Error early if required parameters aren't set

[imdanique]

Description of the bug

Hi,
Thanks for the great pipeline!

I'm using samurai v.1.2.3.1 and encountered this issue:

I wanted to skip the alignment part and run "liquid_biopsy" subworkflow starting from bam files with the following command:

nextflow run samurai-1.2.3.1 -profile singularity --input samplesheet_bam.csv --outdir samurai --binsize 500 --analysis_type liquid_biopsy --caller ichorcna --aligner false

However, I noticed that it doesn't start after seeing this:

[-        ] DINCALCILAB_SAMURAI:SAMURAI:BAM_QC_PICARD:PICARD_COLLECTMULTIPLEMETRICS -
[-        ] DINCALCILAB_SAMURAI:SAMURAI:BAM_QC_PICARD:PICARD_COLLECTHSMETRICS       -
[-        ] DINCALCILAB_SAMURAI:SAMURAI:BAM_QC_PICARD:PICARD_COLLECTWGSMETRICS      -
WARN: No PoN specified: CNA calling performance can be impacted
ERROR ~ Argument of `file` function cannot be null

I tried experimenting with options but no matter what it doesn't start unless I remove --analysis_type liquid_biopsy and run solid_biopsy version.

The test command for liquid_biopsy works just fine on my machine:

nextflow run samurai-1.2.3.1 -profile test_ichorcna,singularity --outdir test -c slurm.config

Is there anything I am missing or is that a bug? I would appreciate your help

[imdanique]

Thanks to the comments in issue #33, I was able to run the liquid biopsy workflow with the following parameters:

nextflow run samurai-1.2.3.1 \
        -profile singularity \
        -c slurm.config \
        --input samplesheet_bam.csv \
        --outdir results \
        --genome hg38 \
        --fasta ref_genomes/hg38_no_alt/grch38_core+bs_controls.fa \
        --fai ref_genomes/hg38_no_alt/grch38_core+bs_controls.fa.fai \
        --dict ref_genomes/hg38_no_alt/grch38_core+bs_controls.fa.dict \
        --analysis_type liquid_biopsy \
        --caller ichorcna \
        --binsize 500 \
        --normal_panel samurai-1.2.3.1/assets/ichorcna/HD_ULP_PoN_hg38_500kb_median_normAutosome_median.rds \
        --ichorcna_gc_wig samurai-1.2.3.1/assets/ichorcna/gc_hg38_500kb.wig \
        --ichorcna_map_wig samurai-1.2.3.1/assets/ichorcna/map_hg38_500kb.wig \
        --ichorcna_centromere_file samurai-1.2.3.1/assets/ichorcna/GRCh38.GCA_000001405.2_centromere_acen.txt \
        --compute_signatures TRUE \
        --run_gistic TRUE \
        --gistic_broad_analysis TRUE \

It would be very helpful if the documentation included a minimal set of required parameters for each type of workflow.

That said, my run did not complete fully and stopped at the GISTIC2 step.

I’ve opened a separate issue for this here: #47.

[lbeltrame]

The problem lies in the fact that when we point the pipeline at some files, they must be something, while an empty value (because you didn't specify it) is something that causes an error.

A solution I have in mind is to fail early if you use a pipeline that requires them, and you haven't set them, is to error early and tell the user about it. Not sure when I will get around to it though.