Sydney-Informatics-Hub / Fastq-to-BAM

Optimised pipeline to process whole genome sequence data from fastq to BAM on NCI Gadi
GNU General Public License v3.0
20 stars 7 forks source link

samblaster Path Hard-Coded to Inaccessible Path #17

Closed DarioS closed 2 years ago

DarioS commented 2 years ago

I noticed all error logs had samblaster errors, such as:

$ cat Logs/Dedup_sort_error_capture/OSCC_45-CC_M.err
Corrupted or missing BAM
Error in samblaster log

I think the reason is in dedup_sort.sh.

# Does not like being in run_parallel script
samblaster=/scratch/er01/apps/samblaster/0.1.24/samblaster

The path is not accessible to end-users.

$ ls /scratch/er01/apps/samblaster/0.1.24/samblaster
ls: cannot access '/scratch/er01/apps/samblaster/0.1.24/samblaster': No such file or directory

This crucial edit which must be done by end-user should be stated in README.md.

calizilla commented 2 years ago

I removed the offending relic line from the script. The samblaster module (global app) is loaded within the driver script dedup_sort_run_parallel.pbs, so there is no need for the user to update path to samblaster within the dedup_sort.sh task script.