Content area
Full Text
REVIEWS
COMPARATIVE GENOMICS AT THE VERTEBRATE EXTREMES
Dario Boffelli, Marcelo A. Nobrega and Edward M. Rubin
Annotators of the human genome are increasingly exploiting comparisons with genomes at both the distal and proximal evolutionary edges of the vertebrate tree. Despite the sequence similarity between primates, comparisons among members of this clade are beginning to identify primate-as well as human-specific functional elements. At the distal evolutionary extreme, comparing the human genome to that of non-mammal vertebrates such as fish has proved to be a powerful filter to prioritize sequences that most probably have significant functional activity in all vertebrates.
In Homers classic, Odysseus is confronted with the impossible task of steering his ship midway through a narrow strait, flanked by two rocks. The slightest veer to one side and he and his crew would be within reach of the monster Scylla, with its six teeth-filled horrible heads that can destroy anything within their reach. Swerve to the other side, and they would now fall prey to Charybdis, a whirlpool that sucks anything that comes close enough down to the bottom of the ocean1.
Odysseus quandary over the route to take evokes a contemporary dilemma that biologists face when they attempt to use comparative genomics to identify functional elements, such as genes, gene regulatory elements and other less well-defined structural components of the genome, buried in otherwise anonymous sequences. If they compare species that are too closely related, then the high degree of similarity between the orthologous sequences will obscure the functional elements within them; by contrast, if they compare species that are too distantly related, then the functional elements will have diverged too much to be readily identifiable2.Until recently, it was assumed that comparing species that are separated by moderate evolutionary times, such as non-primate mammals, would represent an ideal strategy3,4.
The principle of steering a middle course has certainly proved successful: comparative analyses of the human and mouse genomes, which diverged from each other approximately 75 million years ago, allowed much better annotation of both these genomes than would have been possible had only 1 been available5,6.Moreover,
recent comparisons of multiple moderately related species have been extremely powerful in the analysis of simple genomes such as that of yeast. By comparing the genomes of four...