DNA test to prove/disprove blood relationship going back ten or twelve generations

Question

On my mother's side I am related (in Southern Germany) to the B. family. My grandfather collected data back as far as my 10G-grandfather. I know nothing about his parents, although there is an old family belief that my 10G-grandfather immigrated from Switzerland, where the B. name also occurs. (It is not a particularly common last name, and back to 1589 all members of the B. family in Southern Germany can be traced back to this individual).

I was recently contacted by a Swiss genealogist who shares the B. last name. According to some research that he shared with me, there are three geographic concentrations of the B. name, in Southern Germany, around Zuerich, Switzerland, and in Thurgau, approximately 50 kilometers (30 miles) from Zuerich.

This member of the B. family also indicated that a DNA test had shown that the Zuerich and Thurgau clusters are not blood related.

Would there be any hope of connecting the Southern German cluster into this sample, to settle whether or not there is a relationship between my cluster and one of the other two?

As I said, I assume that one must go back at least ten generations, and I assume that the DNA has been pretty well shaken up, mixed and diluted in the meantime.

Is there any hope of getting meaningful results out of such a test?

How large a sample might be needed?

How diverse would it have to be, that is, how far back along my cluster of the family would I have to go in order to have a hope of definitive results?

Answer 1

There are three types of genetic tests:

• mtDNA
• autosomal
• yDNA

mtDNA is inherited only through direct maternal line. Usually the surnames are the same as the father's surname. So mtDNA won't give you any useful information.

Autosomal DNA is very interesting test. It can say a lot about ancestral geography, i.e. it can help to confirm the ancestry from particular region. But unfortunately it has not very good precision. Also it is useful for confirmation the close relationship (up to xth cousins).

yDNA is inherited via direct paternal line. It is the only test that can help much. There are two types of testing yDNA: STR-markers and sequencing like BigY or Y-Elite. STR tests show the particular marker values. It is relatively cheap test and can give a clue about haplogroup, or, if you are lucky, clade/subclade. So if two persons with the same surname have absolutely different Y-STR results, they can not have a common ancestor (*). The opposite situation when the haplogroup is the same and clade/subclade too - means nothing. You should go deeper with Y sequencing like Big-Y and Y-Elite. These tests analyze almost whole Y chromosome and can give the answer to the question when lived the common ancestor of two males. Also they detect so-called private SNPs - the mutations which are specific to one family. And when you will know what are the private SNPs, you will be able to make express (and cheap) test for particular person for these particular SNPs. If they are positive, it means that this person belongs to your family, otherwise - not.

OK. Let's talk about amount and prices. You should test as much persons as you can. One person from each branch is absolute minimum. Better - two persons from each stem. The prices for Y sequencing is great. So it is usually the limitation factor for the testing of every person :-(

(*) I mean in the timeframe that you mentioned. Of course, we all have the common ancestors, but the time when they lived may vary. And such results usually don't have genealogical value.