Unifying package managers, workflow engines, and containers: Computational reproducibility with BioNix

Justin Bedő; Leon Di Stefano; Anthony T Papenfuss

doi:10.1093/gigascience/giaa121

Unifying package managers, workflow engines, and containers: Computational reproducibility with BioNix

Gigascience. 2020 Nov 18;9(11):giaa121. doi: 10.1093/gigascience/giaa121.

Authors

Justin Bedő^{1

2}, Leon Di Stefano^{1

3}, Anthony T Papenfuss^{1

4

5

6

7}

Affiliations

¹ Bioinformatics Division, Walter and Eliza Hall Institute of Medical Research, 1G Royal Pde., Parkville, VIC 3052, Australia.
² School of Computing and Information Systems, University of Melbourne, Melbourne, VIC 3010, Australia.
³ Department of Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, 615 N. Wolfe Street, Baltimore, Maryland, U.S.A.
⁴ Peter MacCallum Cancer Centre, 305 Grattan St., Melbourne, VIC 3000, Australia.
⁵ Department of Medical Biology, University of Melbourne, Melbourne, VIC 3010, Australia.
⁶ Sir Peter MacCallum Department of Oncology, University of Melbourne, Melbourne, VIC 3010, Australia.
⁷ School of Mathematics and Statistics, University of Melbourne, Melbourne, VIC 3010, Australia.

Abstract

Motivation: A challenge for computational biologists is to make our analyses reproducible-i.e. to rerun, combine, and share, with the assurance that equivalent runs will generate identical results. Current best practice aims at this using a combination of package managers, workflow engines, and containers.

Results: We present BioNix, a lightweight library built on the Nix deployment system. BioNix manages software dependencies, computational environments, and workflow stages together using a single abstraction: pure functions. This lets users specify workflows in a clean, uniform way, with strong reproducibility guarantees.

Availability and implementation: BioNix is implemented in the Nix expression language and is released on GitHub under the 3-clause BSD license: https://github.com/PapenfussLab/bionix (biotools:BioNix) (BioNix, RRID:SCR_017662).

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Computational Biology*
Language
Reproducibility of Results
Software*
Workflow