DNA puzzle

[Pauline]

I have a DNA match (A) that has long been puzzling me. I now have a bit more information to go on - though still no answers.

A and I share 135 cM across 8 segments. A's son B has now tested and he and I share 98 cM across 4 segments. My sister has both these matches in her list and shares 59 cm across 5 segments with A, and 6.6 cM across 1 segment with B.

I cannot identify from their tree how I might connect with A & B but location-wise, it could possibly be on my father's side via his paternal grandfather, in the Nottinghamshire/Derbyshire/Leicestershire area.

Shared matches, however, mostly suggest the connection might likewise be on my father's side, but via his paternal grandmother in the Gloucestershire area. I also have a couple of unknowns in my shared matches, and one seemingly anomalous match (C) on my mother's side (although C does have some ancestry in Gloucestershire). My sister also has one unidentified shared match (D) and I can see a connection between D and A in the Leicestershire/Derbyshire area, but not how D connects to us.

One of the shared matches in the Gloucestershire area is a documented 3rd cousin (E - our shared 2g grandparents being Thomas & Jane) with whom my sister and I both share about 25 cM across 2 or 3 segments. E shares 120 cM with A. Other shared matches in Gloucestershire are more distant documented cousins via ancestors of both Thomas & Jane, though mostly via Thomas.

Jane was born illegitimate, and her father is as yet unknown, so she is an obvious point of investigation, though this wouldn't explain the connections being mostly with her husband Thomas.

I have done lots of investigations with all the available trees, searching for others sharing the same surnames and locations, but I have so far drawn a complete blank.

Has anyone any suggestions of how I might proceed with this, please?

 

[peter]

IMPORTANT NOTE

Please be very careful about posting match information - this is a public forum and the people you have been matched with will be able to identify themselves from the information given. In most cases you can convey sufficient information without providing the precise length of matches, and certainly without giving names or initials - thanks to Pauline for taking those precautions in this case.

This forum might seem like a cosy discussion between trusted friends and colleagues but it visible to everyone in the world. If we are careless about the way we use DNA information it will have unwanted consequences.

 

[peter]

As a general point your match with B and your sister's match with B can be safely ignored - the figures for the matches with B are compatible with B having inherited the shared DNA from only one parent. This simplifies the analysis.

Rather than attempt to leap to the answer in one go it's better to take small steps. This helps to ensure that you don't miss the correct solution simply because it's implausible, unattractive, or simply doesn't get considered.

The first step in any situation is to list the possible relationships that are compatible with the amounts of DNA shared. In this respect having matches to both yourself and your sister helps, because it can eliminate some the less likely choices (less likely statistically, not less likely in practice). The coloured chart in my Masterclass isn't quite the latest version from Blaine Bettinger, but for practical purposes it's just as serviceable.

In this case most of the close relationships will be eliminated. The next step is to consider the number of generations difference - for example, if you and your match are around the same age, it's relatively unlikely that you are their great-aunt (or vice versa), or that the two of you are cousins twice removed.

The third step is to consider whether you are likely to share one common ancestor or two. Where illegitimacy is involved it's usually only one - which means that you would focus on the half-cousins and the half-aunts. But just because illegitimacy doesn't necessarily mean that you don't share two common ancestors - it depends where in the chain the illegitimacy was. For example, the child of your father's full sister is a full 1st cousin to you irrespective of who the father was.

Tip: I find it often helps to draw out an example family tree and work it through.

 

[Pauline]

Thank you, Peter, for your very helpful input.

I already know that A is quite a bit older than me, and my guess has tended to be that we have one shared ancestor rather than two. I can also guess at the more likely sex of this ancestor (but saying which might be giving too much away). I'm also guessing at a possible two generation difference because of age differences going back in our trees.

The problem from there is that:

1) I cannot definitely place the most likely suspect(s) in the right place at the right time, although they may not be too far away

2) I cannot then explain how shared matches C & D and my two unidentified shared matches (these two are related to each other) link into this

3) I am not confident that I am not misreading things, and that the link is not quite so straight forward as it might first appear

 

[peter]

Best not to zoom in too closely on particular individuals - there's a tendency to focus on males in your own tree (and therefore females in the other person's tree) rather than the other way round. From the point of view of an observer both are equally likely.

Also you can't know that you have one common ancestor rather than two - it's too early to tell. More likely, perhaps, but that's something to consider at a later stage.

You now have your list of possible relationships, work through them one by one and see what must have occurred in each case for it to be correct. Don't disregard a hypothesis simply because you don't think the two people could have been in the same place at the same time - for example, people have affairs on holiday (or so I'm told).

Don't worry for now about C, D, and the other shared matches - it'll just cramp your thinking.

 

[Pauline]

... there's a tendency to focus on males in your own tree (and therefore females in the other person's tree) rather than the other way round.

I guess that's understandable. If a male ancestor had an extra-marital affair (or whatever) any resulting child is likely to be in someone else's tree. If a female ancestor had one the resulting child is likely to be in your tree and may change who your ancestors are. However, there may be good reasons, including DNA evidence, to make one more likely than the other.

