Running a NextFlow pipeline from nf-core

[wong-nw]

Start logging steps and taking notes to set up an nf-core pipeline run

[wong-nw]

1. Find a pipeline of interest from https://nf-co.re
2. Navigate to the usage page to determine necessary input files
3. Log on to biowulf and set up an interactive session
4. Set up the singularity cache to use the user's data directory instead of home (https://ccbr.github.io/HowTos/docs/containers/singularity-biowulf.html#cache-directory)
5. module load singularity and module load nextflow
6. Test the workflow with built in tests
   Example: `nextflow run nf-core/airrflow -r 4.3.0 -profile test,singularity --outdir airrflow test
7. Prepare necessary inputs
   • samplesheets
   • path to input files (e.g. fastq.gz)
   • path to config file
8. Run according to recommended specs for project.

[kelly-sovacool]

need to add conf/biowulf.config & conf/slurm.config when running

[bianjh-cloud]

I tried running https://nf-co.re/rnaseq/3.25.0/, but am getting permissions issue. Re-running the workflow on tmp/ does not seem to work.

sinteractive --mem=400g --cpus-per-task=16

540 10:56 module load singularity
541 10:56 export SINGULARITY_CACHEDIR=/data/$USER/.singularity
542 10:56 module load nextflow
543 10:56 nextflow run nf-core/rnaseq -r 3.18.0 --input samplesheet.csv --outdir results --genome GRCh38 -profile singularity -w /data/$USER/nf-work -resume

---

Command exit status:
126

Command output:
(empty)

Command error:
WARNING: Not virtualizing pid namespace by configuration
/bin/bash: line 10: .command.run: Permission denied

[kelly-sovacool]

are you running the command in a directory where you have write access? e.g. /data/$USER/rnaseq-project/?

if not, I would recommend you do this:

mkdir /data/$USER/rnaseq-project/
cd /data/$USER/rnaseq-project/
module load nextflow
export SINGULARITY_CACHEDIR=/data/$USER/.singularity
nextflow run nf-core/rnaseq -r 3.18.0 --input samplesheet.csv --outdir results --genome GRCh38 -profile singularity -resume

[bianjh-cloud]

Thanks Kelly - I am still getting the same error

[kelly-sovacool]

Does this happen when you run a CCBR pipeline with the equivalent command? e.g. nextflow run CCBR/SINCLAIR ... (fill in the rest of the args)