undefined `centromere_bed` in Wakhan

[iagooteroc]

Description of the bug

Hi! I'm trying to run the pipeline but I am getting an error in Wakhan and it seems to be due to the params.centromere_bed not being set somehow:

Apr-10 10:13:48.952 [Actor Thread 41] WARN: Access to undefined parameter `centromere_bed` -- Initialise it to a default value eg. `params.centromere_bed = some_value`
[...]
 [2026-04-10 10:16:05] INFO: Cmd: /usr/local/bin/wakhan --target-bam PD1007b1_PD1007c_tumor.bam --breakpoints severus_all.vcf.gz --reference Homo_sapiens_assembly38.fasta --genome-name PD1007b1_PD1007c --out-dir-plots . --normal-phased-vcf PD1007b1_PD1007c.vcf.gz --contigs chr1-22,chrX,chrY --pdf-enable --centromere null --threads 6
[...]
 [2026-04-10 10:16:05] ERROR: Input file does not exist: /usr/local/lib/python3.10/site-packages/src/null

I am not sure why this could be happening. Could it be that conf/modules.config is trying to set "--centromere ${params.centromere_bed}" before it is actually set in workflows/lrsomatic.nf here params.centromere_bed = getGenomeAttribute('centromere_bed') ? Doesn't make a lot of sense.
Thank you.

Command used and terminal output

I am launching the pipeline like this:

nextflow run IntGenomicsLab/lrsomatic \
   -r 1.0.0 \
   -c ${CONFIG} \
   --genome GRCh38 \
   --input ${INPUT} \
   --outdir ${OUTDIR} \
   -resume \
   -profile singularity


Relevant output:

Apr-10 10:13:48.952 [Actor Thread 41] WARN: Access to undefined parameter `centromere_bed` -- Initialise it to a default value eg. `params.centromere_bed = some_value`
[...]
 [2026-04-10 10:16:05] INFO: Cmd: /usr/local/bin/wakhan --target-bam PD1007b1_PD1007c_tumor.bam --breakpoints severus_all.vcf.gz --reference Homo_sapiens_assembly38.fasta --genome-name PD1007b1_PD1007c --out-dir-plots . --normal-phased-vcf PD1007b1_PD1007c.vcf.gz --contigs chr1-22,chrX,chrY --pdf-enable --centromere null --threads 6
[...]
 [2026-04-10 10:16:05] ERROR: Input file does not exist: /usr/local/lib/python3.10/site-packages/src/null

Relevant files

These are the contents of my -c ${CONFIG} configuration file:

aws {
    client {
        anonymous = true
    }
}

executor {
    name = 'slurm'
    queueSize = 10
}

singularity {
    enabled = true
    autoMounts = true
    runOptions = '-B $SINGULARITY_TMPDIR:/tmp'
    //Used to allow Singularity to access bashrc variables
    envWhitelist = ['SINGULARITY_TMPDIR']
}

process {
    executor = 'slurm'
    clusterOptions = '-N 1 -n 1'
}

And this is my ${INPUT}:

sample,bam_tumor,bam_normal,platform,sex,fiber,clair3_model,clairSTO_model,clairS_model
PD1007b1_PD1007c,[...]/PD1007b1.sorted.bam,[...]/PD1007c.sorted.bam,ont,male,n,r1041_e82_400bps_hac_v430,,ont_r10_dorado_hac_5khz

I added those models to clair3_modelMap and clairs_modelMap in workflows/lrsomatic.nf like this:

def clair3_modelMap = [
    'dna_r10.4.1_e8.2_400bps_hac@v4.3.0': 'r1041_e82_400bps_hac_v430',
]
def clairs_modelMap = [
    'dna_r10.4.1_e8.2_400bps_hac@v4.3.0': 'ont_r10_dorado_hac_5khz',
]

.nextflow.log

System information

• Nextflow version 25.10.3
• Hardware: HPC
• Executor: slurm
• Container engine: Singularity
• OS: Rocky Linux 8.4
• Version of IntGenomicsLab/lrsomatic: 1.0.0

[ljwharbers]

Hi,

Great to hear you're looking into using our pipeline!

We are aware that in the 1.0.0 version there are definitely some bugs in Wakhan. We have solved most of those in the dev version, so I would definitely recommend trying to run this version, which we are releasing as 1.1.0 soon (run with specifying -r dev).

Most of us honestly typically skip wakhan --skip_wakhan, and use the ASCAT output if we are running on version 1.0.0 because there are a bunch of bugs in that version of Wakhan.

When it comes to the error, it is very strange since specifying --genome GRCh38 should automatically take the correct centromere bed file. You can try downloading the bed file from the wakhan github (https://raw.githubusercontent.com/KolmogorovLab/Wakhan/refs/heads/main/src/annotations/grch38.cen_coord.curated.bed) and specifying it in your nextflow call as follows:

nextflow run IntGenomicsLab/lrsomatic \
   -r 1.0.0 \
   -c ${CONFIG} \
   --genome GRCh38 \
   --input ${INPUT} \
   --outdir ${OUTDIR} \
   --centromere_bed /path/to/downloaded/centromere.bed \
   -resume \
   -profile singularity

But once again, would highly recommend trying our dev version and letting us know if you run into any issues there.

[iagooteroc]

Hi Luuk,

Thank you so much for your quick response. I was able to finish running the pipeline with --skip_wakhan, but it is helpful to know I can provide the centromere file with --centromere_bed.
I will also keep an eye out for the new version. Thanks!