Preprint mandates gather momentum

By Richard Sever, PhD. Chief Science & Strategy Officer, openRxiv

Preprints speed up science by enabling researchers to disseminate reports of new findings immediately. They also represent a simple way to provide free access to those findings, which has long been the goal of the Open Access movement. Organizations that fund biomedical research are increasingly recognizing this, and philanthropic funders like the Gates Foundation, the Chan Zuckerberg Initiative, the Simons Foundation, and Aligning Science Across Parkinson’s now require grantees to post manuscripts as preprints as a condition of funding. In doing so, they are implementing the ‘Plan U’ proposal, which argued that preprint mandates are the easiest route to universal free access and could stimulate much-needed evolution of scientific publishing and academic assessment.

In a significant move for a major biomedical research funder, the Howard Hughes Medical Institute (HHMI) has now also adopted Plan U. HHMI investigators, scholars and scientists employed at HHMI’s Janelia research campus must now “​​make original research articles that represent major contributions from their laboratories publicly available as…preprints” as part of a new access policy centered on preprints rather than journal articles (HHMI scientists can still publish their work in traditional journals as well, but there are some restrictions on payment of article-processing charges).  

HHMI notes that “shifting the focus to preprints…encourages a more transparent scholarly ecosystem in which researchers have more control over dissemination of their scholarship”, a key goal of openRxiv. Importantly, bioRxiv and medRxiv, the two preprints servers operated by openRxiv, are among the servers HHMI recommends, along with arXiv and chemRxiv. The requirement that HHMI scientists use non-profit, community servers is welcome given the dominance of academic publishing by commercial for-profit companies, as is the recognition that this could promote a more healthy scholarly ecosystem. 

An important aspect of the new HHMI policy is the requirement for scientists to post both an initial preprint and a revised preprint drafted following peer review. The new policy thus recognizes the importance of peer review, but as a process in which authors receive and respond to feedback from other experts, rather than as a certification decision by a journal editor based on this. This is consistent with HHMI’s support for eLife, a journal that recently abandoned accept/reject decisions following peer review, as well as previous proposals from HHMI leadership that peer review should be a service not a selection process.  

The shift in focus may prompt arguments that there are important roles journals play besides peer review – for example, performing various integrity/data checks on manuscripts. These are important, but the importance of these checks varies between fields, and some journals do not perform them at all. This kind of content verification may be better achieved by dedicated third-party services in a decoupled and more transparent way. Indeed, bioRxiv and medRxiv have done just this by partnering with groups like Sciscore, Dryad, and Dataseer to provide independent data and reagent/resource verification. openRxiv will collaborate with these organizations and funders like HHMI to expand the spectrum of content checks available and so ensure readers (both human and machine) have more trust signals they can use to assess information. The new HHMI policy is thus not just a smart move that ensures immediate, free access to the research it funds but also a push towards a better scholarly communication ecosystem. Other funders should follow their lead.