My long-term reader will be aware that I very much enjoy chasing down my various genealogical connections, including those that are revealed to me by DNA evidence. About two years ago, a new connection popped up on Ancestry.com – someone with a rather more exotic name than my usual Anglo-Celtic relatives; let’s call her Zenobia.

What was especially interesting was that she was literally the second closest of all of the DNA connections that has turned up on Ancestry.com – the strength of the connection was such that she was most likely my second cousin. (Or possibly a half first cousin once removed, or a first cousin twice removed, or theoretically great-great-great-aunt or great-great-great-niece, but those last seem unlikely.)

From the other connections who we had in common, it was fairly clear which of my sets of great-grandparents had contributed to Zenobia’s DNA – and she was not connected to my other grandparents, which rules out a number of possibilities. So she had to be my second cousin, or just possibly the child of a second cousin with an unusually strong DNA link to me.

I know that side of the family quite well, and there was no relative known to me who fitted Zenobia’s profile. There are obviously a number of possible explanations for this. I sent speculative “are you my Ancestry.com connection?” messages to everyone with Zenobia’s name who I could find on social media. Perhaps unsurprisingly, I got no replies. I filed it all under “unexplained”, and moved on.

Then a year ago, there was a new development. One of my second cousins on that side of the family, who we will call Lucy, the daughter of my parent’s cousin Quentin, also got her DNA analysed by Ancestry.com. Her DNA turned out to be connected to mine at exactly the level you would expect for a second cousin, so no surprise there – but her connection with Zenobia was that of a first cousin, not a second cousin. So that made it look likely that both Zenobia and Lucy had the same grandparent.

The connection between Lucy and Zenobia is not close enough for them to be half-sisters (let alone full sisters), so Quentin could be ruled out as Zenobia’s father. Quentin has two sisters and one brother, Eddie. Given that it is more likely for a man than a woman to have an unknown child, my working hypothesis was that Zenobia was Eddie’s long-lost daughter. I hope that this chart adequately illustrates the relationships as I imagined them.

I do not know Eddie well – he lives on a different continent to me and we have met only a few times – but I know him well enough to propose a catch-up Zoom call. After an exchange of pleasantries I told him what I had discovered. He was startled, to put it mildly, and said (I paraphrase) that for Zenobia to be his daughter, certain events would have needed to take place, and he has absolutely no recollection of any such events occurring. We left it there for the time being, a bit awkwardly.

The mystery has now been resolved. I heard yesterday from Zenobia herself, responding to one of the social media messages I had sent months ago. She said that she had been very puzzled by the DNA results, because her ethnic background is exactly what you’d expect from her real name, and very much not the Anglo-Celtic heritage that I share with Eddie, Quentin and Lucy.

However, she has had a bone marrow transplant.

If you have a successful bone marrow transplant, all of the DNA in your circulatory system including blood and saliva is completely replaced by your donor’s DNA. Your solid organs, skin and bone keep your original DNA, but most DNA tests use saliva or blood. This includes Ancestry.com, who actually recommend that those who have received bone marrow transplants should not use their services (a warning that Zenobia presumably ignored).

I contacted Eddie immediately to let him know. He was understandably relieved. Zenobia’s bone marrow donor must have been either one of Eddie’s own children or one of his sisters’ children. (The DNA link with Lucy is not strong enough for it to be her or a sibling.) Eddie himself thinks it is likely to be one of his own numerous adult children, and for what it’s worth I agree with him, but he has not yet checked with them directly. All in good time.

The fact that I didn’t know that one of my second cousins is a bone marrow donor is considerably less earth-shattering than the now disproven possibility that I had identified Zenobia as a long-lost relative. As I said up top, DNA doesn’t lie, but it may not tell the story you think it is telling. It’s a salutary lesson in not jumping to conclusions before you have all the facts.