Home

Awesome

A Simple SGE Template Preprocessor Engine

qsubsec is a template language for generating script files for submission using the SGE grid system. By using this system, you can separate the logic of your qsub jobs from the data required for a specific run.

Overview

The qsubsec utility processes qsubsec-formatted template files to generate code sections that can be submitted for execution (for example through the SGE scheduler). Before processing, template files are checked for token placeholders and these are filled in. The resulting Python code is then executed to generate a set of sections.

The stages in template processing are:

Processed sections can be output in multiple formats (currently bash and qsub), allowing the same templates to be run on multiple different systems.

For more information on the TFF token definition syntax and the qsubsec template syntax, see the documentation directory.

Installation

Installation should be as simple as:

git clone https://github.com/alastair-droop/qsubsec.git
cd qsubsec
python setup.py install

You will need python3 (>=3.3) to use qsubsec.

If you do not have admin privileges, you can install this locally using python setup.py install --user.

After installation, you can verify that you have the correct version using qsubsec -v.

Although qsubsec can be run on most machines, the qsub executable must be available and functional for automatic qsub submission to work (using the -s argument with the qsub format).

Licence

These tools are released under the GNU General Public License version 3.

Reference

If you use qsubsec for any published work, please reference the original paper as:

Alastair P. Droop; "qsubsec: a lightweight template system for defining sun grid engine workflows", Bioinformatics, Volume 32, Issue 8, 15 April 2016, Pages 1267–1268, https://doi.org/10.1093/bioinformatics/btv698

BibTex:

@article{doi:10.1093/bioinformatics/btv698,
author = {Droop, Alastair P.},
title = {qsubsec: a lightweight template system for defining sun grid engine workflows},
journal = {Bioinformatics},
volume = {32},
number = {8},
pages = {1267-1268},
year = {2016},
doi = {10.1093/bioinformatics/btv698},
URL = { + http://dx.doi.org/10.1093/bioinformatics/btv698},
eprint = {/oup/backfile/content_public/journal/bioinformatics/32/8/10.1093_bioinformatics_btv698/3/btv698.pdf}
}