Transparent, reproducible data.

T he scientific research paper remains the main mode of sharing scientific information. Two key aspects contribute to the unabated success of this format, despite the myriad possibilities for new ways to share research in an online world: Independent validation by peer review and the juxtaposition of research results—presented in figures, tables, and videos—with the authors’ textual interpretation of these results. For many years, publishers have focused largely on rendering text for publication and search engines are limited to indexing the title, abstract, introduction, results, methods and discussion sections of a paper. Yet, the claims made in a paper rely on data selected by the authors and synthesized into figures that illustrate the findings for the human reader. Figures are the heart of the paper and should open a vista on the research data. Unfortunately, that view is currently often limited to data visualizations and illustrations. As we have reported previously (Lemberger, 2010; Pulverer, 2014a,b), we encourage the presentation of Source Data, and indeed links to Source Data are already available in many figures in the journal. Source Data provides access to minimally processed versions of the data underlying figures such as data files used to plot graphs and experimental replicates or less cropped versions of representative data shown in a figure panel. Source Data can also include higher resolution renderings of data compressed in figures for efficient downloading. Source Data provides a foundation to a paper that affords a much richer view of the findings discussed and enables reanalysis and reuse by the interested reader, as well as computational access. (The data are published under a CC-0 license.) It also provides an efficient means to archive data in close juxtaposition with the paper that discusses the findings. A central tenet of publishing is to ensure reproducibility. This requires the transparent and detailed reporting and sharing not only of the data, but also the underlying methods and reagents. Our goal is to develop method sections that satisfy these conditions, including detailed experimental protocols and reagents that can be identified unambiguously, and which link directly to the relevant figure panels and data. We are conscious that these enhancements will need to tie into the development of digital laboratory management tools that allow submission of this information without unreasonably burdening the authors.