June 13, 2007 | Sean Eddy accepted the 2007 Benjamin Franklin Award and then proceeded to poke a few good-natured holes in Franklin’s sterling open access reputation.
Eddy noted ironically that the U.S. patent office was founded (by Thomas Jefferson) the year Benjamin Franklin died in 1790. Thus, even if Franklin had wanted to apply for a federal patent, he could only have done so posthumously. Eddy also noted that Franklin was once embroiled in his own “open access dispute” in the 1730s and 1740s. When Philadelphia postmaster Andrew Bradford wouldn’t carry Franklin’s competing newspaper, Franklin engineered the firing of Bradford, and took over as postmaster, generously agreeing to carry his competitor’s newspaper. But in 1740, the first American magazines were founded, and Franklin refused to carry Bradford’s new magazine.
Four centuries later, the open access hasn’t changed so much. In 2001, Celera published its assembly of the human genome in Science, but refused to share the sequence without restrictions. Eddy said he uses Celera’s DVD as a coffee coaster, because the terms barred him from redistributing the sequence data.
In the aftermath of the controversy, the National Academy of Sciences convened the Cech Report, reaffirming the quid pro quo of science publication. “If you’re not disclosing anything, why should you get credit?” said Eddy. “All members of the scientific community ... are playing on the same playing field. [If] Celera publishes a paper, their genome sequence should be available. [If] I publish a paper, my software should not be made available to academics only, but to [everyone].”
Eddy devoted most of his talk to the topic of data dissemination and open access. “We want not just humans to be able to read the literature, but computers to be able to read the literature.” Said Eddy: “The bioinformatics community is the canary in the coal mine for this kind of issue. We’re the ones trying to integrate large datasets. If the dataset is there but not integratable, we know about it first.... The literature affects us very much.”
Eddy credited his fellow Franklin Award nominees as providing “great examples of the kind of bioinformatics access that we need.” He praised Don Gilbert for creating influential programs and databases such as a freely available biosequence search tool, as well as FlyBase. Steven Salzberg is a “vocal leader in the community” and Robert Gentleman was credited for “influential things” including BioConductor (for microarray analysis).
“For me, open access is secondary. The goal is to create systems that work,” said Eddy. “Open access is necessary but not sufficient.” Eddy made a strong case that it is worth academics taking time away from their top priority — writing papers — to “create tools and make sure that they get used by the community.” There are several keys to doing that successfully, including choosing the right problem, preparing a simple manual, making it freely and readily available, and ensuring that it is integrated into the community’s system.
Eddy is best known for a program called HMMER, for running profile hidden Markov models on sequence families. The software is easily downloadable, and runs on anything, said Eddy. There is one page of instructions for installation, another page for getting started. He also said that the creation of HMMER was a major factor in his obtaining tenure, even though he never published a paper on it.
Integration with other systems is also hugely important. InterPro is a compilation of 11 protein search tools, including Eddy’s Pfam tool, organized by the European Bioinformatics Institute to help groups doing protein domain analyses. Eddy noted that a handful of other databases use different software packages that are not readily available — which benefits him. “I get market share,” said Eddy. “My stuff becomes better known, more widely used and [has] more impact on the community.”
Curiously, Eddy noted that the International Society for Computational Biology has a policy against open source software. Some members, he said, think they should be able to make money from industry for writing software. “Some people with a similar mindset as me have infiltrated the board of directors,” and will work to change the system.
The Benjamin Franklin Award is presented by the Bioinformatics Organization. The award has been given at the Bio•IT World Conference & Expo for the past four years.
Photo by Mark Gabrenya
Subscribe to Bio-IT World magazine.