I have traced one branch of my family back to a Hannah Wilson 1763-1850 based around Kendal. There are a lot of Hannah Wilson’s around at that time/place and it is difficult to go further.
Some researchers have her descended from a family with the name Wakefield via a grandmother Hannah Wakefield (1705–1738) born at Challon Hall, Preston Patrick, Westmorland, England and dying in Kendal. I have not found any corroborating evidence so far.
As a bit of an experiment really I searched all my DNA matches on ancestry.com - you can search for all matches that have a specific name in their linked tree.
I found about 30 trees with the name Wakefield in their trees. They were all distant DNA relatives with around 9-13 cM matches. Many are based in the US and none of them have any other obvious links into my tree.
Do you think this is enough to be significant or is it just a coincidence? I’m deciding if I should put much effort into this line of investigation or if I should spend my time elsewhere. Any thoughts?
We welcome any query on Who When Where. If you have previously posted it on another forum (including the old WDYTYA forum), please state this in your opening post - this will save people redoing the research which has been done before: they can look at it and possibly go further with it.
How many distant DNA hits are significant?
-
AdrianBruce
- Posts: 358
- Joined: 14 Jun 2020, 18:57
- Location: South Cheshire
Re: How many distant DNA hits are significant?
Don't quote me on these numbers, but the suggestion is that 7cM matches (which I don't think Ancestry report any more) were genuine matches (identical by descent) only half the time - the others were just identical by chance. Remember the DNA alphabet is not huge and while that's a gross over-simplification, in the same way that DNA segments "mutate" further apart, so it's also possible that segments "mutate" closer together.
The assessment of 7cM matches being 50/50 came from studies where some people and both parents had done DNA tests. If the child matched someone but neither parent matched that someone, then (leaving aside jokes about the milkman) the match of the child to that person must be purely by chance.
Now, this shows that the chance of a small match being the result of a "mutation" between child and parent is about 50%. What this survey didn't do is assess what the odds were of the "mutation" being on the previous generation... It must happen, I suggest.... Therefore, I believe that there is a reasonable chance that matches in the 9-13 cM range are also identical by chance.
Conversely, having the coincidence of surname would suggest differently!
I had (and still have) similar situations in my Ancestry matches, where there were numerous low level matches including people who, judging by their profile pictures, were African-American. It's difficult to see how my distant ancestors or their cousins fit into such a picture. I thought I might have found one possibility in an emigrant to late 1800s Massachusetts but their descendants seemed to fly past these other trees. I had another similar level match to a white American genealogist from Virginia who seemed to have a well populated tree many generations back. So many generations back, that her tree seemed to go past the degree of cousinship usually implied by that degree of match, with no sign of any names that I recognised.
Frankly, if you're looking for a distant match and have no clue, getting a DNA match isn't going to get you any more of a clue....
The assessment of 7cM matches being 50/50 came from studies where some people and both parents had done DNA tests. If the child matched someone but neither parent matched that someone, then (leaving aside jokes about the milkman) the match of the child to that person must be purely by chance.
Now, this shows that the chance of a small match being the result of a "mutation" between child and parent is about 50%. What this survey didn't do is assess what the odds were of the "mutation" being on the previous generation... It must happen, I suggest.... Therefore, I believe that there is a reasonable chance that matches in the 9-13 cM range are also identical by chance.
Conversely, having the coincidence of surname would suggest differently!
I had (and still have) similar situations in my Ancestry matches, where there were numerous low level matches including people who, judging by their profile pictures, were African-American. It's difficult to see how my distant ancestors or their cousins fit into such a picture. I thought I might have found one possibility in an emigrant to late 1800s Massachusetts but their descendants seemed to fly past these other trees. I had another similar level match to a white American genealogist from Virginia who seemed to have a well populated tree many generations back. So many generations back, that her tree seemed to go past the degree of cousinship usually implied by that degree of match, with no sign of any names that I recognised.
Frankly, if you're looking for a distant match and have no clue, getting a DNA match isn't going to get you any more of a clue....
Adrian Bruce
Re: How many distant DNA hits are significant?
Adrian I agree with everything you say. DNA doesn't build trees for you!
It was really a question as to if whether people think that this number of "distant relative" hits is significant enough to justify some good old fashioned detective work in Cumbria or if I would be better served following other branches. Or put another way - what's the chances of it being a clue, rather than a red herring?
It was really a question as to if whether people think that this number of "distant relative" hits is significant enough to justify some good old fashioned detective work in Cumbria or if I would be better served following other branches. Or put another way - what's the chances of it being a clue, rather than a red herring?
Re: How many distant DNA hits are significant?
I would agree with Adrian BUT If all else has failed then maybe it is worth spending a morning/afternoon/day on it.
I have to relate that earlier this year someone contacted me with a DNA match of 7cM and the name was SMITH!! My heart sank, but because we both had a Smith line in the same area I agreed to take a loook and after a few days of searching records and DNA matches we did find the link and I have a new 5th/6th cousin. The problem was the transcription indexes. On one we found the Father's name as we needed it to be, on all the others it was a different Smith. We emailed Berkshire Record Office, who I have always found to be very helpful and prompt. They checked their transcription and it was the wrong Smith but, when they checked the original record, it was the right one!
So it is possible, but it just depends on who has tested, and if you can marry up the results.
I took my DNA test with Ancestry, but I have also uploaded the results to My Heritage.
As Adrian said Ancestry now disregard any match below 8cM. As yet they do not have any DNA Tools to help.
My Heritage, however, has the auto cluster report, which will group your DNA matches according to your chromosome match, but will usually only work for 20cM + matches if memory serve s me correctly. They also have the chromosone browser whereby you can match yourself to up to 7 (?) other DNA Matches and it will show you which chromosome you all share if any. This works with with any level of DNA Match.
I have to relate that earlier this year someone contacted me with a DNA match of 7cM and the name was SMITH!! My heart sank, but because we both had a Smith line in the same area I agreed to take a loook and after a few days of searching records and DNA matches we did find the link and I have a new 5th/6th cousin. The problem was the transcription indexes. On one we found the Father's name as we needed it to be, on all the others it was a different Smith. We emailed Berkshire Record Office, who I have always found to be very helpful and prompt. They checked their transcription and it was the wrong Smith but, when they checked the original record, it was the right one!
So it is possible, but it just depends on who has tested, and if you can marry up the results.
I took my DNA test with Ancestry, but I have also uploaded the results to My Heritage.
As Adrian said Ancestry now disregard any match below 8cM. As yet they do not have any DNA Tools to help.
My Heritage, however, has the auto cluster report, which will group your DNA matches according to your chromosome match, but will usually only work for 20cM + matches if memory serve s me correctly. They also have the chromosone browser whereby you can match yourself to up to 7 (?) other DNA Matches and it will show you which chromosome you all share if any. This works with with any level of DNA Match.
Sally