CALCULATE_FRIP fails when no peaks are found

[peneder]

Check Documentation

I have checked the following places for your error:

• [ x] nf-core website: troubleshooting
• [x ] nf-core/cutandrun pipeline documentation

Description of the bug

First, thank you for this great pipeline!
I noticed the following bug: For samples for which no peaks are called (i.e. the .peaks.bed.stringent.bed file is empty), the following error occurs during CALCULATE_FRIP : "RuntimeError: .peaks.bed.stringent.bed does not seem to be a recognized file type!"

My impression is that the problem lies here, in bin/frip.py. It seems like it would make sense to check if the .bed file is empty before calling crpb.CountReadsPerBin, and return 0 as the frip value in this case.

Note that I have replaced the real sample/group name with in the error log below due to confidentiality.

Steps to reproduce

Steps to reproduce the behaviour:

1. Command line: nextflow run nf-core/cutandrun -r 1.0.0 -bg --input $(pwd)/config/CUTandRUN/2021-12-20.samplesheet.csv --outdir $(pwd)/res/CUTandRUN/nextflow_CUTandRUN_pipeline/2021-12-20/nextflow -profile docker --genome hg38 --spikein_genome K12-MG1655 --max_memory "120.GB" --max_cpus 20 -resume 13659c93-98fb-41b6-a4b6-28729f6b59cd
2. See error:
   Error executing process > 'NFCORE_CUTANDRUN:CUTANDRUN:CALCULATE_FRIP (H2BK15ac_Dox_s03_R1)'

Caused by:
Process NFCORE_CUTANDRUN:CUTANDRUN:CALCULATE_FRIP (<group>) terminated with an error exit status (1)

Command executed:

frip.py
--bams ".bam"
--peaks ".bed"
--threads 12
--outpath .

python --version | grep -E -o "([0-9]{1,}.)+[0-9]{1,}" > python.version.txt

Command exit status:
1

Command output:
Calculating .target.markdup.bam using .peaks.bed.stringent.bed

Command error:
WARNING: Your kernel does not support swap limit capabilities or the cgroup is not mounted. Memory limited without swap.
Matplotlib created a temporary config/cache directory at /tmp/matplotlib-94b1cu4t because the default path (/.config/matplotlib) is not a writable directory; it is highly recommended to set the MPLCONFIGDIR environment variable to a writable directory, in particular to speed up the import of Matplotlib and to better support multiprocessing.
Traceback (most recent call last):
File "/home/peter/.nextflow/assets/nf-core/cutandrun/bin/frip.py", line 44, in
reads_at_peaks = cr.run()
File "/opt/conda/envs/packages/lib/python3.8/site-packages/deeptools/countReadsPerBin.py", line 356, in run
imap_res = mapReduce.mapReduce([],
File "/opt/conda/envs/packages/lib/python3.8/site-packages/deeptools/mapReduce.py", line 85, in mapReduce
bed_interval_tree = GTF(bedFile, defaultGroup=defaultGroup, transcriptID=transcriptID, exonID=exonID, transcript_id_designator=transcript_id_designator, keepExons=keepExons)
File "/opt/conda/envs/packages/lib/python3.8/site-packages/deeptoolsintervals/parse.py", line 591, in init
ftype = self.inferType(fp, line, labelColumn)
File "/opt/conda/envs/packages/lib/python3.8/site-packages/deeptoolsintervals/parse.py", line 166, in inferType
raise RuntimeError('{0} does not seem to be a recognized file type!'.format(self.filename))
RuntimeError: .peaks.bed.stringent.bed does not seem to be a recognized file type!]

Expected behaviour

CALCULATE_FRIP on empty .bed files should return 0 instead of an error.

Log files

Have you provided the following extra information/files:

• [x ] The command used to run the pipeline
• The .nextflow.log file (sample names are confidential)

System

• Hardware: HPC
• Executor: local
• OS: Debian
• Version 4.9.88-1+deb9u1 (2018-05-07) x86_64

Nextflow Installation

• Version: 21.10.0.5640

Container engine

• Engine: Docker
• version: Docker version 18.09.0, build 4d60db4

[chris-cheshire]

Hi @peneder, this bug is fixed in the upcoming v1.1 version which should be out in a week or two. If you need to run the pipeline before then i suggest you run the dev branch using -r dev

[chris-cheshire]

Fixed in - https://github.com/nf-core/cutandrun/releases/tag/1.1