There are many platforms on which you can deposit your preprint (let’s call it a “preprint” even if you don’t intend to publish it in a journal). These platforms are of three types: preprint and e-print servers, institutional repositories and open archives.
You should carefully consider the characteristics of these platforms before depositing the preprint you want to submit to a PCI. In particular, you should ask yourself the following questions:
- Does my institution recommend a particular archive for preprints?
Research institutes or universities may recommend that affiliated researchers deposit preprints in their institutional repository, or may have a specific directory in an open archive.
- Is the scientific scope of my article appropriate for the platform?
arXiv was long considered to be restricted to mathematics, physics and computer sciences. However, this is no longer the case, as it now accepts any article including quantitative analyses in biology, finance, statistics, electrical engineering, systems science, and economics. Note that almost all biology papers include quantitative analyses.
- Is my preprint of a type compatible with the platform?
Most preprint servers, repositories and archives accept any sort of preprint, but some impose restrictions. For instance, bioRxiv is restricted to preprints presenting “Results” and does not host reviews, opinions, or perspectives.
- Will the peer-reviewed status of my article be clearly displayed on the platform?
To our knowledge, no preprint server, repository or archive correctly displays the peer-review status of their content yet. bioRxiv shows annotations/reviews from Review Commons in the upper right corner of the corresponding preprint home page. bioRxiv also displays a link to PCI reviews and recommendations in the “Preprint discussion sites covering this article” section, found just below the article’s abstract and copyright statement. However, for these preprints, as for non-evaluated preprints, bioRxiv displays a sentence stating that “This article is a preprint and has not been certified by peer review“. The same type of warning is displayed in Preprints.org and ResearchSquare. This sentence is only removed for preprints that are subsequently published in journals. This sentence will not be removed even if your preprint is recommended by PCI, giving readers the incorrect impression that your preprint has not been peer-reviewed. Reviews and endorsements of bioRxiv preprints are also listed on the Hive platform, developed by eLife.
- Do I want to post a final version of my preprint if it is published in a journal?
The “author accepted manuscript” or “postprint”, i.e. the unformatted version of the accepted manuscript that still belongs to the authors (i.e. the version before copyright transfer), may be used for green Open Access (“self-archiving”).
It may not be possible to deposit postprints on some servers, repositories and archives. For instance, bioRxiv does not allow authors to upload postprints (https://www.biorxiv.org/about/FAQ), whereas HAL and arXiv do. You can read more on the subject at this post.
Most preprint servers, repositories and archives accept all articles for which the authors hold copyright, and, in cases of submission to a journal, if the journal’s copyright transfer agreement does not conflict with the servers’ licenses. The Sherpa Romeo website is a useful resource for checking whether and when the journal to which you submitted your preprint allows you to archive the postprint.
- Is the platform well-indexed and are its articles easy to find?
This is an extremely important question. Some platforms are well-indexed (e.g., arXiv, HAL, bioRxiv, OSF-preprints and its branded preprint servers, etc.) by major search engines (Google, Google Scholar), whereas others are currently less visible (e.g., Zenodo). In general, institutional repositories and open archives are not referenced in scientific databases (e.g. PubMed, Europe PMC, Dimensions), but preprint servers may be (e.g. arXiv is indexed in Dimensions but not in Europe PMC, and only bioRxiv preprints with NIH fundings are index in PubMed).
These technical features are continually evolving, and it is, therefore, a good idea to do a quick search for updates before posting your article.
- Do I want to bypass the classic publication system?
Some servers seek to facilitate the transfer of preprints to journals by considering themselves as a preliminary step in the publication process (e.g., Research Square). Others, such as arXiv or OSF-preprints, have taken a more independent position, seeking to disseminate research results as quickly as possible, without considering the fate of preprints.
By answering these questions (along with others relating to metadata richness, services provided, full html availability, etc.), you should be able to make the best choice of platform for your preprint. This choice is an important one, because switching servers/repositories after an initial deposition is discouraged, to avoid the creation of multiple DOIs for the same article. Moreover, there is usually an “indexing” advantage to the first server used for a particular article on well-known search engines. It is, therefore, worth taking the time to make the right choice for your work.
- Additional criteria and modifications suggested by colleagues after the first publication of the post:
Commitment to preservation of works. For example, at the very least, authors are advised not to choose a location where the terms state that works can be deleted/removed without notice.
There was an inaccuracy in the figure (corrected 22 Sept 2020): @arxiv does not limit articles to those not published in journals, authors may post e-prints of published articles if they retained the copyright or have the license to do so from the publisher.