On 25 March, the OMF’s Director, Michael Barton, hosted a rich discussion amongst colleagues from the Public Library of Science (PLOS) and CERN IT, as well as an engaged audience.
The event began with a presentation from Lauren Cadwallader, the Open Research Manager at PLOS. Lauren presented the methodology and findings of a pilot study for peer reviewing biology and physiology models when publishing research. During the pilot study, 71 researchers opted into having their computational models peer reviewed. The most common reason researchers had for opting in was their interest in promoting reproducible practices. One key conclusion Lauren presents is that top-down policies may be adopted if the community is already amenable to such practices. In this case, a scientific community keen on open science practices seemed welcoming of a model peer review process.
The second presentation was from Lars Holm Nielsen, Head of Open Science Repositories at CERN IT. Lars delivered an analysis of Zenodo, the online platform for the sharing, curation and publication of data and software. At a basic level, Zenodo functions as a generalist platform for the sharing of an enormous swathe of data and software components, along with their versions. Lars explained how version control – as enabled by Zenodo – is reflected in academic literature, where we see that researchers tend to cite the latest version of any such component. At a broader level, Zenodo is interoperable with a wider range of research tools and practices. For example, Lars mentioned how Zenodo can be integrated with citation graphs, and allows data to feed into relevant repositories, such as the Global Biodiversity Information Facility.
The third presentation involved observations from PLOS Computational Biology Editors-in-Chief, Jason Papin and Feilim Mac Gabhann. Jason argued for model sharing practices from three perspectives: an instructor supporting their students, a researcher responding to technical clarifications, and an editor requesting more detail from an author. Jason built on this to explain that model sharing also shortens the review cycle; where code is available, peer reviewers can more quickly give more targeted feedback.
Following the three presentations, Michael Barton moderated a rich Q&A session with the audience. Watch the full workshop below, join us on 09 April for the following workshop, and sign up to our Google Group to learn how you can get involved with the ModelShare program!
📸 Image by Alan Warburton / © BBC / Better Images of AI / Plant / Licenced by CC-BY 4.0