In the spring of 2002, the Department of Energy's Joint Genome Institute (JGI) completed the draft sequence of the genome of Ciona intestinalis, a sea squirt. A few months later JGI reported the first results of an analysis of the genome of Fugu rubripes, a puffer fish, whose draft sequence had been completed the preceding autumn. JGI also announced it had begun sequencing the genome of the African frog Xenopus tropicalis.

Paul Richardson, head of functional genomics at JGI, holds a sea squirt. To gain insight into the human genome, researchers are surveying many hundreds of millions of years of evolutionary time, seeking clues in our ancient relatives.

Why this odd array of organisms? (Not to mention other sequencing projects, including a tree, a wood-eating fungus, the house mouse, and a few dozen bacteria.) It all began with the Human Genome Project, during which JGI sequenced chromosomes 5, 16, and 19. To accomplish this, JGI installed industrial-scale sequencing machines that could determine the order of hundreds of base pairs of DNA every second--a rate that continues to increase.

Some of JGI's sequencing capacity is now zeroing in on useful or dangerous organisms like microbes. But much effort is aimed at a larger scientific question.

"The overall reason for genome work is to gain insight into the function of the human genome," says Paul Richardson of Berkeley Lab's Genomics Division, head of functional genomics at JGI. "To see significant changes in the genes, especially well conserved genes, you might have to go several hundred million years back in time to a common ancestor, something like Ciona."

Sea squirts stay put as adults, but as tadpoles they swim freely, propelled by a tail stiffened by a kind of forerunner of the backbone. Remarkably for so primitive an organism, squirts possess hearts, nervous systems, immune systems, and an analog of the thyroid gland.

If it's no surprise that many of our relatives' genes, even relatives as distant as the sea squirt, are much like our own and perform similar functions, it's a short step to the hypothesis that other similar sequences may also be genes, albeit with functions yet unknown. Indeed, the first payoff of JGI's comparison of the human genome with that of the puffer fish was the prediction of 961 previously unidentified human genes. Knowing they're there is the first step to finding out what they do.

According to Dan Rokhsar of Berkeley Lab's Physical Biosciences Division, head of computational genomics at JGI, Fugu rubripes was especially suited to this task because, odd as it seems (a deadly Japanese delicacy, Fugu fatally poisons over a dozen people a year), "almost three-quarters of human genes find identifiable counterparts in Fugu." And the genes are easier to find: Rokhsar notes that because it lacks much of the noncoding DNA found in humans, the puffer's compact genome is just one-eighth human-genome size.

There's more to comparative genomics than finding new genes. The sea squirt and puffer fish genomes are expected to yield significant knowledge of how genes are regulated and how they interact within cells. Researchers look for very short sequences of DNA that are the same in squirt, puffer, human, and other organisms. Though often widely separated from genes, these are likely to be the promoters, enhancers, and transcription factors that signal genes when to turn on and off.

Fugu Rubripes. The genome of the Fugu fish contains essentially the same genes and regulatory sequences as the human, but the genes are easier to find, because Fugu lacks much of the noncoding or "junk DNA" found in humans. Fugu's compact genome is just one-eighth human-genome size.

Beyond comparative genomics lies functional genomics. Frog researchers hope that when JGI completes the draft genome of Xenopus tropicalis, the fast-breeding frog will become the premier vertebrate model for functional genomics.

Many factors besides regulatory sequences are involved in the functioning of genes, including the phenomenon of alternative splicing, in which coding sequences in a single gene may rearrange themselves to produce quite different proteins. Hardly rare, a third or more of all human genes are estimated to have alternative splices.

In the past, functional studies have been done one gene at a time, but the massive Human Genome Project has inspired a move toward studying the complex networks of relationships by which genes conduct most of their business.

"As historic and important as the Human Genome Project is," says Eddy Rubin of Berkeley Lab's Life Sciences Division, interim director of JGI, "it's only the first step in determining how genes work--and why they sometimes don't work the way they should."

The Joint Genome Institute is jointly operated by Lawrence Berkeley, Lawrence Livermore, and Los Alamos National Laboratories. Sequencing of Fugu rubripes, Ciona intestinalis, Xenopus tropicalis and other organisms is carried out in concert with many other research groups.

-- Paul Preuss