Using Containers in Supercomputing Environment

Running nf-core pipeline at CSC

This tutorial needs accesss to Puhti, which requires that:

In this tutorial, you will learn to:

💬 nf-core is a community effort to collect a curated set of analysis pipelines built using Nextflow. nf-core facilitates standardisation of Nextflow pipelines with best practices, guidelines, and templates for bioinformatics community. These workflows are modular, scalable, and portable. Here we use single-cell RNA-seq pipeline as an example nf-core pipeline.

Login to Puhti supercomputer

In this tutorial, we will use Puhti supercomputer. First login to Puhti using SSH (or by opening a login node shell in the Puhti web interface):

ssh <username>@puhti.csc.fi    # replace <username> with your CSC username, e.g. myname@puhti.csc.fi

And navigate to a scratch directory to submit Nextflow job:

mkdir -p /scratch/<project>/$USER/                               # replace <project> with your CSC project, e.g. project_2001234
cd /scratch/<project>/$USER/  && mkdir -p nf-core && cd nf-core

Wrap nf-core pipeline as a normal batch job

Here is an example batch script to run nf-core pipeline on Puhti:

#!/bin/bash
#SBATCH --time=01:00:00
#SBATCH --partition=small
#SBATCH --account=<project_2xxxx>
#SBATCH --cpus-per-task=4
#SBATCH --mem-per-cpu=4000

# Let's use the fast local drive for temporary storage
export SINGULARITY_TMPDIR=$PWD
export SINGULARITY_CACHEDIR=$PWD

# This is just to avoid some annoying warnings
unset XDG_RUNTIME_DIR


# Load Nextflow on Puhti
module load nextflow/23.04.3

# nf-core pipeline example here
nextflow run nf-core/scrnaseq  -profile test,singularity -resume --outdir .

‼️ Note! If you are directly pulling multiple images on the fly, please set $APPTAINER_TMPDIR and $APPTAINER_CACHEDIR to either local scratch (i.e. $LOCAL_SCRATCH) or to your scratch folder (/scratch/<project>) in the batch script. Otherwise $HOME directory, the size of which is only 10 GB, will be used. To avoid any disk quota errors while pulling images, set $APPTAINER_TMPDIR and $APPTAINER_CACHEDIR in your batch script.

Please make sure to specify the correct version of the Nextflow module as some nf-core pipelines require a specific version of Nextflow.

copy and paste the above script to a file named scrna_nfcore.sh and replace your project number with <project_2xxxx> in slurm directives.

Finally, submit your job as below:

sbatch scrna_nfcore.sh

💬 When you are running a nf-core pipeline directly using Nextflow as above, all the GitHub scripts of a given pipeline are download to a directory: $HOME/.nextflow/assets/nf-core/. You can navigate to the directory and explore Nextflow script and configurations settings. Alternatively, you can clone a GitHub repository of a nf-core pipeline and run the pipeline from the same directory.

Monitor the status of submitted Slurm job

   squeue -j <slurmjobid>
   # or
   squeue --me
   # or
   squeue -u $USER

(Bonus) Think of the following aspects of Single-cell RNA-seq pipeline

  1. Where are the actual codes (local path) of Nextflow pipeline located? (tip: check slurm output file for more details or use nextflow info nf-core/scrnaseq)
  2. What kind of Nextflow profiles are used in the above batch script? Also, find out how those profiles are described in the configuration file?
  3. Find out the different mondatory command line options associated with scrnaseq pipeline?( hint: use the following command: nextflow run nf-core/scrnaseq –help)
This site is open source. Improve this page.