F1000—or Faculty of 1000—is an online publisher of life science articles that appear in databases called F1000Prime and F1000Research. Recently I Skyped with the company’s Vitek Tracz, a scientific publishing pioneer and founder of F1000 and previously of BioMed Central and Current Opinions journals. Tracz discussed why he started the business, problems with peer review, how F1000’s approach has affected scholarly publishing, and their upcoming launch of a comprehensive workspace for scientists.
Tracz told LJ that while the organization has decided not to change the name, F1000 “is now faculty of 10,000, and it’s continuing to grow.” After the success of Current Opinions, he and other scientists decided to create something similar for the Internet age, going about it by finding the top faculty in the world in “big subjects”—cardiology, cell biology, immunology, for example—and asking each one to name the top ten categories in the relevant subject area and the ten best experts in those areas.
F1000 has in this way gathered around 5,000 high-level experts (“All of the recent Nobel Prize winners in biology and medicine are involved,” says Tracz), each of whom is asked to invite younger scientists to work with them, and who are the drivers behind the company’s F1000 Prime recommendation engine. “Young experts are quite different from older experts,” says Tracz. “They are more active, they read more, they see more, [whereas] older experts can apply their judgment better.” Together these professionals are asked to notify F1000 when they discover an interesting article and state why they find it of value.
Faculty who contribute to F1000 are unpaid, explains Tracz, but they gain access to the top literature in their fields as well as to F1000 and “over time membership became quite a prestigious thing to have…and they do it because they think it’s a benefit to the community.” The beauty of the system, Tracz remarks, is that “unlike traditional measures of article or journal success, such as impact factor or citation counts…users can see the names of people who recommend an F1000Prime article and why they did so.”
After a period of time the organization decided to start an additional project, F1000Research, as it recognized that while open access was well established and already a great success, it only went partway toward solving some of the problems related to traditional scientific publishing.
Peer review is mortally sick
The first problem identified by Tracz with scientific publishing, even when it is open access, is timing. It is “unbelievable and ridiculous,” he says, “that science information should become available to people who want to know it with a year delay.” Secondly, according to Tracz, “peer review has become mortally sick, and the reason for the sickness is secrecy…it just begs for misuse.” Anonymous peer review involves a conflict of interest, says the F1000 founder. Since peer reviewers are experts in the field of the papers they review, they can be in competition with the authors of those papers and therefore may have an interest in delaying publication. Also, since scientific subjects are quite narrow, it is a problem that the final decision on whether or not to publish papers lies with editors. They are usually not experts in the subdisciplines of the papers they see, and an editor’s main interest, says Tracz, is the journal—in increasing its impact factor, for example—not necessarily disseminating sound scientific findings.
Tracz explains that when a paper is submitted to F1000Research, “it first goes through a cursory hygiene check to make sure it is not obvious nonsense,” continuing, “if it’s not science we won’t publish it, but otherwise we publish papers immediately, everybody can see them, they get a DOI [digital object identifier], and they are searchable and citable.” Peer review happens after publication, and F1000 has convinced PubMed and others to accept a new form of citation, “a dynamic citation.” When a paper is published, the citation will say “awaiting peer review,” for example, with the review status being updated at various milestones. The paper is then indexed in PubMed after it passes the peer review stage.
Also revolutionary is that authors choose their peer reviewers from a list of possible people, with some restrictions—they can’t select someone who has a conflict of interest, such as a scientist who collaborated on the work. Tracz remarks that sometimes an author can’t find a suitable reviewer, especially if the field of the paper is very narrow, and they may then suggest a reviewer.
Reviewers have three options: they can approve a paper, approve it with reservations or questions, or mark it as “not approved.” In the latter two cases, Tracz explains, the referee must write a signed report explaining the decision, and these documents are citable. The author can respond, and so it goes, back-and-forth, with any new versions of the paper getting version numbers and new citations. “Our referees responses are quicker than I ever experienced before,” says Tracz, who also notes that since the whole process is open and reviewers know their evaluations may be cited, comments are “polite and useful.” These narratives often include observations on what’s going on in the field and are “a fantastic source of commentary.”
Reproducibility of the work described in F1000Research papers is significantly increased as authors must also provide their data in a form that is accessible and usable. The data is also part of the peer review.
A workspace for science
In May, F1000 will be releasing a suite of software and services for scientists to write, annotate, share, and discuss scientific literature of various types—papers, lectures, explanations, etc. In beta testing now (see vimeo.com/111517635) the suite will include a web-based platform and a powerful plugin built on top of Microsoft Word, as, according to Tracz, most scientists use that program. These authoring tools are accompanied by literature discovery mechanisms, one of which Tracz describes as “a bit of magic that’s quite amazing”: scientists can annotate any website and save it to their workspace for their collaborators to see. Other materials, journal articles for example, can also be annotated and shared.
In addition, the system uses an algorithm that allows scientists to receive beneficial reference suggestions as they write. A reference tool is also available, along with a publishing one that adapts a paper to the format of a journal to which it is being submitted, with the system preserving the guidelines of major publishers. Activity is securely stored, so that a group of researchers can, for instance, look back at previous iterations of a paper or other output. If the scientists are publishing their work with F1000, they will eventually be able to do so with one click directly from their workspace.
Tracz defines his motivation for all this work: “Researchers themselves should be in control of the whole process of communicating science. Communicating science is the center of science.” He concluded the discussion with a lofty aim: “We believe that we are trying to change science publishing to science services.”