Configuration file
Nextflow allow to separate the configuration for the pipeline execution from the workflow logic.
MetaPhage comes with some pre-configured profiles, but it can be convenient to simply use a configuration file for the setup and then add -c mycluster.conf to the pipeline execution.
What to configure
- Executor (local, slurm, PBS…)
- Dependencies (Singularity, Docker,…)
Executor
Using a local machine
When running the pipeline in a local machine, you can use the local executor, supplying the details about your available CPUs and memory. MetaPhage will require at least 16 CPUs and 64 Gb of RAM.
params {
max_cpus = 64
max_memory = 128.GB
max_time = 72.h
}
process {
executor='local'
}
Using slurm
A predefined template is available for slurm, but should be customized with the available queues in your HPC.
process {
executor='slurm'
withLabel:big_res {
cpus = 32
memory = 128.GB
time = 72.h
queue = "your_queue_name"
}
withLabel: med_res {
cpus = 16
memory = 32.GB
time = 48.h
queue = "your_queue_name"
}
withLabel: low_res {
cpus = 8
memory = 16.GB
time = 24.h
queue = "your_queue_name"
}
}
Dependencies
Miniconda
If you created a Miniconda environment with the requested dependencies, just actvitate it before running the workflow.
Singularity
If you want to use Singularity, you have to download the image first and then either add -with-singularity PATH_TO_IMAGE to the pipeline execution, or add
// Change the value to the path of your singulariity image
process {
container = '$projectDir/containers/metaphage.simg'
}
Docker
A Docker image is also available, as described here.