Late last week in Barcelona, Spain, the Centre for Genomic Regulation (CRG) hosted a workshop entitled Nextflow: Reproducible In silico Genomics. From introductory to state-of-the-art sessions, hackathons plus informal opportunities to discuss use cases and more in detail, this workshop attempted to address the interests of all participants. Judging from social media posts, it was an engaging and productive session for all involved.
For anyone who’s just so much as glanced at Nextflow, the very positive upshot of this recent workshop is hardly surprising, as this software addresses a challenge fundamental to the Life Sciences. Appearing soon after the very codification of DNA into sequences of amino acids, the challenge of ensuring the reproducibility of results in genomic-analysis pipelines emerged. Ironically with the ongoing uptake and success of computational methods in biology, and the Life Sciences in general, this reproducibly challenge became significantly exacerbated.
Not only have Singularity containers also experienced significant uptake in the Life Sciences in general, they also share with Nextflow a natural affinity for ensuring reproducibility and mobility. Moreover, from a technical perspective, Singularity containers permit an extremely natural, appropriate, and complementary means for effectively and efficiently containerizing those applications executed within a Nextflow-mediated workflow.
Recognizing the combined value of this solution for the Life Sciences, members of both communities recently collaborated in developing a series of lab notes. Owing to the success of this collaboration, and in order to communicate the collective output in a single document, the technical white paper Containerized Genomic Workflows with Singularity has recently been developed. The executive summary of the white paper is as follows:
The workflow manager Nextflow and the container platform Singularity combine to create the ideal solution for many of the problems prevalent in scientific fields such as machine learning and genomics.
With Nextflow and Singularity, problems such as mobility, security, and scalability are a thing of the past. Even the “reproducibility crisis” of the data-intensive fields now have an easy solution.