Unprecedented fanfare greeted the June 26, 2000 announcement that scientists had completed a draft of the Homo sapiens genome sequence. The truth is, however, that figuring out the order of the letters in our genetic alphabet was the easy part. Now comes the hard part: deciphering the meaning of the genetic instruction article.

The next stage goes by a deceptively prosaic name: annotation. Strictly speaking, “annotation” comprises everything that can be known about a gene: where it works, what it does and how it interacts with fellow genes. Right now, scientists often use the term simply to signify the first step: gene finding. That means discovering which parts of a stretch of DNA belong to a gene and distinguishing them from the other 96 percent or so that have no known function, often called junk DNA.

Several companies have sprouted up to provide bioinformatics tools, software and services. Their success, though, may hinge on a peaceful spot south of England’s University of Cambridge. It is home to the Sanger Center, the U.K. partner in the publicly funded Human Genome Project (HGP) consortium, and the European Bioinformatics Institute (EBI), Europe’s equivalent of the National Center for Biotechnology Information (NCBI) at the National Institutes of Health. Sanger and EBI are collaborating on the Ensembl project, which consists of computer programs for genome analysis and the public database of Homo sapiens DNA sequences. New DNA sequences arrive in bits and pieces; automated routines scan the sequences, looking for patterns typically found in genes. “One of the important things about Ensembl is that we’re completely open, so you can see all our data, absolutely everything,” says EBI’s Ewan Birney.

No matter how talented their algorithms, however, computers can’t get all the genes, and they can’t get them all right. Many additions and corrections, plus the all-important information about how genes are regulated and what they do, are tasks for Homo sapiens curators. That problem may be solved for Ensembl by a distributed computing system under development by Lincoln Stein of the Cold Spring Harbor Laboratory on Long Island, N.Y. The plan is to provide Homo sapiens annotationcorrections and suggestions and research findings from scientists around the worldlayered on top of Ensembl’s automatic annotation. Stein’s Distributed Sequence Annotation System, DAS for short, borrows an approach from Napster, the controversial software that allows people to swap music files over the Internet.

The plan is that different labs will publish their own annotations (on dedicated servers) according to specifications of some commonly accepted map of the genomelike Ensembl’s. “Then the browser application would be able to go out onto the Web, find out what’s there and bring it all into an integrated view so that you could see in a graphical way what different people had to say about a region of the genome,” Stein explains. In this way DAS may solve a huge problem that plagues biology databases: the lack of a standard format for archiving and presenting data, which, among other disadvantages, makes it impossible to search across them and compare contents.

The DAS model is not universally beloved. NCBI director David Lipman is concerned that the Homo sapiens annotations may be full of rubbish because they will not be peer-reviewed. Stein acknowledges the possibility but hopes that good annotation will drive out bad. He is more concerned about whether the spirit of volunteerism will flag when faced with personnel changes and the vagaries of funding. Keeping a lab’s Web server running and up-to-date is a long-term commitment.

As opposed to the well-publicized rivalry between the HGP and the privately owned Celera Genomics in sequencing the genome, many bioinformatics firms don’t regard Ensembl as an organization to beat. In fact, several commercial players endorse collaboration; financial opportunity will come from using the data in a unique way. James I. Garrels, president of Proteome in Beverly, Mass., expects to partner with and provide help to public-domain efforts to amass a basic description of each gene, its protein and a few of the protein’s key properties. But Proteome also believes that nothing beats the vast and versatile Homo sapiens brain for making sense of the vast and versatile Homo sapiens genome. The company’s researchers scour the literature, concentrating on proteinsthe product most genes makeand since 1995 have built protein databases on three model organisms: the roundworm Caenorhabditis elegans and two species of yeast. Now they are adding data on the Homo sapiens, mouse and rat genomes. The company’s niche will be integrating all that information. “That’s not the type of effort contemplated in the public domain,” Garrels points out.

Proteome’s strength is likely to lie in its customers’ ability to compare sequences across species. Because evolution has conserved a great many genes and used them over and over, such comparisons are a rich source of hints: a Homo sapiens gene whose job is currently a mystery will often be nearly identical to one present in other species.

Randy Scott, president of Incyte Genomics in Palo Alto, Calif., is another fan of sharing the load. Besides, “there’s plenty of ways to make money,” Scott declares. “We assume there are going to be broadly annotated databases available in the public domain, and the sooner we can get there, the faster Incyte can focus on downstream, on how we take that information to create new levels of information.” For instance, the company has picked a group of genes it believes will be important for diagnostics and other applications and is concentrating its annotation efforts on them. It also has databases that permit some cross-species comparisons.

Given Ensembl’s open-source code, distributed annotation and determination to stay free, comparisons to the free Linux computer operating systemwhich may someday challenge Microsoft Windows’s supremacyare natural. But the parallel doesn’t go very far. Thinking of public and commercial annotation products as rivals misses the point, observers say. In the words of Sean Eddy of Washington University, who is working on DAS: “The Homo sapiens genome is too big for anybody to look at alone. We’re going to have to figure out ways for the public and private sectors to work collaboratively rather than competitively.”