Cluster Options

Run pipeline on compute cluster

Variant calling can be run in parallel on each sample using the -cluster argument. Set pbs resources such as resources (-l), email (-M), queue (-q), flux_account (-A) and notification (-m) under the [scheduler] section in config file.

For more details, refer to the UMICH flux website for a detailed explanation of each pbs specification.

Possible options for the -cluster option (Supported system: pbs):

cluster: The pipeline is optimized for this option. When this option is set, the pipeline will run variant call jobs for each sample on individual small compute clusters or on a single large cluster for core snp generation step.

local: This is the default option. When this option is set, the pipeline will analyze each sample one after the another. This will not make use of multiple cores present in your system or clusters. Use this option for small numbers of samples or for testing purposes.

parallel-local: This option will run the pipeline and analyze samples in parallel but on a local system. This is preferred for an input sample size of less than 20 and a multiple core local system.