3. Your environment

In the SGE environment, a scheduler system will queue your jobs and execute them as soon as there are free nodes, based on the resources (threads, memory, etc.) that you specified. The commands for all the processes that you need to run should be written into a shell script file and submitted to the scheduler. You should open a file with a text editor (for example vi, nano or pico), and write at the beginning of the file a header similar to the following:

 #$ -S /bin/sh
 #$ -cwd
 #$ -q amd.q,large.q,intel.q
 #$ -l h_vmem=20G
 #$ -e GenomeAnnotaiotn_error.e
 #$ -N GenomeAnnotation_experiment
 #$ -o GenomeAnnotation_out.o

Then, in the body of the script you should first load the appropriate modules. In the SGE environment, several bioinformatics modules are ready be loaded. After that, you could write Unix commands, calls to installed software, and so on. Alternatively, you can run commands and software from the command line. Unless you log in into bert using an interactive shell, the bert front end has stringent limits to the process that you run from command line. In this tutorial some short processes can be run from command line, while other requiring more resources will be run through the scheduler.