Scale of time and space in admixture

Screenshot 2016-09-30 08.23.21As you may know in Britain there is a new direct to consumer genetic testing service, Living DNA. Debbie Kennett has a post up where she talks about how it works and why it’s different. For now it is British focused, and leverages haplotype-based methods with the PoBI database to give really fine-grained analysis to their customers on those sceptered isles.

Brought to you by the same people who brought you FineStructure, this is a major offering in this space.

But there’s a more general issue I want to comment on. In the video accompanying the website, one of the presenters states that people are “very surprised how admixed” humans are. This depends on scale.

Let’s start at a time, 0. In the short term you are not very admixed usually. Outside of the Americans and other settler societies admixture is not common on the scale of recent generations. But as you go further back, you become quite admixed. E.g., The Geography of Recent Genetic Ancestry across Europe showed lots of admixture on 1,000 year scales within Europe. David Reich’s lab has shown lots of admixture between very diverged populations on a 5,000 year scale. But as you go further back the ‘admixture’ gets lower and lower, and at some point you hit the species barrier, and you see the same genealogies coalescing again and again through the bottleneck.

Now, look at space. As distance is ~0 you’ll see lots of admixture. But as you push further and further away, the admixture drops. At some point the admixture is quite boring. Ergo, all the “you are 100% European” results from DTC companies.

The two dimensions look quite different. Admixture increases as you go further back in time…until it doesn’t. You’re hitting the “species” taxonomical barrier. In terms of space, admixture decreases the further you go away from your focal point of interest.

Posted in Uncategorized

Comments are closed.