One of the things that I’ve found is that genetic tests such as Ancestry and 23andMe seem to have a difficult time discerning the status of “half-” relatives. In my case, I have an uncle on 23andMe that was classified as a first cousin because he is the half-brother of one of my parents, meaning that we share less DNA than we would if he were a full sibling. Even though his daughter also took a 23andMe test and is linked to him as his daughter, he is still listed as a first cousin on my list of DNA matches, even though the family tree seems to get it right listing him as my uncle and her as my cousin — the algorithm behind the family tree function that 23andMe provides seems to have been refined, and improved upon, with recent updates, which I like a lot…

However, had I not known that this first cousin of mine was actually my uncle because of contact with his daughter (and, again, the family tree function) I would probably never have known that someone who really seems like they are my aunt was classified as my first cousin on Ancestry. She seems to be a half-sister to my other parent if the information that I’ve been putting together is correct, is so much older than me that it seems implausible that she would be my first cousin, has a child who Ancestry has guessed is my “second cousin” who seems like he is much closer to me in age, and shares nearly the same percentage of DNA in common with me as my established half-uncle. I sent her a message on Facebook inquiring about this, because, you know, having a sibling that you might have had absolutely no idea about is a hell of a thing to miss. I’m not sure if she’ll respond. She may or may not know about this sibling. But it would really do for both Ancestry and 23andMe to refine their algorithms a bit better to correctly identify when there is a “half-” relationship like this. I’m not sure if they even can, but at least looking into it and trying wouldn’t be all bad.