2026-10-27 –, Naturalis 02
In 2023 eLife launched a ‘Publish, Review, Curate’ (PRC) model of publishing which emphasised the publication of preprints, preprint review, and ‘post-publication’ curation. In this session you will discover how PIDs can create a public record connecting the revisions of a paper with public peer reviews, metadata (e.g. funding), and related research outputs (data and code). You’ll explore the practical considerations when registering metadata for ‘new’ types of research output that don’t neatly map to existing content-based schemas in scholarly infrastructure, and hear how we navigated the tension between best practice and author requirements.
After this session, you will understand:
- How to architect a PID strategy that transparently links preprints, article versions, peer reviews, data and software.
- The challenges that accompany mapping new or alternative models of publishing to the existing scholarly infrastructure.
Crossref’s schema is content-based, requiring metadata registrants to declare what type of content they are registering (e.g. journal article, posted-content, book chapter, grant, etc.). This poses a challenge when releasing new research output types. In eLife’s case, in 2023 we introduced the concept of a ‘Reviewed preprint’ (although the concept is not entirely new, cf. ‘refereed preprints’), it doesn’t neatly map to Crossref’s existing content types. Is it a preprint, a journal article, something else (should Crossref update their schema to accommodate this new item)? There has been some recent debate on this particular issue in the context of MetaROR [1], [2].
In addition we had competing requirements: a) it is best practice to mint a DOI for each revision, and doing so enables us to effectively link peer reviews which pertain to a particular version. b) we know our authors prefer the simplicity of having one DOI per ‘article’ (across all versions), for citation counts, and to limit the proliferation of records downstream. Having a stable DOI allows authors to always point to the latest version of the article in grant applications, which is particularly important in a publishing model that emphasises article-level (even version-level) research assessment over metrics such as Impact Factor. We opted to do both (a unique DOI for each version and then an ‘umbrella’ DOI which always links to whichever version is the latest, making extensive use of Crossref's 'relations' to disambiguate and link all the versions).
The session will detail the path eLife took, and more importantly why we took it, as well as outlining the positives (transparency, etc.) and negatives (the impact downstream of having multiple DOIs for the same article) of the approach. If there’s space to, I may also compare it to the approach taken by MetaROR, which could prove fruitful for discussion and further thought.
[1] Waltman L., and Brasil A. (2025) Innovation in scientific publishing and its implications for Crossref DOI registration practices - Request for input, Crossref Blog. https://www.crossref.org/blog/innovation-in-scientific-publishing-and-its-implications-for-crossref-doi-registration-practices-request-for-input/
[2] Waltman L., and Brasil A. (2026) Innovation in scientific publishing and its implications for Crossref DOI registration practices - MetaROR’s approach, Crossref Blog. https://www.crossref.org/blog/innovation-in-scientific-publishing-and-its-implications-for-crossref-doi-registration-practices-metarors-approach/
