What you can do with a dozen genomes

I really enjoyed the ISR Distinguished Lecture by Eric Green on Wednesday. It reinforced my excitement about the idea that having a dozen genomes will allow us to obtain qualitatively different information than we’ve been able to obtain from a single genome. In addition to the alignment-based methods he described, there is the (rather amazing) possibility of reconstructing the ancestral sequence (see Blanchette et al. 2004, a very nice paper by an all-star cast) and methods of assigning gene function based on patterns of duplication and loss (e.g. the recent paper by Li, Pellegrini and Eisenberg in Nature Biotechnology). It strikes me that with 20-30 appropriately related genomes one could deduce whether individual nucleotides within a conserved block are under selection, an incredibly powerful tool (of course, I’m thinking about ESEs). I suspect that, like many new methods, comparative genomics will yield insights in ways that will not be fully appreciated until the data are at hand. It is exciting, and it reminds me of the excitement we all felt during the late 70s, when the first sequences were being obtained.