Marking a MRCA couple as DNA confirmed

Question

Hi everyone, there's something I don't understand about the DNA confirmation rules so would like to have some discussion.

In the [[Help:Triangulation]] page. section "Do you include the MRCA?", it explicitly says that if the documented MRCA is a couple, then not to mark either of the MRCA as "Confirmed with DNA".  Only the legs should be marked.

This seems sensible to me, as the match only proves that at least one of the two in the couple are a common ancestor of the people in the legs.   

However in the rules for one-to-one confirmation, [[Help:DNA_Confirmation#Which_Relationships_to_Mark_as_Confirmed]]  it says that both of the MRCA couple should be marked.

I don't see how this is any different, and why the same rationale doesn't apply.   The evidence only tells us that at least one of the two in the couple is a common ancestor, especially if it was a low cM match

For example, using the names from the diagram on the Confirmation page: supposing Helen's father is actually a David ,  and then Bob shows up with a DNA match to Helen and a bunch of strong unknown matches,  the fact that Jason was marked as DNA Confirmed to be Helen's father is going to interfere with his search .

It doesn't seem to make sense that even though a one-to-one match is less evidence than a triangulated match, it requires adding greater DNA confirmation markers.

Context - I was asked to mark the DNA confirmation of both of the MRCA couple for a 2C1R with 56cM match (the expected value for full 2C1R from DNApainter is 122cM),  I have complied but I feel dishonest in doing so.   I also have another 2C1R with a 37cM match where the same question would arise.

Comments

I have a MRCA who matches to both parents, while some match to only one I will check them with segment matching and see if there is any sense to it.

---

I have a situation where I know that my g-grandfather had 2 wives.  I also have 1 second cousin (who I knew personally until his death 3 years ago) who shares my g-grandfather but we have different g-grandmothers.  We both have had autosomal DNA tests, with results in line with expectations based on the very well documented paper trail of our relationship.

I put the DNA confirmation statement on only the g-grandfather's profile and regularly see an error indicating that the statement is missing from both g-grandmothers' profiles (I also manage my half-2nd cousin's profile), when I actually know with certainty that our relationship does not trace to either of our respective g-grandmothers.

---

Well, I had 5 people on thesame family line from MRCA only two triangulated with me and 1 other. I lowered the cm to 3 and got a couple more around 4 cm to triangulate with the 1st two so I guess everyones segments are from different ancestors. I am keeping a segment spreadsheet to see if I can identify segments to ancestors. Starting with those who have good trees and hoping I can prove no trees at a later date.

---

I think it comes down to the proved connections as posted in your statements. So, for example, if you have several sons of a MCRA and they all have different wives, then it is presumed that the MCRA and his wive can both be confirmed. Then, say you have an additional who ties in a generation farther back, then you can eliminate the wife and be fairly certain the overall is coming from the Male MCRA only. That's what I use...

Answer 1

One-to-one (for 3rd cousins and closer) depends on the total amount of shared autosomal DNA. That DNA would come from their shared ancestral couple (not just one of them).

Comments

In other words, matches from an ancestor requiring triangulation are often/usually only one segment, which could only come from one of the ancestral couple. In fact, you only triangulate one segment. However, matches from an ancestor not requiring triangulation are often/usually multiple segments, even on multiple chromosomes, leading to the presumption that some of the segments come from each of the ancestral couple.

Answer 2

Maybe I can answer my own question here using back-of-the-envelope math:

Taking the probabilities from the meioses chart in the AncestryDNA white paper,  and using Bayes' Theorem and a prior probability of 97% that the documented parents of a child are correct,  then I get 84% chance that a 56cM match for 2C1R still means both the parents are correct.  Which is probably high enough to use the word "Confirmed".

But this doesn't take into account the number of segments match (and it's not apparent from the white paper how to do so easily , without running simulations).

Comments

Matt, to confirm I share both halves of a couple with a person (let's say John Smith and Annie Green), I should see shared DNA with that person along John's Smith and Annie Green's family. Preferably each of their parents. I never assume I am related to both halves of the couple without this evidence because half relationships happen all the time. When I create DNA statements (I'll use your 2C1R example), I add the statements all the way up to the child of the couple I'm descended from and mark his parents (the couple) as confirmed with DNA. But I've done the work to know this to be true. And yes, sometimes the amount of shared DNA comes out lower than usual due to recombination. I hope that makes sense.