How many sequenced genomes are enough? The minimum number for comparative genomics, researchers say, depends on what you want to learn. The optimum number is a still a mystery.
For identifying cis-regulatory regions such as enhancers and promoters, the genomes of three species that are roughly equidistant evolutionarily is the bare minimum, and more is better, according to Lincoln Stein. Stein, who is at Cold Spring Harbor Laboratory, is first author of a paper on the Caenorhabditis briggsae genome in the first issue of PLoS Biology, an open-access online journal published by the Public Library of Science. He pointed out that having the genomes of at least three species permits distinguishing between the signal—bases conserved because they actually do something—and noise—bases conserved simply because they haven't mutated yet. “As you add more species to the alignment, the signal remains the same while the noise decreases because there are fewer and ...
