Two years ago, outspoken U.C. San Diego bioinformatics professor Richard Belew told Science, "In an emerging field like computational biology -- where it seems likely that we will have the worst of both software and biomedical intellectual property issues -- the stakes are enormous. Genomic projects are already generating huge wealth; active academic researchers ... will have great influence on how these assets are ultimately distributed."

For scientists who work with genomes and proteins, possibly the most radical position they can take is that their research is for the public good, and therefore their data should be available in the public domain. The problem is, few members of the general public are well-trained enough to appreciate the value of a software program that, for example, aligns your cDNA sequence to a gene or set of genes. Nor would many care to use their computers to predict the secondary structure of a protein.

Thus, for many scientists, putting data in the public domain means sharing it with other scientists. Doing this might mean placing your newly discovered protein structures in a public database. Or, if you want to publish your findings, it could mean working with Michael Eisen on the Public Library of Science, a free, peer-reviewed scientific publishing project the U.C. Berkeley professor initiated to combat the problem of having to buy hundreds of prohibitively expensive science journals in order to "share" knowledge with his peers. Like many in the life sciences, Eisen dislikes the commercialization -- and, for all practical purposes, privatization -- that occurs when scientists place their valuable findings in journals that other scientists can't access because their universities or labs haven't got a subscription.

It's this same problem that led European Bioinformatics Institute team leader Ewan Birney to announce emphatically at a recent conference, "I have vowed never to publish in the journal Science." Birney -- whose work is globally renowned -- is one of the founders of the Open Bioinformatics Foundation, a nonprofit that advocates open-source software development in genomics and serves as the umbrella organization for several popular projects.

Open-source development -- in which software code is by definition made publicly available to all -- has become a thorny issue for many scientists whose contracts stipulate that all their intellectual property belongs to the university or company that employs them.

Chris Dagdigian, a life sciences technology consultant in Boston who builds genome data-crunching clusters of computers running Linux-based operating systems, explains that for many researchers, developing open-source tools is crucial because otherwise each piece of software has to "reinvent the wheel" and may cost the developer several days of valuable work when she could be gathering new data. While it's common to hear that private companies don't want their employees to "give away" software by working on open-source projects, what most people don't realize is that the problem is even more pressing in academia.

The Bayh-Dole Act, passed in 1980, allows universities to claim ownership of patents on academic discoveries made using federal funds. What this means is that the university, as well as the government, stands to profit from what researchers discover. As a result, universities have set up technology transfer departments whose sole purpose is to determine which discoveries the university is going to patent and market to industry. In other words, universities have become minicompanies, wanting to keep all their scientists' research proprietary in case it might turn out to be lucrative.

Licensing issues for software development have long been a focus of conflict between programmers and software companies. But the immense financial bonanzas that many believe will accrue to biotech companies -- which by necessity are heavily involved not just in software development but also in research that could affect human health -- is making such license issues all the more pressing.

Cynthia Gibas, a bioinformatics professor at Virginia Tech, is very concerned about the way her university and others are using the Bayh-Dole Act to keep her colleagues from contributing to open-source software projects. Although her department allows her to work on open source, she says, "I'm interested in this because we don't have a uniform policy about this across the university. The university doesn't train us in intellectual property issues. Many researchers have to deal with labs where they believe that anything that goes out the door is a potential loss of profitability."

With so many people going into bioinformatics, she notes, it's going to be difficult for them to develop decent software if they're not allowed to participate in already-existing open-source projects. Gibas may be "a commie pinko from way back," as she puts it, but she's also a pragmatist. "I think universities should be allowed to get some butter from these patents because they're so underfunded anyway. But I'm worried when the attitude becomes 'let's grab everything.'"

One solution that Gibas and many other academics favor right now is creating an open-source contract, which would allow academics (and possibly people in industry, too) to make a formal agreement with their employers that allows them to contribute to open-source projects. A group called Openinformatics.org is trying to organize academics around this issue, proposing a number of different kinds of open-source contracts.

Perhaps the only academic to negotiate a blanket open-source contract with his university so far is computational biologist Steven Brenner at U.C. Berkeley, who says that it took several months and hundreds of dollars in legal fees to come to an agreement with the U.C. technology transfer office.

"I just wanted them to sign a document saying I could modify the copyright [license] on my work and contribute to open source," he says. Eventually, after a great deal of explaining, the tech transfer office came around. Now Brenner is free to contribute to a project called "bioperl" -- which he helped found -- as much as he likes.

Open-source contracts may be the most elegant solution for scientists who want to share their work. "I know many people at U.C. Berkeley produce open-source semi-illicitly," Brenner says. "And I've been contacted by many people who want the [open-source contract] arrangement."

A more radical response to privatization of the genome and bioinformatics software would be to open up the entire scientific process. This is exactly what Jeff Bizzaro, founder of Bioinformatics.org, proposes to do. His site, which has attracted thousands of members from across the globe, hosts several open-source bioinformatics projects. But the site isn't just about software. It's also about making the process of scientific discovery public and collaborative. Bizzaro encourages scientists to post their findings on the site so that, for example, two scientists could conduct complementary experiments halfway across the globe from each other.

Bizzaro and his cohorts at Bioinformatics.org believe that the scientific process has become too competitive and proprietary; ultimately we should share data in the same way a political progressive would suggest we share wealth: democratically, openly and with ethical integrity.

Recent Stories