Also you can't know that you have one common ancestor rather than two - it's too early to tell.

Depending on documented evidence, sharing two ancestors could presumably suggest an adoption? Again, DNA evidence might indicate whether this is more likely to be an adoption in or an adoption out.

You now have your list of possible relationships, work through them one by one and see what must have occurred in each case for it to be correct. Don't disregard a hypothesis simply because you don't think the two people could have been in the same place at the same time - for example, people have affairs on holiday (or so I'm told).

I have already done a fair amount of this, but will revisit it. As well as holidays (and I'm not sure to what extent our ancestors took them) there are those who may have travelled with their jobs.

Don't worry for now about C, D, and the other shared matches - it'll just cramp your thinking.

OK, I'll try not to for now, but presumably they will still impinge on the situation somewhere.

 

[peter]

I guess that's understandable. If a male ancestor had an extra-marital affair (or whatever) any resulting child is likely to be in someone else's tree. If a female ancestor had one the resulting child is likely to be in your tree and may change who your ancestors are. However, there may be good reasons, including DNA evidence, to make one more likely than the other.

The children of single mothers were often adopted or brought up by grandparents as their own, so they won't necessarily be obvious even if they are in your tree.

Depending on documented evidence, sharing two ancestors could presumably suggest an adoption? Again, DNA evidence might indicate whether this is more likely to be an adoption in or an adoption out.

Sharing two ancestors could indicate an adoption, but it also happens when the happens when there is an illegitimate birth in a more recent generation. For example, if your mother's brother fathered an illegitimate child, irrespective of circumstances the common ancestors would be your grandparents, and that child would be your 1st cousin, not a half 1st cousin.

 

[Pauline]

For example, if your mother's brother fathered an illegitimate child, irrespective of circumstances the common ancestors would be your grandparents, and that child would be your 1st cousin, not a half 1st cousin.

Yes, of course - and you’d already pointed that out above. I think my brain must have knocked off early yesterday evening!

 

[Pauline]

And just to confirm my brain is now working again, the list of possible relationships which I had extracted from the Masterclass chart still looks to me to be complete:

Full blood - 1C3R, 2C, 2C1R, 2C2R, 2C3R, 3C, 3C1R
Half blood - 1C1R, 1C2R, 1C3R, 2C, 2C1R, 2C2R, 3C, 3C1R

For reasons I probably shouldn't mention here, I have been focusing on a particular sex of people from my tree and a particular degree of relationship (with varying removals) with A.

But I appreciate that there may be other possibilities.

 

[peter]

I think you missed out a couple of relationships:

Full: 1C2R
Half: Great-great aunt/great-great niece

Perhaps you can eliminate some of the possible relationships based on relative ages? I'm thinking of the great-great aunt, the 3R relationships, and perhaps the 2R as well? It depends what you meant by 'quite a bit older' above.

General note: it's important to bear in mind that in theory men can father children at any age - my 3G grandmother married a man old enough to be her grandfather, and they produced one child, who was a half-sister to his daughter from his first marriage, but 52 years younger. Undocumented relationships could involve even greater age disparities.

 

[peter]

You could also tentatively remove others from the list where the size of the match is unlikely (though not impossible). See the table in Masterclass which gives theoretical shared DNA, eg 56cM for 3rd cousins, so for a 3C1R you'd divide that by 2 (28cM) and for a half 3C1R you'd divide it by 2 again (14cM). I'd put both of those relationships on the 'last resort' pile.

The important thing about this process is that we're sifting the relationships based on objective criteria. The longer we can delay getting our feelings involved the better (of course, it's far easier for me because it's not my family).

At the end of the day you might conclude that the most likely relationship is the same one that you guessed at the beginning - and that would be great. But I've been involved in so many of these quests that I know those first impressions are sometimes wrong, and when this happens there's always the possibility that someone gets hurt unnecessarily.

 

[Pauline]

A was born around the same time as my parents, so she can't be any kind of niece to me. A half gg aunt would be a half sibling of one of my great grandparents, and all of my gg grandparents died before she was born.

I don't think I can eliminate 2R relationships, though I could maybe eliminate 3Rs. The theoretical shared DNA chart would allow for a half 1C2R and that would be theoretically possible.

I guess it's hard for anyone to be completely objective about undocumented relationships concerning their own ancestry, but at the same time anyone researching their family history, particularly if they do a DNA test, needs to be prepared for unexpected skeletons to come tumbling out of the cupboard. It's already happened to me and was one of the things that first got me into family history.

And after all my investigations into the various available trees an undocumented relationship is looking like a distinct possibility.

 

[peter]

I guess it's hard for anyone to be completely objective about undocumented relationships concerning their own ancestry, but at the same time anyone researching their family history, particularly if they do a DNA test, needs to be prepared for unexpected skeletons to come tumbling out of the cupboard. It's already happened to me and was one of the things that first got me into family history.

That's absolutely true. But some people assume that the most obvious answer, or the one that fits their preconceptions of what their antecedents were like, is the correct solution.

