Test profile fails at purple module

[mariesieger]

Description of the bug

I ran the test profile to test the oncoanalyser pipeline, but it fails at the purple module. I already tried to solve the issue by expanding time limits in the .config file, but it still interrupts after 1h

nextflow_v2.zip

Command used and terminal output

$ nextflow run nf-core/oncoanalyser   -revision 2.3.0   -profile test,docker   -c nextflow_v2.config   --outdir /data/trr305/Multiomics/TransVar_Project/analysis/oncoanalyser_test

Command output:
  10:25:24.542 [INFO ] Purple version 4.3
  10:25:24.549 [INFO ] reference(subject_a.normal) tumor(subject_a.tumor)
  10:25:24.549 [INFO ] output directory: purple/
  10:25:24.678 [INFO ] using ref genome: V38
  10:25:24.906 [DEBUG] loaded 1 alternative transcripts from 1 genes
  10:25:24.959 [DEBUG] loaded 39500 gene records
  10:25:26.322 [DEBUG] loaded 39500 genes with 39501 transcripts records and 247411 exons
  10:25:26.631 [INFO ] loaded 53284 germline deletions frequencies from file: cohort_germline_del_freq.38.csv
  10:25:26.699 [INFO ] reading Amber QC from amber/subject_a.tumor.amber.qc
  10:25:26.701 [INFO ] reading Amber BAFs from amber/subject_a.tumor.amber.baf.tsv.gz
  10:25:26.707 [INFO ] reading Amber PCFs from amber/subject_a.tumor.amber.baf.pcf
  10:25:26.742 [INFO ] Amber average tumor depth(30) ambiguous BAF(0.581)
  10:25:26.743 [INFO ] reading Cobalt tumor segments from cobalt/subject_a.tumor.cobalt.ratio.pcf
  10:25:26.745 [INFO ] reading Cobalt ratios from cobalt/subject_a.tumor.cobalt.ratio.tsv.gz
  10:25:28.764 [INFO ] reading Cobalt reference segments from cobalt/subject_a.normal.cobalt.ratio.pcf
  10:25:28.894 [INFO ] loaded 4 somatic SVs from subject_a.tumor.esvee.somatic.vcf.gz
  10:25:28.917 [INFO ] loaded 103 somatic variants from subject_a.tumor.pave.somatic.vcf.gz
  10:25:28.917 [INFO ] sample gender is female
  10:25:28.917 [INFO ] output directory: purple/
  10:25:28.941 [INFO ] loaded 0 germline SVs from subject_a.tumor.esvee.germline.vcf.gz
  10:25:28.942 [INFO ] applying segmentation
  10:25:28.942 [INFO ] merging reference and tumor ratio break points
  10:31:25.430 [DEBUG] chimerism analysis: weighted-mean(NaN)
  10:31:25.431 [INFO ] fitting purity
  10:31:25.436 [DEBUG] tumor evidence: svHotspotCount(1) svFragmentReadCount(47)
  10:31:25.466 [DEBUG] building copy numbers
  10:31:25.495 [INFO ] maxDiploidProportion(1.000) diploidCandidates(1) purityRange(1.000 - 1.000) hasTumor(true)
  10:31:25.566 [INFO ] enriching somatic variants
  10:31:25.571 [DEBUG] modelling somatic peaks
  10:31:25.572 [DEBUG] somatic peak uses 99 variants
  10:31:25.691 [INFO ] load(3.0 tml=3.0000) msiIndels(0 perMb=0.0000) burden(103.0 perMb=0.0360)
  10:31:25.691 [DEBUG] charting variants: total(snvs=103 indels=0) downsampled(103 snvMod=1 indelMod=1)
  10:31:25.907 [DEBUG] generating QC stats
  10:31:25.934 [INFO ] load 1 germline variants from subject_a.tumor.pave.germline.vcf.gz
  10:31:26.606 [INFO ] generating drivers
  10:31:26.638 [INFO ] generating charts
  10:31:26.658 [INFO ] generating subject_a.tumor.circos.png via command: /usr/local/bin/circos -nosvg -conf purple/circos/subject_a.tumor.circos.conf -outputdir purple/plot -outputfile subject_a.tumor.circos.png
  10:31:26.658 [INFO ] generating subject_a.tumor.input.png via command: /usr/local/bin/circos -nosvg -conf purple/circos/subject_a.tumor.input.conf -outputdir purple/plot -outputfile subject_a.tumor.input.png
  10:31:26.682 [DEBUG] executing R script via command: Rscript /tmp/script3208635633421707300.R subject_a.tumor purple/ purple/plot/
  10:31:26.683 [DEBUG] executing R script via command: Rscript /tmp/script10048119809232375301.R subject_a.tumor purple/ purple/plot/

Command error:
  /usr/local/bin/purple: line 6: warning: setlocale: LC_ALL: cannot change locale (en_US.UTF-8): No such file or directory

Work dir:
  /data/trr305/Multiomics/TransVar_Project/analysis/oncoanalyser_test/work/c7/8aec260e9012da8432a50b9163b8be

Container:
  quay.io/biocontainers/hmftools-purple:4.3--hdfd78af_0

Tip: view the complete command output by changing to the process work dir and entering the command `cat .command.out`

 -- Check '.nextflow.log' file for details
ERROR ~ Pipeline failed. Please refer to troubleshooting docs: https://nf-co.re/docs/usage/troubleshooting

 -- Check '.nextflow.log' file for details

Relevant files

.nextflow.log

System information

Nextflow version: 25.04.7
Hardware: server
Executor: local
Container engine: Docker 28.4.0
OS: Ubuntu 22.04.5 LTS
Version of nf-core/oncoanalyser: v2.3.0-g234fd82

[scwatts]

Hi @mariesieger, I can see PURPLE is running more or less as expected until chart generation, where it then hangs for ~50 minutes then failing due to hitting the time limit. Under normal circumstances PURPLE runs on the test data within 5-10 minutes

My initial suspicion is that there is some server configuration impacting use of /tmp/, which PURPLE uses when creating the charts

Are you able to provide any further information on your compute environment? It would also be useful if you're able to navigate into the PURPLE work directory and manually try executing the command, then monitoring in htop or similar