The Genographic Project is now moving beyond uniparental lineages with Geno 2.0. Spencer Wells kindly invited me to a conference call last month where he outlined a lot of the details, so I’ll hit the salient points for readers of this weblog:
* They’re unveiling a new SNP-chip and a new project which moves beyond the Y and mtDNA to the autosome. But they’re also expanding their coverage of uniparental markers.
* Though there are “only” autosomal 130,000 markers, Wells and his collaborators have selected a subset of markers which are highly informative of population structure (e.g., high Fst). Their SNPs are biased toward those with moderate levels of polymorphism across many populations to maximize the power of diagnosis of differentiation.
* They tried really hard to get rid of ascertainment bias. This means that in many previous chips there is a tendency to work off the polymorphism in Europeans, and then examine worldwide variation using this ruler. The problems with this method are obvious. One of the scientists on this project outlined how they worked to look for SNPs which are very informative for populations where ascertainment bias is a particular problem, Oceanians and Amerindians. I was impressed …