Copyright (c) 2017 Satoshi ITO, Masaaki YADOME, and Satoru MIYANO
Virtual Grid Engine (VGE) is a kind of middleware for running bioinformatics software pipelines on large-scale supercomputers which do not support any grid engine survices. VGE employs master-worker model. It first reserves processors and/or cores by running the job which is parallelized by MPI, then asign divided small tasks onto its worker processes. VGE is written in python.
VGE uses Message Passing Library through MPI4PY module. It also uses socket communication between VGE master process and VGE jobcontroler process. Dependent software and their versions are below:
- Python (2.7>=)
- MPI4PY (2.0.0>=)
- MPICH or OpenMPI (2.0>=)
Download source archive and extract it. Then type following commands:
$ cd VGE
$ python setup.py install --user
S. Ito et al., "Virtual Grid Engine: Accelerating thousands of omics sample analyses using large-scale supercomputers," 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Madrid, Spain, 2018, pp. 387-392. doi: 10.1109/BIBM.2018.8621285