One thing I like in Nextflow is how “publishDir” can be made dynamic, avoiding duplicated processes.
In my “FASTQC” workflow, I wanted the same process to handle both raw and trimmed reads while publishing outputs to different folders.
Instead of creating separate logic, I used a conditional publishDir
publishDir(
path: { task.process.contains(‘TRIMMED_FASTQC’)
? “${params.outdir}/shortReads/checkQuality/trimmed/”
: “${params.outdir}/shortReads/checkQuality/raw/”
}, mode: ‘copy’
)
The trick is that I reused the same module with include and assigned it a new process name.
include { FASTQC as TRIMMED_FASTQC } from ‘./modules/fastqc’
Since “task.process” reflects the aliased process name, the same “FASTQC” module automatically publishes results to the correct location..
A small Nextflow trick that keeps pipelines easier to maintain. 