Easiest way to design Bioinformatics pipeline

Question

[Nai]

Hello Everyone!

I am newbie for programming to make bioinformatics pipeline. I would like to know basic steps or starting steps for bash scripting in nextflow. I would like to say , how to start pipeline from QC to DEGs analysis

[Pierre Lindenbaum]

I would like to say , how to start pipeline from QC to DEGs analysis

it can be quite complicated https://github.com/hoelzer-lab/rnaflow/blob/master/main.nf

You should start small

[Mensur Dlakic]

I think the easiest way is to copy a similar pipeline. I don't mean this flippantly. In most cases pipelines have already been designed that are either identical or very similar.

Answer 1

[ATpoint]

To be honest, I would first of all write a regular bash script that contains the individual pipeline steps. Get a small test dataset that completes within seconds to see whether you're getting the commands right. Then you can dive into existing tutorials to wrap that into Nextflow. Don't get me wrong, Nextflow and other workflow managers can be valuable to orchestrate, parallelize and containerize your workflows, but it is an entire new language/ecosystem with its own logic and pitfalls that (at least for me) took quite some time to get my head around, even now that I use it quite a bit. Writing end-to-end pipelines with good reporting and some documentation and tests takes time, just saying. Nextflow (and others) are not easy or simple (as often advertised), but powerful once you get the how-to's.

Answer 2

[Matthias Zepper]

I recommend following along the tutorial Reproducible, scalable, and shareable analysis workflows with Nextflow by Sateesh Peri, Michael Cipriano and Matthew Hunter Seabolt.