Job submission in HPC cluster - DR-genomics/Genomics-pipelines GitHub Wiki
Note: No jobs should be run directly on the login node. All jobs should run using the HPC's cluster queuing system (for eg., standby, comm_mmem_day, comm_mmem_week, etc.,) or research group specific queues.
Thornyflat cluster - bioinfo software/tools are packed within a singularity container "/shared/containers/Bioinformatics_2021.sif". To access tools (such as blast, muscle, repeatmasker, etc.,) within this container, perform any analysis via this container. See below for reference.
Login to your account in the cluster using: ssh -X [email protected] followed by ssh -X [email protected]
To perform a simple blastn analysis, open a new text file using any text editor in home directory (~/):
vi blastn.pbs
Press "i" to get into insert mode. Copy and paste the below script in the text file.
#!/bin/sh
#PBS -N JOB_Blast #### job name
#PBS -l nodes=1:ppn=4 #### specify no. of nodes and processors required
#PBS -l walltime=20:00:00 #### maximum limit for walltime
#PBS -m ae #### -m Email options; a => email sent when job is aborted; e => email sent when job ends
#PBS -q cfb0001 #### Queue name
module load singularity/3.7.0 #### Load singularity container to access all bioinformatic software installed within it.
module load lang/python/cpython_3.8.6_gcc93 #### Load python
cd /users/dramacha
##### To use BLAST installed within the singularity container "Bioinformatics_2021.sif", execute the container (singularity exec <container name>) first before running blast command.
singularity exec /shared/containers/Bioinformatics_2021.sif blastn -query SbiCDS.fasta -db SbiCDS.fasta -evalue 1e-10 -outfmt 6 -perc_identity 80.0 -out SbiCDS.aba.blast
Now, press "ESC" and type :wq (write and quit). From the command prompt, submit the blastn job by typing:
qsub blastn.pbs
A job ID will be assigned after job submission. To check the status of the above job:
qstat -u <username>