Experimenting with reproducibility in bioinformatics

Reproducibility has been shown to be limited in many scientific fields. This question is a fundamental tenet of the scientific activity, but the related issues of reusability of scientific data are poorly documented. Here, we present a case study of our attempt to reproduce a promising bioinformatics method [1] and illustrate the challenges to use a published method for which code and data were available. First, we tried to re-run the analysis with the code and data provided by the authors. Second, we reimplemented the method in Python to avoid dependency on a MATLAB licence and ease the execution of the code on HPCC (High-Performance Computing Cluster). Third, we assessed reusability of our reimplementation and the quality of our documentation. Then, we experimented with our own software and tested how easy it would be to start from our implementation to reproduce the results, hence attempting to estimate the robustness of the reproducibility. Finally, in a second part, we propose solutions from this case study and other observations to improve reproducibility and research efficiency at the individual and collective level.