Y-DNA and First Cousins

[DavidK]

Two of my second cousins on my maternal side recently had their fathers tested at FTDNA. The Y-111 results are confusing, with a genetic distance of 4.

The two men supposedly share the same paternal grandfather, but I personally have Y-111 matches of genetic distance 3 and 4 with no common ancestor in at least 5 and 6 generations, respectively.

In my Ancestry Autosomal DNA results, the daughter of one of these men matches me with 293 cM across 12 segments (above the 2C average of 233; range 46-515). Her father hasn't done an atDNA test.

The children of the other man match me with 258 cM/11, 187 cM/8, and 99 cM/5, respectively. Their father matches me with 439 cM across 19 segments (bang on the 1C1R average; range 141-851).

These atDNA results seem consistent with the paper trail, but the Y-111 results do not. The TiP report says the probability that they are first cousins is 29.47%. I realize that this is not zero, but...

The Y-111 differences, in case they're helpful, are as follows:

DYS19: 15 vs. 15-16
DYS442: 12 vs. 13
DYS446: 15 vs. 13

Both of these men have brothers who are still alive, and may be willing to test. And there's also a third man who shares the same paternal grandfather via another of his sons. Would it be worthwhile testing any of them?

I'd appreciate advice on how I might investigate further or, frankly, if I should bother. Perhaps this is just the nature of Y-DNA?

 

[Tim Clarke]

Two of my second cousins on my maternal side recently had their fathers tested at FTDNA.

You say 'your maternal side' which immediately means that there can be no Y-DNA matches with yourself since Y-DNA only travels in the male line. I hope I read your message correctly.

 

[peter]

Y-DNA changes as a result of random mutations (some STRs are more likely to mutate than others), so the best you can hope for - without other relevant evidence - is to get a statistical estimate of how close the relationship might be.

If I look at a match of mine which has a difference of 4 at 111-markers this is what FTDNA tell me:



In other words there is a 12% chance that we share the same grandparents - but there's also a 12% chance that our common ancestor is more than 8 generations back. Or to put it another way, there's a 76% chance that our common ancestor is between 3 and 8 generations back - an enormous range, in genealogical terms, yet there's 1 chance in 4 it could be outside the range.

So yes, it's in the nature of Y-DNA that we get results which are indeterminate - which many people will find confusing, especially if they didn't take advice before testing.

This wide range is also why I don't generally recommend Y-DNA tests - there are occasions when they are the only solution, but those occasions are few and far between. In this instance testing two such close relatives was probably a waste of money since atDNA tests would have shown if they shared the same grandfather (and much more besides).

 

[DavidK]

You say 'your maternal side' which immediately means that there can be no Y-DNA matches with yourself since Y-DNA only travels in the male line. I hope I read your message correctly.

That's why we tested male-line descendants on my maternal side. My own Y-111 matches are, of course, on my paternal side; I mention them only for comparison purposes.

Y-DNA changes as a result of random mutations (some STRs are more likely to mutate than others), so the best you can hope for - without other relevant evidence - is to get a statistical estimate of how close the relationship might be.

The overarching goal is (was) to establish a surname project for purported descendants of the ancestors of my maternal grandfather, who was from north Armagh, in Ireland. There are a number of branches of the McAtamney family, but the records begin only at about 1840. We have, for example, the marriage of my 3G grandparents, William and Isabella, but their fathers' names are not given. At that distance, atDNA is not proving useful.

We tested a couple of close male-line family members in order to establish a reference. We have Y-25 results from a descendant of Bernard, who would be a peer of William's, but who does not appear in the records (he emigrated to America). We're also in touch with several descendants of Eugene, another peer of William's, who married twice and had at least 20 children. A next step would have been to get one or more of them tested, but that doesn't seem like it would provide useful information.

Thanks for reading!

 

[peter]

At that distance, atDNA is not proving useful.

It should be - as long as it is the oldest generations that are testing atDNA can reach back as far as the second half of the 18th century, sometimes earlier. However you can't just look at one line when you use atDNA - you have to look at them all, otherwise the clues will be all jumbled up.

However it all depends what you are trying to achieve. If all you want to know is which living McAtamneys share your McAtamney ancestors then Y-DNA is the answer. But there aren't many questions to which Y-DNA is the answer, and it can be an expensive learning curve for those who embark on a project that is doomed to failure.

I set up a Y-DNA surname project because I'd like to prove where my surname came from (there are two conflicting theories). It's a fairly modest goal, but even so I'm no closer to the answer 8 years down the line.

 

[DavidK]

It should be - as long as it is the oldest generations that are testing atDNA can reach back as far as the second half of the 18th century, sometimes earlier.

I'll follow up on this as a new thread, so it's more useful to atDNA researchers.

If all you want to know is which living McAtamneys share your McAtamney ancestors then Y-DNA is the answer.

I was hoping for a little more granularity, to be honest; as in, approximately where ancestors connect. Clearly, that's not realistic at the current scale!

In theory, if I got other 1C1R McAtamney males tested, I could start to track mutations and extrapolate the "original" STRs back, generation by generation, but that would be fraught and expensive.

I sent a Y-111 kit to a paternal 3rd cousin, but since then I've started to find atDNA matches to our shared 2G grandparents. A zero genetic distance match would tick off a box, but it could just as easily be 4 or more. He hasn't taken the test yet. I'm inclined to nix it and save the test for someone else.

 

[peter]

I was hoping for a little more granularity, to be honest; as in, approximately where ancestors connect. Clearly, that's not realistic at the current scale!

In theory, if I got other 1C1R McAtamney males tested, I could start to track mutations and extrapolate the "original" STRs back, generation by generation, but that would be fraught and expensive.

I wouldn't have thought that testing such close relatives is going to help much.

I sent a Y-111 kit to a paternal 3rd cousin, but since then I've started to find atDNA matches to our shared 2G grandparents. A zero genetic distance match would tick off a box, but it could just as easily be 4 or more. He hasn't taken the test yet. I'm inclined to nix it and save the test for someone else.

Didn't you have to specify who would be testing when you placed the order? If so that might be a problem.

 

[DavidK]

Didn't you have to specify who would be testing when you placed the order? If so that might be a problem.

I gave FTDNA a ship-to name & address, but didn't commit to a testee.