ENCORE. A practical implementation to improve reproducibility and transparency of computational research.

Abstract

Reproducibility of research outcomes has become increasingly important within all research fields, including the (bio)medical domain. Despite ongoing (open-science) initiatives, computational research is often still difficult to reproduce. Although many papers are available that layout clear guidelines to improve computational reproducibility, the translation and application of these guidelines into practice seems to be an obstacle. We propose ENCORE (ENhancing COmputational REproducilbity) as a practical implementation of previously published guidelines to improve reproducibility and transparency through a tight integration of all project components (data, code, results, concepts, documentation). ENCORE comprises a standardized file system structure (sFSS), pre-defined (Markdown) files for documentation, a GitHub repository for software version control, a HTML-based sFSS browser (FSS Navigator), and ENCORE documentation, to set up a self-contained project package and support researchers to use ENCORE in a consistent manner. ENCORE can be used for any type of computational project. It improved reproducibility and transparency of our computational projects but, at the same time, made clear that achieving full reproducibility requires additional steps. Perhaps the most significant challenge towards improved reproducibility is the lack of clear and direct incentives for and intrinsic motivation of the scientists.