Gerry Rubin is one of the best-respected Drosophila geneticists. He has published many seminal papers on the functions of fruit fly genes and he led the international project to sequence the Drosophila genome. Rubin now publishes most of his articles under Open Access. Open Access Now talked to him about his role as an Open Access advocate at the Howard Hughes Medical Institute, where he is a vice-president and Director of the Janelia Farm Research Campus.

Rubin the Funder
HHMI is setting a pioneering example for other funding agencies in its support for Open Access publishing. As one of the largest philanthropic institutions in the world, supporting over three hundred distinguished investigators across the United States, HHMI now has an official policy of paying the additional costs of charges for Open Access publishing up to an initial limit of US $3,000 per investigator per year. That represents a potential budget of around $1 million.

"We are committed to the general process and would be happy to shift the money that our investigators now spend on journal subscriptions to upfront charges", says Rubin. "As a funder I would rather pay up front for the publication of the work that we supported and make it available to everybody. We want anyone to be able to see the results of that work. We don't want them to have to pay some other third party - who had nothing to do with funding the work or doing the work - a large sum of money to see the results of the work. That doesn't make sense."

"The funders need to recognize that publication of the results is a valid cost of doing the research. I mean, we pay for it anyway. We often pay for the publication as page charges, and we pay to buy subscriptions to journals. People are now used to paying for color illustrations and reprints. It would be better if the people that did the work paid for its dissemination. There has to be a business model where publishers can cover their expenses." Rubin feels that charges to the author of up to several thousand dollars per article would be reasonable. "The work costs so much more than that." He agrees that it might be difficult for publishers or societies that rely exclusively on journals' prices to survive or make a profit. But he questions the justification for learned societies that cannot attract enough membership fees and other income.

We want everyone to be able to see the results of the work we paid for

At the moment HHMI is encouraging its investigators to publish in Open Access journals by paying any 'author charges'. But Rubin says they aren't yet ready to make Open Access an obligation, partly because there aren't yet enough high quality Open Access journals. But he says HHMI is doing other things to support Open Access publication, such as ensuring that the research evaluation process is not biased towards the traditional prestigious journals. "People should really judge the papers by the content rather than the journal. We give our reviewers all sorts of instructions and criteria and we remind them that we are judging the quality of the work. We ask investigators to send their five most important papers, and the reviewers actually read those papers. So they are not really counting the number of papers in certain journals."

Rubin and HHMI are encouraging other funding agencies to adopt similar Open Access policies. "I know that the Max Planck Society is very interested and strongly in favor of supporting Open Access", he notes. And he claims that most researchers are sympathetic to these issues, but have reservations about the effects that publishing in prestigious journals has on their careers. A fairer review process will obviously help. "And there needs to be a few more journals like the Journal of Biology (published by BioMed Central) and the forthcoming PLoS Biology that have some stature - so that people don't feel that their careers will be hurt." Rubin acknowledges that the transition will take time. "It is the beginning of a process. We needed a uniform definition of what Open Access publishing is - so that not everyone can claim that just by making the article available online six months after publication that it is Open Access."

Defining Open Access
To this end, Rubin hosted a meeting in April of over thirty academics, editors, publishers, and number of funding organizations at the HHMI headquarters in Chevy Chase, Maryland. The purpose of this Open Access Publishing Workshop was to bring together participants from key constituencies of the biomedical research community to discuss the principles of Open Access publishing and the concrete steps that can be taken towards promoting the rapid and efficient transition to a fully Open Access scientific literature.

"The aim of the meeting was to come up with an accepted formal definition of 'Open Access Publication' and a set of recommendations for implementation", says Rubin (see the box for the meeting's definition of Open Access). "Harold Varmus (President and Chief Executive Officer of Memorial Sloan- Kettering Cancer Center) and I agreed to synthesize the comments into a form that is somewhat of a consensus. It's definitely a compromise consensus and it is unlikely that any one participant would agree with every word. There is something here for every participant to dislike. But we tried to steer a middle-of-the-road course, striking a compromise between the 'purist' view and ways that make it more acceptable for other publishers to join in." This was obviously a difficult task and some of the participants have chosen not to have their names associated with the document. The 'purists' feel that all scientific reports should be available immediately and free-of charge in a centralized repository that allows full-text searches and freedom to download and disseminate any data. But publishers have concerns about their subscription income and copyright agreements.

Rubin stresses that each participant attended in a personal capacity and not necessarily as a representative of their institution. "The purpose was not to be the final word on the definition but to propose something that is written out in black-and-white in enough detail that people can actually discuss what it means. We needed a definition and a set of criteria that we can agree upon and we have made a good stab at that. It needs further discussion and finetuning. And we got a view of what practical things we can see librarians, institutions, scientists and journal editors doing to move forward. Hopefully it will also get people to begin to talk about it. A lot of people are really not aware of it."

To me the most important reason for Open Access is for data mining

Rubin the researcher
Many high-profile scientists are sympathetic to the need for Open Access publication. But Rubin is one of the few who actively publishes his work in Open Access journals. In the last year he has published more than ten articles in BioMed Central journals. These include a report of the finished Drosophila genome sequence in Genome Biology, and the maiden paper in the Journal of Biology last year. The latter describes clustering of coexpressed genes in the fly genome and has been downloaded more than a staggering 17,000 times from the journal's website. "Certainly I didn't lack anything on visibility. A lot more people knew about that paper or saw it than would have if it were in Science or Nature or wherever."

"I would say that 80% of the papers I have published in the last two years have been in Open Access journals, but I tend to let the first author choose where they want it to go", says Rubin. "I think it's inevitable that we switch to Open Access. But there is still an uphill battle to be won. We're trying to do what we can to speed it up. There are more and more people, like me, who are committed to it and try to steer the papers in that direction. But we do need more alternative journals and more people who will take that viewpoint."

For Rubin, the reasons to support Open Access publishing are obvious. "To me the most important reason for Open Access is for data mining. Let me give you an example. Suppose someone writes a paper about a specific set of genes and microarray experiments. And they write a comment somewhere in the middle of the Discussion section that one of the seventy genes that they have identified appears to be a ortholog of the Drosophila whatever gene. There's currently no search that could reveal that to me; I would need a search engine that could perform complete full-text searches. It's just like DNA sequences - the reason why it's really good to have everything in the GenBank and EMBL databases is that everything's there. If it's been sequenced, you'll be able to find it there and you don't have to search other places for it. If the same were true for the literature, I'd be able to easily find data and results that are relevant to my work. It's the biggest thing that you can't do now".

Rubin thinks that wider acceptance of Open Access will be driven by the availability of search engines. "More and more people will be finding information by these kinds of searches. If you do a search and you get your competitor's paper but you don't get your paper because the publisher wouldn't let the text be searched by this engine... then they will see that the work doesn't get cited." He concludes that "We need more people thinking about what all this means for their ability to access the scientific literature and to carry out research in the biomedical sciences."

http://www.hhmi.org/