yet complete so there is a warning that this may change.  Currently it says that I am 96% Orcadian and 4% Middle Eastern which suggests perhaps a Jewish ancestor some 4 or 5 generations ago.

I'm a bit confused that they are using Orcadian to match the whole of the British Isles, as although it makes sense in the case of my Celtic origins, I would have expected German, French and other Western European influences to have had a significant impact on the population of England too.

I think that there seems to be some issue with the attempt to select comparison communities which will be relatively unmixed, hence using Orcadians, but I suspect that there are few other areas of Europe even represented in the database, as most populations are quite mixed.

This definitely raises more questions than it answers.

[DevonCruwys]

The results from the Population Finder test should not be taken too literally. FTDNA are only using a very limited range of reference populations:

https://www.familytreedna.com/faq/answers.aspx?id=22#1039

It's rather like trying to paint a picture when you only have a limited palette of colours. There is no "British DNA" so it has to be mixed up from what is available which is often Orcadian and Middle Eastern. I'm finding that large number of the people in my Devon DNA project are coming out with small percentages of "Middle Eastern" ancestry in the Population Finder test, and many of them have solid English pedigrees going back for many generations. The Middle Eastern percentage doesn't usually mean that you have any Jewish ancestry or any other exotic ancestry. It just means you're a bog standard Brit. This guest blog post by Doug McDonald provides an explanation of the process:

http://dna-explained.com/2012/09/09/doug-mcdonald-on-biogeograpical-analysis/

You can submit your raw data files to various third-party sites for alternative readings:

http://www.isogg.org/wiki/Admixture_analyses

FTDNA are supposed to be updating the Population Finder results at some point but it's anyone's guess as to when this might happen. They are holding their group administrators' conference this weekend so I'm hoping there might be news of new developments coming from that.