A radical open-access journal is to be launched that will rely entirely on post-publication peer review.
F1000 Research is the latest initiative by the Faculty of 1000, whose large existing network of senior scientists select and evaluate top published papers in biomedicine.
The new life sciences journal will also publish non-standard research outputs such as incomplete data sets, negative results, preliminary analyses and "thought experiments", which are typically rejected by standard journals.
Submissions deemed to have passed a basic in-house "sanity check" will be posted immediately, and the journal, which will be funded by article fees, will then invite specific experts to post a review. Other readers will also be able to comment on the paper or the reviews, and authors will be encouraged to amend their papers in the light of that feedback.
Rebecca Lawrence, director of new product development at F1000, said many operational details still had to be decided, and the journal would test its approaches by posting trial articles before launching more formally later this year.
"It is crucial to have discussions out in the open so that people know who said what and can then make up their own minds," she said, adding that F1000 Research would also try to link papers to any discussion about them elsewhere on the web.
She said the journal would have to think about what limits it would set on the kinds of material it would accept, but it did not want to be overly restrictive, particularly at the beginning.
Retractions would be contemplated only in the event of misconduct or "dangerous" inaccuracies such as inaccurate drug doses.
But the journal would clearly display whether an article had been peer reviewed and, if so, whether it had been "approved". She expected researchers to self-police their submissions for quality because "if everybody said [a submission of theirs] was a load of rubbish it wouldn't do their career much good".
This view was echoed by Cameron Neylon, a senior scientist at the Science and Technology Facilities Council's Rutherford Appleton Laboratory and a member of the journal's advisory group.
Dr Neylon even wondered whether fear of bad reviews would discourage researchers from submitting non-standard outputs.
He said the journal was an important experiment and the "next logical step" in the wake of PLoS ONE's demonstration of researchers' appetite for rapid, simplified peer review.
He said post-publication peer review would overcome the "binary" judgements about an output's value imposed by traditional peer review, the onerous nature of which also discouraged researchers from publishing all their results.
But Dr Neylon was unsure how long it would be before the new approach gained widespread acceptance.
Paul Groth, an assistant professor in the Knowledge Representation and Reasoning Group at VU University Amsterdam, said the biggest challenge for F1000 Research would be to attract good-quality reviews, which "will be key in establishing the full scientific quality of the posted work". But, as an advocate of "alternative metrics" in science publishing, he said the journal was a "good starting point".