RootsChat.Com
General => The Common Room => Topic started by: Catservant on Thursday 02 March 23 13:07 GMT (UK)
-
I share an ancestor with someone who according to the new 'possible relationships' feature is most likely to be my 2nd cousin. His son is, however, most likely to be the son of my 3rd cousin. I appreciate that what is most likely is not definitely the case, I'm just confused as to why the 'most likely' data doesn't match up for father and son.
Can anyone who has seen this new feature explain this. thanks
-
The feature is based on the amounts of shared DNA and the probabilities of relationships for the length of the match. For any given relationship there is a range of possible match lengths, and the probabilities of matches at that relationship level for varying match lengths can be viewed in tools such as DNA Painter - https://dnapainter.com/blog/introducing-the-updated-shared-cm-tool/
DNA inheritance is to an extent random, which is one reason why it is possible for a particular relationship to exist over a range of match lengths, and for almost any given match length, there will be an overlap of different possible relationships and probabilities.
All that My Heritage and other providers are doing when they suggest a relationship between two matches is to indicate the likeliest probability for the match length that exists. But the most likely probability isn't necessarily the actual relationship between the matches. Matches at the same length could be related in different ways at varying levels of probability, and there will always be two people who are related at the lowest level of probabilty, or that probability would not exist for the match length given.
Whilst the son will have inherited 50% of each of his parents' DNA, it doesn't follow that the 50% of the DNA that he has inherited from his father is, or includes, the entirety of the segments that are included in the match between his father and you. So if he has inherited only some of the segments that you and his father share, your match length with the son will be correspondingly lower, and the highest probability calculated for a match at that length suggests a different relationship to that expected.
But that doesn't mean that either of the two suggested relationships are correct. Either, or both could be completely wrong. You can use the suppositions to guide your research, and a tool such as DNA Painter can give additional guidance as to the expected range of relationships for each match. The definitive relationships between you are only going to be confirmed though by verified documentary evidence - the "traditional " research which links families, ancestors and decendants by BMD, census, parish and all the other historical records available.
I believe that the My Heritage possible relationships are presented in the form of a table, which suggests a number of relationships, in descending order of probability. Effectively, they are presenting similar information to the DNA Painter tool, and the fact that the highest shown relationship might, for example, state 3rd cousin with a probability of 55.8%, doesn't mean that the lowest indicated relationship of 2nd cousin's son at a probability of 2.3% isn't the correct relationship between you, or indeed any of the intervening relationships in the list. 55.8% of people sharing the same amount of DNA can expect to be related at the highest level, but 2.3% might be related at the lowest level, with varying probabilities inbetween.
-
Thanks for your detailed and helpful reply. I’m veering to 2nd cousin as they share ancestors on my grandmother’s maternal and paternal sides.
-
I’m not sure which “new feature” you are referring to, but I think Phil has just about covered it. MH have “cM Explainer” which I think is fairly new. You will find it under Tools.
I’ve found MH’s relationship estimates to be reasonably accurate for the few relationships I’ve been able to work out, but for some reason, despite us sharing 3541.9cM, they decided that my daughter was my niece. ::)
Once you have a DNA match with someone, looking at the paper trail is probably more accurate than relying on their relationship estimates.
There could be some examples of children born to unmarried daughters who were brought up by or even registered as children of her parents, (so brought up by grandparents as their child rather than grandchild) which, if descendants were unaware, may skew the paper trail and the estimated DNA relationship. Though if it’s only one generation, the further back you go probably the less noticeable the difference cM wise.
I have many examples where my daughter shares very similar cM amounts with matches as her grandfather, which is probably why relationship estimates as quite vague such as 3rd to 5th cousin.
-
I’m not sure which “new feature” you are referring to, but I think Phil has just about covered it. MH have “cM Explainer” which I think is fairly new. You will find it under Tools.
Ruskie, when you go to Review DNA Match, I believe the OP is referring to the table that now appears, named "Possible Relationships NEW". At the bottom of the table there is a link to "Show more relationships plus diagram" which opens a chart with the option to show probabilities for the MRCA.
Another new feature, basically a reworking of the DNA Painter type presentation.
-
Thank you Phil. I didn’t notice that new “feature” hiding there. :)
I’ve only had a quick glance, but it doesn’t look to offer anything different.