In the vast majority of DNA cases that have passed across my desk the initial assumption has been wrong - just think how much unnecessary heartache might have been caused had they mentioned their hypothesis to their match and/or to their own family members.

Only this week there was a re-showing of the documentary about the reasons why the police took so long to catch the Yorkshire Ripper, mistakes that cost lives. Everyone makes mistakes, but nobody should make careless mistakes. Following a logical process and making objective decisions is the best way I know to avoid bias.

 

[Pauline]

So I have been mulling over everything here overnight and this is where I am now at.

Firstly, because my 3rd cousin E shares almost as much DNA with A as I do, I am presuming that I can focus my investigations in the part of my tree which I share with E. I am also thinking that E and I will presumably share essentially the same relationship with A - although maybe removed to a different extent?

Secondly, although I have made guesses as to the most likely possibilities here, this is not to the extent of making definite assumptions, only about investigating the more likely possibilities first. I am of course conscious of the need to respect A's feelings here, but E has already approached me about this, and it won't be an issue with anyone I am likely to mention it to in my close family.

The children of single mothers were often adopted or brought up by grandparents as their own, so they won't necessarily be obvious even if they are in your tree.

DNA evidence doesn't really support the possibility of an adoption into this particular line in my tree, and it backs up my documented tree at least as far back as my 4 x gt grandparents (subject to my gg grandmother having an unknown father).

However, I appreciate that documented ancestors could still be supported by DNA if an ancestor's documented parents were actually their grandparents, but this is only possible if the documented parents were old enough to be grandparents and if there was a whole "sibling" who could instead be their parent. Neither is a possibility with any of my ancestors in this particular line of my tree until I get back to my 3 x gt grandfather, where it is just about feasible for him to be son of his oldest sister.

Anyway, I will now start working through all the possible relationships as objectively as I can, but starting with the more likely and bearing in mind all the available DNA evidence.

 

[peter]

Firstly, because my 3rd cousin E shares almost as much DNA with A as I do, I am presuming that I can focus my investigations in the part of my tree which I share with E. I am also thinking that E and I will presumably share essentially the same relationship with A - although maybe removed to a different extent?

You and your sister have a long list of possible relationships; you can do the same for E (E's list may be a little longer)

If E is your 3rd cousin then E is the same generation as you, so if you're a generation younger than A, then E must be too. So the number of removes will be the same.

 

[Pauline]

E's list may be a little longer

Yes, there are one or two extra possibilities for E but essentially it doesn't really make that much difference to things.

Even so, although I have identified all those known ancestors which I could theoretically share with A (from her tree and mine) easily enough, identifying where the actual link might have occurred is a task of a much bigger magnitude, which may not lead to a successful outcome. For while the liaison which makes us related has to involve an ancestor of at least one of us, it could have been with someone one or two generations (or more) removed from an ancestor of the other.

As far as my tree goes that is not too much of a problem as I am looking in a specific part of my tree and have a reasonably broad knowledge of the descendants of those particular ancestors. The same doesn't apply in A's tree, though, and trying to identify all the likely descendants of all her feasible ancestors, particularly those as yet unknown, would be an absolutely huge task.

Being realistic here, it looks to me to be something of a needle in a haystack search - with the added complication of knowing that even if I were able to make reasonable progress with this, the most likely outcome is still going to be a long list of possibilities without much hope of narrowing it down.

So even after doing all this, it seems there is a good chance I would end up still not knowing how we might be related. Or am I missing something here?

 

[peter]

Shared matches are the key - you've previously said that you know which part of your tree you should be looking, but it seems you haven't done the same for A.

 

[Pauline]

Shared matches are the key - you've previously said that you know which part of your tree you should be looking, but it seems you haven't done the same for A.

As mentioned above, I have already looked at the matches I share with A, but they haven’t given me any clues as to whereabouts in A’s tree to look.

 

[jorghes]

Have you checked A's tree in comparison with any of the other shared matches with A to see if you can spot any similarity in those trees - i.e. attempting to figure out A's relationship with any of them?

I have a similar issue - one of the test results has two relations with whom we had no known relationship and didn't have any shared links with any of the results that I knew the link. When I searched their trees, I was able to locate their relationship to each other and am currently searching through the most likely portion of their trees by constructing my own version of it, to see if I can find where our lines might overlap.

 

[Pauline]

Have you checked A's tree in comparison with any of the other shared matches with A to see if you can spot any similarity in those trees - i.e. attempting to figure out A's relationship with any of them?

Hi Jorghes, thank you for the input. Yes, I've been through all the available trees - forwards, backwards, inside out and upside down! I've also been checking and extrapolating trees to help with this.

One of my sister's shared matches with A (D) is a 3rd cousin of A, but none of the other matches my sister shares with D are related to A, nor can I work out how any of them are related to me and my sister. I have very thoroughly checked out the connection between A and D but it doesn't seem to be where I connect to either of them.

I haven't been able to identify how any of my other shared matches are related to A. With two of the matches I share with A, I have been unable to identify how they are related to me, A or any of the other matches in the shared match list, but by extrapolating the trees of these two unidentified matches, I can see how they are related to each other.