Is it normal for DEXSEQ_EXON to take a long time? I see my pipeline crashed because it reached time limit (16h) my params and run command:

nextflow run nf-core/rnasplice -r 1.0.3 -resume -c custom.config -profile docker -params-file nf-params.json

# params:
{
  "input": "samplesheet.csv",
  "contrasts": "contrasts.csv",
  "source": "genome_bam",
  "outdir": "outs",
  "fasta": "Homo_sapiens.GRCh38.dna_sm.primary_assembly.fa.gz",
  "gtf": "ensb110_modified.gtf.gz",
  "star_index": "star",
  "salmon_index": "salmon\/",
  "skip_alignment": true,
  "edger_exon": true,
  "dexseq_exon": false,
  "dexseq_dtu": false,
  "min_samps_gene_expr": 3,
  "sashimi_plot": false,
  "suppa": false,
  "rmats": false
}

Yes, I had the same problem. You can speed it up a bit if you give it a lot of CPU cores (e.g. 36), but even then it's going to take at least 24 h. I have a config file similar to this:

process {
        withName: 'DEXSEQ_EXON' {
                // if your scheduler needs any special options, put them in clusterOptions
                time = 48.h  // I hope that is safe
                cpus = 36    // make sure you have that many or adjust as needed
                maxRetries = 0  // if it still doesn't work, we need to adjust the config file before rerunning
        }
}

I hope that works for you as well. 🤞

Thanks sir, yes I have done something similar and this is useful

It can process more exons in parallel if you have more cores. I have not done any thorough testing to what extent the increase in speed is correlated to the number of cores, so I don't know if there is a point beyond which the benefit from an extra core becomes negligible (or where that would be). But in my limited experience the process got further with 36 cores than with 12 or 24.