Reproducible bioinformatics project: a community for reproducible bioinformatics analysis pipelines

Area of Science:

• Bioinformatics
• Computational Biology
• Scientific Computing

Background:

• Research reproducibility is crucial for modern science and industrial applications.
• Ensuring reproducibility requires accessible data and clearly described computational workflows.
• Complex bioinformatics analyses face reproducibility challenges due to tool and library versioning.

Purpose of the Study:

• Establish a non-profit, open-source project for reproducible bioinformatics.
• Provide a schema and infrastructure using Docker images and R packages.
• Facilitate the creation and distribution of robust, reproducible bioinformatics workflows.

Main Methods:

• Developed a schema and infrastructure based on Docker images and R packages.
• Utilized ad hoc Ubuntu Docker images to simplify workflow integration.
• Implemented workflows using R functions within a standardized R package structure.
• Required R vignettes and example datasets for workflow documentation and usability.

Main Results:

• Created the Reproducible Bioinformatics Project (RBP).
• Enabled the integration of workflow modules into Docker images.
• Standardized workflow implementation in R for homogeneity and reusability.
• Provided a system for distributing reproducible bioinformatics analyses.

Conclusions:

• The RBP offers a general schema and infrastructure for distributing reproducible workflows.
• Guarantees consistent analysis replication for end-users.
• Ensures reproducibility across different UNIX-like architectures.