Home // LCSB // Research // Bioinformati... // Reproducible Research Results (R3)

Reproducible Research Results (R3)

The goal of the R3 project is to provide infrastructure and methods for data processing and storage in order to improve research quality and increase overall reproducibility of scientific results.

With increasing complexity of generated data and associated analyses, there is a growing concern in the scientific community about the reproducibility of findings in many peer-reviewed scientific articles [1]–[5]. Currently, a description of experimental procedures and end results are required in the publishing process, with much less focus on sharing raw structured data sets. Moreover, with the rapid pace of technology, many workflows or computational environments get outdated very quickly, thus are difficult to re-establish. This leads to a global problem where many research findings cannot be reproduced based on the contents of a published article.

At the LCSB, we are implementing a “reproducible research results” layer to standardize main steps of a classical research publication workflow, such as structuring and standardizing data, capturing lab protocols and experimental methods in electronic lab notebooks, source code versioning, workflow management and freezing of project states via virtualization technologies such as virtual machines or linux containers.

  1.  M. Baker, “Irreproducible biology research costs put at $28 billion per year,”Nature, Jun. 2015.
  2. “Code share.,”Nature, vol. 514, no. 7524, p. 536, Oct. 2014.
  3. C. G. Begley and L. M. Ellis, “Drug development: Raise standards for preclinical cancer research.,”Nature, vol. 483, no. 7391, pp. 531–3, Mar. 2012.
  4. J. Rung and A. Brazma, “Reuse of public genome-wide gene expression data.,”Nat. Rev. Genet., vol. 14, no. 2, pp. 89–99, Feb. 2013.
  5. Open Science Collaboration, “Estimating the reproducibility of psychological science,”Science (80-. )., vol. 349, no. 6251, pp. aac4716–aac4716, Aug. 2015.