Would a Y-DNA test be useful for a NPE?

[MarkyP]

Hi all,

I'm helping a cousin to try and trace his biological father. He has done the autosomal test which has been uploaded to all the various sites. We haven't got huge matches but the ones we do have should theoretically point us in the right direction. However this doesn't seem to be the case we can only seem to get as close as Great Grandparents in some cases, 2x Great Grandparents in others, and are unable to tie the various families together. I'm wondering if a Y test might help? I'm not too familiar with how they work, but am I correct in thinking they would be better for descendants further back? If it is worth doing one, which one should it be?

Thanks in advance,

Mark.

[SouthseaSteel]

As ever with DNA, it might do!! Any DNA test depends on who else has tested, by whatever process and on what platform. I also strongly believe that most people who take a Y test would have done an autosomal test too - but you just dont know that.  However, even without a clear match (such as zero genetic distance on a reasonable number of markers i.e. > 37) you might get an inference if for instance there are a lot common surnames listed.  However, that isnt guaranteed by any means. I have my surname confirmed going back 6 generations both with autosomal DNA and documentation, but you wouldnt really recognise my surname as being statistically significant in my list of matches.

If you have the money available to do it and you want to find out, you have nothing to lose.

 

[MarkyP]

Thank you, it's the genetic distance that I fail to get my head around, it doesn't mean that if my cousin goes for the Big Y on FTDNA he'll stand more of a chance of finding someone of a closer match to him?

[Alexander.]

With Y-DNA I think it is best used when trying to solve a specific genealogy problem. If just wanting to know your ancient paternal haplogroup, it will tell you that too, but not everyone will have close matches right away.

Y-DNA has been very useful to help chip away at a couple of my brick walls, one involving an NPE in the 1820s (a little beyond the reliable scope of autosomal DNA). I have upgraded a few to BigY over the past couple years and although it didn't "solve" every problem I was investigating, in each case it has given me additional information to pursue. I did just get a new BigY match this week which I'm investigating further, so even if you don't have matches right away, you will probably get some more with time.

BigY won't give you closer matches but it will tell you more accurately when your branches split, and will usually give you a modern haplogroup assignment. BigY tests both the usual STRs as well as SNPs, which gives you a lot of detailed genetic information.

In your case, to confirm the paternal line with Y-DNA, if no existing matches in the database, you might need to find a male cousin related to the possible father to also test. Looking at autosomal matches might be the easiest place to find such as candidate. Even a fairly distant cousin is okay when it comes to Y-DNA, as long as it is a direct paternal relationship on each line. A Y-37 test on each line may be sufficient as a starting place. Most people will be happy to test, especially if you pay for it.

[SouthseaSteel]

You asked a specific question (what is genetic distance), which is good as it is that which is most relevent to your very specific goal (finding your cousin's father).

Firstly the results from a Y test bear no resemblance whatsoever to Ancestry like autosomal results.  You do get a list of matches, and their surnames, who are contactable, but they are listed with their "genetic distance" from the tester.  At a VERY basic level, genetic distance is the number of mutations between the two parties Y chromosone. And as mutations only occur from one generation to another, genetic distances are an "indication" of how many generations difference there is between the two testers. A genetic distance of zero is no mutations suggesting the very closest of relationships.  Then there is 1, then 2, then 3 etc.

Zero gd is instantly very interesting and you need to contact that person to discuss.  A gd of 1 is marginal but it is still worthy of contact.  A gd of 2 is pretty useless tbh and this is my closest match.  (ftDNA actually suggests in my case we are descended from a common ancestor born in 1600 - 1700 which is absolutely no use to you, nor me for that matter).

Now, the other big difference is that bigger does not mean better.  If you test at 37 markers, like me, and get matches no closer than a gd of 2, then testing any more markers up to Y700 is a waste of time in what you are trying to achieve.  If I look at my 12 marker results I have 10s of zero gd matches and I have many at 25 markers too. But when sampled over 37 matches I have no zero gd matches.  Doing what you are trying to do, then it would be pretty much a waste of money and time me upgrading to more markers in this case.

Now, if you find matches at zero (or 1) gd then it is worth upgrading but ONLY if the other person has upgraded too.   This is quite easy as you do not need to retest you just pay an upgrade fee and voila, fresh more detailed analysis. 

For what you need I suggest you do a 37 (or maybe a 67) marker test and see what you get.  Any zero gd matches and you are coooking on gas straight off the bat. If after chatting to this match, you may want to upgrade, but only after ensuring that they have already done more markers than you or at least intend to.

Like I said before, the distribution of match surnames might be interesting and useful, maybe not.

In addition, you will also get a bunch of numbers and letters that constitute a haplo group. Not relevant in finding your cousins father but to some its interesting.

Good luck

[Alexander.]

Zero gd is instantly very interesting and you need to contact that person to discuss.  A gd of 1 is marginal but it is still worthy of contact.  A gd of 2 is pretty useless tbh and this is my closest match.  (ftDNA actually suggests in my case we are descended from a common ancestor born in 1600 - 1700 which is absolutely no use to you, nor me for that matter).

I would disagree a little here. I look into all GD 2 and 3 matches at this level and some are helpful, some not. I have a GD 2 match for one of my kits at 37 markers, and if you look at the actual STRs there is only one that differs - one mutated up to 13, one down to 11. All other markers are a perfect match. This makes it a GD 2, but only one marker mutated, so the estimated relationship was significantly more distant than actually. Similarly I have seen some that have mutated up and back down over a few generations, so that there appeared to be a GD 0 when there was actually two mutations (appears closer than it is). So STRs are good to give a rough estimation of relationship but not nearly as reliable as the SNPs you get with the Y700. In any case, I definitely agree that a 37 marker test is a great place to start, you can always upgrade in the future.

If you are looking to get a test now, FTDNA does have it's Father's Day sale going now which would save you a bit.

[Biggles50]

As advised a yDNA test might or might not yield results.

There is no need to go for the most expensive Big Y as you can upgrade retrospectively without having to supply another sample.

What you may find is that you have a few matches with the same surname, but hundreds with different surnames.

I took the ftDNA 111 test and that included a Haplogroup of which I am a part.  It is pretty useless in reality and upgrading to the Big y will only give a more accurate Haplogroup membership, again the usefulness to a genealogist researching their family tree is questionable with the migration paths being the main benefit.

In my own case, my two closest matches had the same surname and a third a slight variation so that gave me confirmation of the paternal family that I am a part of.

Hence a yDNA test could provide a Family Surname that could aide the research.

[MarkyP]

Thank you all for your advice, I think the 37 marker test is the way to go then, and see what we get. Thanks again.

[Biggles50]

Thank you all for your advice, I think the 37 marker test is the way to go then, and see what we get. Thanks again.

Do come back post results and let the group know the outcome