BCC2020

→ Abstract


The presenter(s) will be available for live Q&A in this session (BCC West).

Jaqueline J. Brito 1,*, Jun Li 2, Jason H. Moore 3, Casey S. Greene 4,5, Nicole A. Nogoy 6, Lana X.
Garmire 2, Serghei Mangul 1,7

1 Dept. of Clinical Pharmacy, School of Pharmacy, University of Southern California, USA
2 Dept. of Computational Medicine & Bioinformatics, University of Michigan, USA
3 Dept. of Biostatistics, Epidemiology, and Informatics, Institute for Biomedical Informatics,
University of Pennsylvania, USA
4 Dept. of Systems Pharmacology and Translational Therapeutics, University of Pennsylvania, USA
5 Childhood Cancer Data Lab, Alex's Lemonade Stand, USA
6 GigaScience, Hong Kong
7 Quantitative and Computational Biology, University of Southern California, USA
*Email: britoj@usc.edu

Project Website: https://github.com/Mangul-Lab-USC/enhancing_reproducibility
License: CC BY 4.0 License

Computational methods reshaped the landscape of modern biology, generating new channels of
communications to publish and share the most recent techniques and methodologies. While the
dependence on computational tools of the biomedical community increases steadily, the
mechanisms ensuring open data, open software, and reproducibility are heterogeneously
enforced. Institutions, funders, and publishers offer different guidelines, or no guideline at all.
For instance, publications may cite software artifacts, key to reproduce research results, that
may become unavailable or depend on packages no-longer supported. Publications lacking fully
reproducible research significantly limit the role of reviewers in evaluating technical strength
and scientific contribution. Moreover, incomplete ancillary information for an academic
software package will likely bias and restrict any subsequent research produced with the tool.
In this presentation, we provide eight recommendations across four different domains to
improve three main principles: reproducibility, transparency, and rigor in computational
biology. These are the main principles which should be emphasized in life sciences curricula,
especially as assays and pipelines grow more complex than ever. We propose that a
combination of lowering the learning curve needed to maintain the three principles and more
strict guidelines are key to ensure adoption by the community. Ultimately, our
recommendations target fostering a sustainable data science ecosystem in biomedicine and life
science research.
Keywords: Reproducibility; Open science; Reproducible research; FAIR principles.