ClairS crashing with spaces in input file name

[RenzoTale88]

Hello,
I come across an issue when the software is provided input files with spaces in their path.

$ ~/Software/ClairS/run_clairs --tumor_bam_fn Test\ data/demo_tumor.remap.bam --normal_bam_fn Test\ data/demo_tumor.remap.bam --bed_fn Test\ data/demo.bed --ref_fn Test\ data/GCA_000001405.15_GRCh38_no_alt_analysis_set_chr20.fna --threads 4 -o Test\ data/TESTING_BREAK/ClairS_test -p ont_r10_guppy

[INFO] Create folder Test data/TESTING_BREAK/ClairS_test

[COMMAND] /home/talentia/Software/ClairS/run_clairs --tumor_bam_fn Test data/demo_tumor.remap.bam --normal_bam_fn Test data/demo_tumor.remap.bam --ref_fn Test data/GCA_000001405.15_GRCh38_no_alt_analysis_set_chr20.fna --threads 4 --platform ont_r10_guppy --output_dir Test data/TESTING_BREAK/ClairS_test --bed_fn Test data/demo.bed

[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/logs
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/split_beds
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/candidates
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/pileup_tensor_can
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/fa_tensor_can
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/vcf_output
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/tmp_vcf_output
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/logs/clair3_log
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/clair3_output/phased_output
[INFO] Create folder /home/UID name/Test/Test data/TESTING_BREAK/ClairS_test/tmp/clair3_output/vcf
gzip: /home/UID.gz: No such file or directory
gzip: name/Test/Test.gz: No such file or directory
gzip: data/demo.bed.gz: No such file or directory
[INFO] --include_all_ctgs not enabled, use chr{1..22,X,Y} and {1..22,X,Y} by default
[WARNING] No contig in --bed_fn was found in the reference, contigs in BED /home/UID name/Test/Test data/demo.bed:

This seems to be related to how the workflow decompresses the input files, particularly the input BED and genotyping VCF here and here.
This is particularly relevant when the space is within the user name, making it impossible to run the software.

Thanks
Andrea

[zhengzhenxian]

@RenzoTale88 Thanks for raising this. The space in the input path leads to parser error. We have tried to add path quotation but find it hard to cover all the parameters and especially for the softwares we are using, such as Clair3, WhatsHap, and LongPhase.

We would suggest you remove the space in all your input parameters, including the input path of BAM, REF, BED, and genotyped VCF.

[RenzoTale88]

@zhengzhenxian thanks for the quick reply. Unfortunately, the issue is not always this easy to sort out: it is fairly common for users to have spaces in their user names, which in turn makes their home the folder having space in it. This would make the issue impossible to circumnvent for these users. I appreciate the complexity of fixing this though!