somaticseq failing for same command it had previously successfully run

[csreej27]

I am getting the above error message for the below mentioned command , which has been in the past run and results successfully generated.

python3.7 /home/ec2-user/.local/bin/makeSomaticScripts.py paired --normal-bam /path/to/normal.bam --tumor-bam /path/to/tumor.bam --genome-reference /home/ec2-user/efs/sridhar/Somatic_resources/hg19/hg19.fa --inclusion-region /home/ec2-user/efs/sridhar/Somatic_resources/hg19/hg19_inclusion.bed --output-directory somaticseq_results_retrial_check --dbsnp-vcf /home/ec2-user/efs/sridhar/sreejita/hg19_dbsnp_sorted/new_sorted_dbsnp_138.hg19.vcf --threads 12 --container-tech docker --run-mutect2 --run-vardict --run-lofreq --run-strelka2 --run-somaticseq--run-somaticsniper --run-varscan2 --run-workflow

Although I don't understand why it is saying somaticseq distribution not found because when I just type the following python3.7 /home/ec2-user/.local/bin/makeSomaticScripts.py paired --help The help page pops up perfectly well :

[litaifang]

What if you just try to install somaticseq and invoked makeSomaticScripts.py directly (without invoking the full path like python3.7 /home/ec2-user/.local/bin/makeSomaticScripts.py)?

[csreej27]

Hi, As per your suggestion I invoked makeSomaticScripts.py directly and now I am getting the following error :

I replicated the error across various server instances and also different input files

[litaifang]

2 Questions:

• Does the hg19.fa have associated hg19.fa.fai and hg19.dict in the same directory?
• Is bedtools installed and callable in the instance?

[litaifang]

You also had --run-somaticseq--run-somaticsniper without space in them. Don't know if that was your actual command and if that made any difference.

[csreej27]

• the hg19.fa does have the associated hg19.fa.fai and hg19.dict in the same directory
• yes, bedtools is installed and can be invoked directly from anywhere in the instance
• "--run-somaticseq--run-somaticsniper without space" - I believe this might just been a typing error while posting the issue here because I checked my scripts and there was space in there

[litaifang]

That's weird.... can you copy-and-paste the actual script?

[csreej27]

Below was my command :

makeSomaticScripts.py paired --normal-bam /path/to/normal.bam --tumor-bam /path/to/tumor.bam --genome-reference /path/to/hg19/hg19.fa --inclusion-region /path/to/hg19/hg19_inclusion.bed --output-directory test_out_dir --dbsnp-vcf /path/to/hg19_dbsnp_sorted/new_sorted_dbsnp_138.hg19.vcf --threads 12 --container-tech docker --run-mutect2 --run-vardict --run-lofreq --run-strelka2 --run-somaticseq --run-somaticsniper --run-varscan2 --run-workflow Do you want me to send you my actual bash script ?

[litaifang]

Try creating the output directory somaticseq_results_retrial_check. That may be the problem. makeSomaticScripts.py will not create the output_directory if it does not exist already.

[litaifang]

Think in the next version, I'll have makeSomaticScripts.py create the output_directory if it does not exist. Let me know if that's the issue.

[csreej27]

Hi, So somaticseq ran and gave results successfully after two measures you suggested :

1. invoking script directly ( without mentioning full path )
2. creating the output directory prior to running the script

So thank you for suggesting those measures, I really appreciate it .

However I would like to point out that we have faced several concerns over the past year with somaticseq - around its stability . It would be great if you could please look into it

[litaifang]

In terms of stability, are those docker issues or somaticseq issues?