Recreational genotyping and ancestry
There’s a fuss at the moment in Britain over the recreational genotyping companies that purport to tell you where your ancestors came from. One of the stories that provoked this, was the claim that over 1 million Brits are descended from Roman soldiers. For example, in the Telegraph, under the headline “One million Brits ‘descended from Romans'”
The Romans departed abruptly in the early fifth century, leaving behind relics of their rule including Hadrian’s Wall along with a host of towns, roads and encampments.
But perhaps the most enduring sign of their legacy is in our genes, experts claim, with an estimated million British men descending from the invading forces.
The first sign that something is wrong is that ‘one million Brits’ turns into ‘a million British men’. What about the women? The reason for the `one million’ estimate is the same as the reason it’s just men — the ‘experts’ are looking only at male-line descent, via the Y chromosome. In fact, the number of British men descended from Roman soldiers is probably more like 25 million. That is, there’s a general principle that anyone in the distant past is either a direct ancestor of no-one in the present, or of almost everyone in the present.
If you go back 100 generations to the time of Roman occupation of Britain, you would have 1267650600228229401496703205376 ancestors. That’s roughly a bazillion times the number of people alive then, so there is a lot of overlap. If you look just at pure male-line ancestors, from whom you inherited a Y chromosome, you either have none (if you don’t have a Y chromosome) or one. It’s clear why the ancestry-mongers want to simplify their sales pitch by focusing on the Y chromosome (and on mitochondrial DNA, which is inherited through the female line), but it’s not clear why anyone should listen to them.
Even if you live in Britain and your Y chromosome came from someone in Roman legions, it didn’t necessarily come via the British occupation. After all, lots of men have migrated to Britain since then: the Vikings and the Normans back in History, and more recently from all over the world. Some of them would have had Roman-looking Y chromosomes too. And even the idea of a ‘Roman’ Y-chromosome is a bit dodgy. Broadly speaking, a group of Y chromosomes tends to get attributed to the region in the world where it is seen most today (unless that’s, say, the US). There’s no guarantee that this is where the Y-chromosome group was common 1000 years ago.
There is some potential for using whole-genome data to say something more meaningful about relatively recent ancestry, but to be useful even that needs to come with uncertainty estimates, which will often be huge.
Sense about Science have put out a good information sheet, but the basic message is that at the moment anything interesting someone tells you about your distant ancestors based on genetic information, they could tell you equally well without bothering to do any genotyping.
Thomas Lumley (@tslumley) is Professor of Biostatistics at the University of Auckland. His research interests include semiparametric models, survey sampling, statistical computing, foundations of statistics, and whatever methodological problems his medical collaborators come up with. He also blogs at Biased and Inefficient See all posts by Thomas Lumley »