-
Notifications
You must be signed in to change notification settings - Fork 15
Troubleshooting.md
If you get errors like this:
ImportError: libbz2.so.1.0: cannot open shared object file: No such file or directory
Try switching from the Conda provided samtools to the one installed by pip:
cd tools
source ./miniconda/bin/activate STR
pip install --upgrade --force-reinstall pysam
rm bin/samtools
cd bin
ln -s ../bwa.kit/samtoolsCheck that that the reference genome used to generate your bam matches up with the bed file you pass to STRetch. For example are chromosomes labelled "chr1" or just "1"?
Make sure you have allocated sufficient memory. The STRetch_wgs_bam_pipeline.groovy requires 25-32 Gb memory to process a typical 40x PCR-free whole human genome. This number could be higher depending on specific data features like the number of unmapped reads and the number of read pairs that map disparately.
- Update to the latest version of STRetch, delete the 'tools' directory, and then run the install script.
- Delete all intermediate files before re-running the pipeline. Or better yet run it in a fresh directory.
- Try running one of the commands directly from the command line to see if it fails. i.e. the last line from
commandlog.txtis the most recent command it tried to run.
chromosome: STR-ATCATG not found in 54.bam
chromosome: STR-ATCC not found in 54.bam
chromosome: STR-ATCCC not found in 54.bam
This indicates that no reads were found aligned to those chromosomes. This isn't necessarily an issue, for example there may be no ATCATG expansions in your sample.
***** WARNING: File 11.bam has inconsistent naming convention for record:
chr1 70349896 70350046 1-1391/1 0 +
***** WARNING: File 11.bam has inconsistent naming convention for record:
chr1 70349896 70350046 1-1391/1 0 +
This is a common bedtools warning that I can't seem to suppress, but isn't cause for concern.