#21 Permission denied when using A5 for MiSeq on university supercomputer

[Sébastien]

Hi,

I am trying to test assembling a bacterial genome with A5. I have paired-end MiSeq 2x150 bp reads. I uploaded my reads in compressed fastq format (fastq.gz) and the script on my university's cluster computer and ran a job to assemble a genome. Unfortunately the job failed after a few minutes. Here is the extract from the error log:
[a5] '/home/smata/a5_miseq_linux_20150522/bin'/sga correct -t 4 -p 20160103_a5_out.pp -o 20160103_a5_out.s1/20151012_S9_L001_R1_001_combi.fastq.gz.a5unzipped.both.pp.ec.fastq 20160103_a5_out.s1/20151012_S9_L001_R1_001_combi.fastq.gz.a5unzipped.both.pp > 20160103_a5_out.s1/raw1.correct.out
sh: /home/smata/a5_miseq_linux_20150522/bin/sga: Permission denied
readline() on closed filehandle TDPIPE at /home/smata/a5_miseq_linux_20150522/bin/a5_pipeline.pl line 502.
Running cat 20160103_a5_out.s1/20151012_S9_L001_R1_001_combi.fastq.gz.a5unzipped.both.repair.fastq >> 20160103_a5_out.s1/20160103_a5_out.ec.fastq
Done merging libraries
[a5] ERROR: Unable to identify any properly paired reads
[a5] ERROR: Please check that you have provided at least one library of paired-end reads with names conforming to the Illumina read pair convention.

It seems that there is a permission problem with the scrit. Any idea where it comes from ? Is it just me misusing it ?

Any help would be most welcome !

[Aaron Darling]

This is likely to be a problem specific to your cluster. Check the CPU architecture -- is it 64-bit x86 or something else like ARM, itanium, or powerpc? A5-miseq for linux is only built for 64-bit x86. You might also check whether SELinux extensions may prevent certain types of executable files from running. Possibly ask the cluster admin.