More Genomes, But Shallower Coverage

See allHide authors and affiliations

Science  28 May 2004:
Vol. 304, Issue 5675, pp. 1227
DOI: 10.1126/science.304.5675.1227a

For the human genome, nothing less than perfection was acceptable: Each DNA base was supposed to be correctly identified and in its proper position. But for a host of other mammals now in sequencers' sights, perfection may be too slow and expensive.

The National Human Genome Research Institute (NHGRI) is considering a proposal from a new advisory committee to turn its sequencing centers loose on decoding the DNA of a dozen or so mammals. Elephants and bats are among the top candidates, all of which will be based on how distant they are from humans and one another on the mammalian family tree. But in a policy shift that is being hotly debated—and could be approved as early as next month—the centers would make only a quick pass at the sequence of each species. For the human and mouse genomes, repeated passes identified each base seven times or more. Now the plan is to drop this 7x coverage to 2x. As a result, the new data would consist of thousands of small pieces of DNA too disjointed to put back together in a whole genome sequence.

An important reason for taking this approach is money: Less coverage means lower cost, and that means researchers will get four genomes for the price of one. “You have the possibility of sequencing lots of organisms quickly,” says Edward Rubin, director of the Department of Energy (DOE) Joint Genome Institute in Walnut Creek, California.

Next up?

Bats are among the leading candidates for sequencing.


But others worry that there just won't be enough data to work with. Michael Lynch, an evolutionary genomicist at Indiana University in Bloomington, says, “You are just going to be left with a hodgepodge of data.” Adds Maja Bucan, a geneticist at the University of Pennsylvania in Philadelphia, “I don't think that 2x is good enough for the kind of biology that we want to do.” Without extensive data, researchers will have trouble trying to understand a genome's overall structure.

Although the new approach could produce a hodgepodge, at least it will be extensive enough to compare with the more fully detailed human, mouse, and rat genomes, says developmental geneticist William Gelbart of Harvard University. And researchers will be able to find new genes and regulatory regions that they can't find when they compare just a few sequences. Already, the analysis of the poodle genome, which had just 1.5x coverage, demonstrated that skimming a genome can yield useful information (Science, 26 September 2003, p. 1898). Some researchers even argue that they could get away with sketchier coverage. “The idea is to find the sweet spot where there's enough information but [the genome] is not too expensive,” Gelbart explains.

Having multiple genomes to compare will pay big dividends, says NHGRI Director Francis Collins. For example, researchers should be able to find hidden regulatory elements common to most species. Besides, Collins argues, 2x is just “a starting point.” Lynch and others wonder, however, whether sequencers will ever complete these genomes, once the quick-and-dirty approach has yielded its initial trove of data.

Navigate This Article