nf-core/configs: Apocrita Configuration
[!WARNING] This is an unofficial project and not officially supported by QMUL ITS Research. Please do not approach the ITSR team for support with this config.
[!WARNING] Make sure that you are fully aware of the options this config submits to Apocrita, particularly the
highmemflag, to avoid abusing nodes on the Apocrita cluster.
All nf-core pipelines have been successfully configured for use on QMUL’s Apocrita cluster Queen Mary University of London.
To use, run the pipeline with -profile qmul_apocrita. This will download and launch the qmul_apocrita.config which has been pre-configured with a setup suitable for Apocrita.
Using Nextflow on Apocrita
Before running the pipeline you will need to configure Apptainer and install+configure Nextflow.
Singularity
Set the correct configuration of the cache directories, where <YOUR_ID> is replaced with you credentials which you can find by entering whoami into the terminal once you are logged into Apocrita. Run these lines in all job scripts or interactive sessions (or create a Private Module to set them):
# Set all the Apptainer environment variables
export APPTAINER_CACHEDIR=/data/scratch/<YOUR_ID>/.apptainer/
export APPTAINER_TMPDIR=/data/scratch/<YOUR_ID>/.apptainer/tmp
export APPTAINER_LOCALCACHEDIR=/data/scratch/<YOUR_ID>/.apptainer/localcache
export APPTAINER_PULLFOLDER=/data/scratch/<YOUR_ID>/.apptainer/pull[!WARNING] You need to make sure that all these folders actually exist before running for the first time.
Nextflow
Module
https://docs.hpc.qmul.ac.uk/apps/bio/nextflow/
Conda
Install Nextflow in a Conda Environment from the Bioconda channel.
Manual installation
[!WARNING] Do not pipe
curltobashas per the offical Quick start guide, follow the instructions below.
Download the latest release of nextflow. Warning: the self-update line should update to the latest version, but sometimes not, so please check which is the latest release (https://github.com/nextflow-io/nextflow/releases), you can then manually set this by entering (NXF_VER=XX.XX.X).
## Load OpenJDK module
module load openjdk
## Download Nextflow install script
curl -s https://get.nextflow.io -o nextflow
## Inspect the contents, then make it executable
chmod +x nextflow
## Check it runs
./nextflow
## Check it is up to date
./nextflow -self-update
NXF_VER=XX.XX.XYou can then move the nextflow executable to wherever you wish, for example:
mv nextflow ~/bin/nextflowThen make sure that this location is in your $PATH, by creating a Private Module, which should load the OpenJDK module as well as add the location of the executable to $PATH.
Config file
params {
config_profile_description = '(Unofficial) Queen Mary University of London'
config_profile_contact = 'Simon Murray (simon . murray AT ucl . ac . uk)'
config_profile_url = 'https://docs.hpc.qmul.ac.uk/'
}
executor {
name = 'sge'
}
apptainer.runOptions = "-B ${HOME},${PWD}"
process {
//NEED TO SET PARALLEL ENVIRONMENT TO SMP SO MULTIPLE CPUS CAN BE SUBMITTED
penv = 'smp'
//ADD MEMORY TO CLUSTEROPTIONS
clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.mega / task.cpus)}M -l rocky" }
withLabel: process_high {
clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.mega / task.cpus)}M -l highmem -l rocky" }
}
withLabel: process_long {
clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.mega / task.cpus)}M -l highmem -l rocky" }
}
withLabel: process_high_memory {
clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.mega / task.cpus)}M -l highmem -l rocky" }
}
}
singularity {
enabled = true
autoMounts = true
cacheDir = "/data/scratch/$USER/.apptainer/pull"
}