I'm wondering how to use information derived solely by DNA testing when compiling a family tree.
My apologies in advance if the example below is a little complicated.
I know that I (say person A) am related at third cousin level with person B. We have both done DNA testing and on shared DNA matches we have a common 4th cousin (say person C). C has other close relatives with whom we also share DNA (also shown as 4th cousins to A and B).
Person C is totally unknown to A and B and have no shared surnames. Person C's tree goes back to about the same time period as the shared ancestor of A and B (born 1830's) and we have no idea of any ancestors prior to that.
It must be that a marriage occurred sometime early to mid 1800's which connected the respective families. I have found one such marriage on church records which could be the link but at this remove I don't see any easy way of finding conclusive proof.
My question basically is: would you rely on DNA evidence to build an extended family tree when there are no other means of proving the exact relationship? If so, how would you represent them in the family tree.
It's possible that there is no marriage proving the relationship....these things have been known to happen!
I wouldn't put those people into my family tree yet. I'd make a note on the nearest likely common ancestor that person C shows a DNA link but hasn't been proven with records.
That was my original view also but as I do more and more research of my DNA matches I am amazed just how accurate the science is becoming.
In this case, there are six different DNA tests all pointing to the same result.
Whilst the science isn't yet 100% reliable to state whether a person is a 4th, 5th or 6th cousin, I guess we are heading that way. It is then I imagine that we will be looking at DNA evidence in the same way as we regard paper evidence and will include such matches in our trees as being proven.
Perhaps a separate tree in your software - to indicate that it's a DNA tree rather than a records-based one?
Yes, I think that's the best way to go. If any further supporting evidence crops up, I'll then merge the two trees.
You use it for clues and as a means of suggesting further alleyways of paper-based research. That (broadly) is all DNA is good for at this stage. The only close link I have found with my autosomal test is my first cousin once removed. He is shown as my second cousin by the DNA company (Ancestry). The science is not yet sufficiently accurate. There already is enough misinformation out there without adding to it (not suggesting that you would )
Research trees and/or mirror trees are highly recommended in the adoption search groups for people who are using DNA to identify relatives. It's always nice to have a paper trail to back it up, but some times you don't get that, at least not in a direct way. I think you could use it in your case too, as you try to figure out how these different cousins fit into your tree.
I successfully found my father's biological father using DNA plus records from Ancestry.com. In my father's case, there are no records identifying his biological father - his bio mother put down the name of another man on his birth certificate. We first had to interpret the DNA and why he was coming up 50% Iberian with a ton of Portuguese Hawaiian cousins when both his mother and the man listed as his father had strong East coast Irish heritage. We had numerous 2nd and 3rd cousins, all from the same town in Hawaii, who all related back to a family in the early 20th century who had 20 kids.
From there, I had to research the 20 kids and their kids to see if I could find someone who was in western Massachusetts in the summer of 1946, when my bio grandmother became pregnant with my father. Thanks to military records, I found a very likely candidate, and my theory was later confirmed when one of his living children took a DNA test and matched as a half sibling to my father.
If this sounds confusing, it absolutely was. It took months of record-finding and research. It took some leaps of faith, too. I was dealing with lots of people, and my research tree helped to keep it all organized and to see where everyone was fitting in. I called it a research tree and in the description box, I explained that it was for research purposes and that the information may not be 100% accurate.