Document 2: A preprint service supported by an international consortium of funders

Drafted by ASAPbio

As a public good, preprints should obey a single data standard that will enable them to reside in a single database.  This database should be permanent, well-maintained, free for all to use, and easily accessible and searchable. This preprint service should have an outstanding governance structure that will represent the needs of the scientific community and oversee adaptations that will be inevitably needed in the future. Further issues regarding the implementation of this preprint service are considered in Document 3.

**As an eventual outcome of this meeting, we recommend the funding of a preprint service in biology by an international consortium of public and private agencies. The contract for this preprint service should support the initial development of infrastructure, yearly operating costs (with the possibility of metric-driven growth), a system of trustworthy governance, and a commitment that the server will not charge submission fees for at least its first five years.** 

The reasons for recommending a single preprint service supported by public/private funding are:

Best chance of adoption by the biology community.  Communication through preprints is foreign to biologists.  Developing a highly-trusted preprint service in the life sciences that is directly supported by major funding agencies and governed by outstanding scientists will promote buy-in to this form of communication.

High visibility.  Scientists want their work to be widely viewed, since this helps them to establish reputation and priority. In the case of arXiv, having a single platform where scientists go to look for new work in the field immensely aids visibility and helps physicist to establish priority of discovery. Maximum visibility also will ultimately require a search engine that integrates preprints with peer-reviewed publication, which should be considered in the initial development of the preprint service as well.

Ensuring trustworthy governance. Direct funding of a preprint service will enable funders to play a more direct role in the governance and future directions of preprints, rather than being relegated to a more peripheral role.  An international component of governance needs to be considered from the onset for this global resource.

Maintaining quality and standards.  A consortium-funded preprint service could help to define quality control for submission and standards for sharing and maintaining data.

Overall cost and ease of funding.  Direct support of a preprint service by a consortium of funders will be small, certainly in comparison to investments made by funders indirectly or directly towards journal publication and the open access of journal publications.  As an alternative to funding a preprint service, funders could provide financial support directly to the scientist to pay for preprint submissions (e.g. as a specific new line item on a grant).  This is similar to the present-day journal system, but in this model, the scientific community and funders have little say in cost and governance.  Furthermore for such a scientist-payer model to take effect globally, all funding agencies would need to develop a new policy allowing scientists to include preprint costs in their budgets.  Even if achieved, this plan is less democratic, as it favors scientists with larger grants versus scientists who have less funds but would like to benefit from preprint use.

Time. There is a sense of momentum and support for the growth of preprints in biology, but this momentum can be lost as quickly as it is being gained. Funders could play a major role in sustaining this momentum by becoming involved and supporting a preprint service. If funders take a lead role, together with the influential junior and senior scientists, then preprints have a chance of becoming common practice in biology.  If funders take a back seat over the next couple of years, if biologist adopt a wait-and-see attitude to see funders become interested, and if preprint submissions tick upward only at a slow rate, biologists will see preprints as a nice idea in theory but a failed experiment in practice.  It might prove difficult to recover from such a situation.