HaplotypeCallerSpark

CWL Questions
1

I am getting error while running CWL tool for HaplotypecallerSpark. Error is as follows.

‘’’
ERROR [job HaplotypeCallerSparkNew.cwl] Job error:
(“Error collecting output for parameter ‘rawVCF’:\nHaplotypeCallerSparkNew.cwl:37:5: Did not find output file with glob pattern: ‘[‘en.vcf’]’”, {})

‘’’

2

Hello @Ambarish_Kumar. Can you link or paste in the source for HaplotypeCallerSparkNew.cwl and how you are running it? It won’t be possible for anyone to assist you without that. Thanks!

3

This is the web-link for the tool.

I am running it over command-line.

‘’’
sudo cwltool HaplotypeCallerSparkNew.cwl --inputBAM ebolamutantmarkdup.bam --Reference2bitGenome ebola1.2bit --RefIndex ebola.fasta.fai --sampleName eb --RefDict ebola.dict --BAMindex ebola.bam.bai --output_filename en.vcf

‘’’
Input files are contained into following sub-directory.

‘’’


‘’’
4

I am trying with different outputbinding parameters but error persists.

5

I don’t think this command is running as you expect it:

arguments:
- valueFrom: |-
    gatk HaplotypeCallerSpark -R $(inputs.Reference2bitGenome.path) -I $(inputs.inputBAM.path) -O $(inputs.output_filename)
  shellQuote: false

Will actually run as “gatk HaplotypeCallerSpark -R …” rather than “gatk” “HaplotypeCallerSpark” “-R” …

I would consider using the ‘inputBinding’ for each of your inputs and a ‘baseCommand’ set to ['gatk', 'HaplotypeCallerSpark']

For the reference, you have put it in as a listing (can you justify why this is necessary), but have specified the .path attribute. If you wish to reference the file that you have listing in the working directory, you should use the .basename attribute instead.

Alexis