generally not ideal and should be avoided. the dualrnaseq hasn't been updated in 4 years, maybe look at more recently maintained pipelines for patterns
party this 1
g
Guntas Khullar
12/15/2025, 5:14 PM
I understand that the any files that are manipulated should mainly go through the input and output but is there a specific reason/s why that is not encouraged? Maybe some documentation that explains why?
p
Phil Ewels
12/15/2025, 7:15 PM
Because it duplicates disk space usage in the work directories, and there's almost always no need for it.
Phil Ewels
12/15/2025, 7:15 PM
Nextflow is able to pull out specific files and dynamically name them / locate them however you like in the downstream task. So there's very rarely a need to mv or cp them.
Phil Ewels
12/15/2025, 7:16 PM
Note that on blob storage like s3, there is no way to move a file. So doing a mv operation effectively just copies it, with the speed + expense penalties that come with that.