o sooner have the two draft human-genome papers vanished from view than along come replacements. At #5 and #6 are papers announcing draft sequences for the species that supports more human genomes than any other: rice.

Two teams, one representing the private sector and the other public research, announced their draft sequences in the same issue of Science. Just ahead at #5 is the public-sector group at Beijing Genomics Institute (BGI), led by Huanming Yang, which sequenced the indica subspecies that is favored in China and other tropical countries. Close behind is a group led by Stephen A. Goff of Syngenta’s Torrey Mesa Research Institute in San Diego. Syngenta, an agricultural company, chose to sequence the japonica subspecies grown especially in Japan and temperate countries.

In some respects this is a rerun of the human-genome race, right down to the fact that Science, to accommodate the wishes of the Syngenta team, waived the customary agreement that full data be deposited in the public domain. But overall there seems to have been more cooperation between the two groups, with the public effort using private data to build its sequence. Especially remarkable is the speed with which the sequences were completed, notably by Yang's group. Just over three years before publication, the Beijing Genomics Institute was little more than an empty building with a stylish helix by the front door. Now it is a sequencing powerhouse.

In large measure the speed came from choosing the whole-genome shotgun approach pioneered by Celera in its effort to finish the human genome. This approach relies on powerful computers to assemble millions of sequences into a coherent whole. But it still leaves plentiful gaps, and the research community is worried that now that the rice sequence is apparently "done," funding will dry up. The challenge of finishing the rice genome has been taken up by the International Rice Genome Sequencing Project. The IRGSP arose at a meeting in 1997 at which researchers agreed to collaborate in a model based on the Human Genome Project. Its technique of mapping first, then sequencing, is slower but more accurate than the shotgun approach, and will be vital to tie the sequences to specific places on the rice chromosomes. The IRGSP expects to publish a proper sequence by the end of 2003.

Rice is, in many respects, an ideal model species. Its genome, at about 440 million base pairs, is far smaller than the 3,000 million of maize or the 16,000 of wheat. But long stretches of rice genes are mirrored in the same order in the other grasses that make up humanity's staple crops. This synteny will make it relatively easy for researchers to transfer knowledge gained from one species to others. And that may be why the private sector was interested in the first place. As Chris Somerville, director of the Department of Plant Biology at the Carnegie Institution of Washington in Stanford tells Science Watch, "It is my understanding that no one makes much money selling rice seeds." But they can make money putting information from rice to work in maize, wheat, and barley. And, says Somerville, the genome will help breeders in non-profit centers such as the International Rice Research Institute directly and by spurring work on rice as a model system.

The two papers, draft sequences though they are, shed interesting light on rice and its relationships with the rest of the living world. Both projects say that more than 80% of the genes noted in Arabidopsis thaliana, the first plant to be sequenced, have counterparts in rice. But almost half the genes (45%) found in rice have no homologs in Arabidopsis; although many of these may prove to be artefacts, some may also reflect differences between the two great groupings of flowering plants, the monocots (rice) and the dicots (Arabidopsis). On the actual number of genes, Syngenta posits 32,000 to 50,000, while BGI predicts 46,000 to 55,000. Even if the BGI numbers come down somewhat, the total may well be higher not only than Arabidopsis, at around 25,000, but also than Homo, at about 35,000.

There is a salutary message in all this, identified by Jeffrey Bennetzen of Purdue University. He points out that the rice genome continues the fine historical tradition whereby science pushes humanity further away from the center of the universe. How galling to discover that the rice on which so much of humanity depends is even richer in genes than the people who eat it. The two rice-genome teams join Copernicus and Darwin in putting us even more firmly in our place.