eLife’s Computationally Reproducible Articles, Code Ocean, and the Future of the “Print” Journal

On February 20, 2019, eLife formally “published” a working prototype for a computationally reproducible article. Unlike the traditionally published research articles that we are accustomed to, this one enriches the research article with code, data, and interactive figures that can be executed in the reader’s browser, downloaded, and explored. 

The importance of providing dynamic publishing platforms has been growing as research has become increasingly data and software intensive in the sciences.

This is not a new requirement. Back in the 1970s as genetic research was starting to take off, the way the details of newly identified genetic sequences were shared was to print pages and pages of nucleic acid names in print journals.  Clearly this was an inefficient way to communicate the results of that research.  First, the proofreading challenge was enormous. But more importantly, there was no practical way to reuse the discovered data.  Then, in 1982, GenBank was launched and genetic sequences could be shared and compared from a database at NCBI rather than a printed journal.  Journal publishers acknowledged the importance of this new database by requiring sequences to be deposited in GenBank before an article would be accepted for publication. 

Journal publishers continue to rely on external databases and code bases like GitHub (now owned by Microsoft) and Code Ocean to accommodate the data and code components of the research results they publish. And researchers currently use tools like Jupyter Notebooks to create interactive reports.

But these resources are external platforms not associated with the publisher.  And they do not constitute formal publications. In contrast, what eLife offers is a platform for formally publishing the text, data, and actionable code all wrapped up in a formal publication with its own DOI.  This new eLife format can also offer benefits for evaluating the research and facilitating reproducible results.

eLife has been publishing peer reviewed, open access articles in biomedicine and life sciences since 2012. eLife’s launch was supported by Howard Hughes Medical Institute, Max Planck Society, and the Wellcome Trust.  In September 2016 it adopted an article processing charge business model. Its articles are covered in the key abstracting and indexing databases like Medline.  

eLife has been busy.  In a webinar on March 11, 2019, the team at eLife gave an update on Libero Publisher – one component of the open-source publishing technology suite eLife is building. Libero provides end-to-end publishing workflow that could expedite the transition to open access and reduce production costs for independent publishers.  The work on Libero is now being done in collaboration with non-profit Collaborative Knowledge Foundation (Coko), open-access publisher Hindawi Limited, and digital consultancy Digirati.

So, what does all this mean for the future of the traditional “print” journal? Whereas subscriptions to paper journals have been shrinking for years, the PDF version of journals continues to have a solid presence.  We see it as the primary option when we open our weekly alert for Science Magazine.  Publishers like Nature call for new articles to be submitted in PDF or Word format to facilitate the review process.  Readers find page-styled content easy to read.  PDFs are not going away, but they are better suited to research synopses than to full research articles. The growing question, particularly in the sciences, is whether static pages are sufficient to the task of conveying research results.  This is not a new question but it is becoming a more important one.

Also, we should not overlook the fact that creating print page layouts for journals adds to the costs and the production timeline for journals. The push from Plan S for full Open Access has created pressures on the publishing revenues it creates, especially for society publishers.

There are major challenges for publishers going forward.  Although significant, Plan S is not the only challenge publishers face.  Fundamental to the future of successful communication of research results is the need to adequately convey those results in comprehensive ways. In the areas of scientific research, that means providing not just the data but also the code that was used to create the data. eLife is taking a significant step in that direction.  And by packaging that information in a formal publication, it preserves the ability for researchers to cite the article and get recognition for their work.  This new publication format is definitely worth following closely.