Overview

The MSU HPCC, managed by ICER, provides an optimal platform for conducting bioinformatics workflows. This guide explains how to run the nf-core/cutandrun pipeline for CUT&RUN and CUT&Tag analysis, ensuring reproducibility and efficiency.

Key Benefits of nf-core/cutandrun

nf-core/cutandrun offers:

• Reproducible CUT&RUN/CUT&Tag Analysis: Comprehensive and standardized workflows.
• Portability: Runs seamlessly across various computing infrastructures.
• Scalability: Capable of handling small and large-scale datasets.

Prerequisites

• Access to MSU HPCC with a valid ICER account.
• Basic understanding of Singularity and Nextflow.

Step-by-Step Tutorial

Note on Directory Variables

On the MSU HPCC:

• $HOME refers to the user’s home directory (/mnt/home/username).
• $SCRATCH refers to the user’s scratch directory, ideal for temporary files and large data processing.

Note on Working Directory

The working directory, where intermediate and temporary files are stored, can be specified using the -w flag when running the pipeline. This helps keep outputs and temporary data organized.

1. Load Nextflow Module

Ensure Nextflow is loaded:

module load Nextflow

2. Create an Analysis Directory

Set up a directory for your analysis (referred to as the Analysis Directory):

mkdir $HOME/cutandrun_project
cd $HOME/cutandrun_project

• Modify $HOME/cutandrun_project to better suit your project.

3. Prepare Sample Sheet

Create a sample sheet (samplesheet.csv) with the following format:

sample,fastq_1,fastq_2,replicate,antibody
sample1,/path/to/sample1_R1.fastq.gz,/path/to/sample1_R2.fastq.gz,1,H3K27me3
sample2,/path/to/sample2_R1.fastq.gz,/path/to/sample2_R2.fastq.gz,1,IgG

Ensure all paths to the FASTQ files are accurate.

4. Configure ICER Environment

Create an icer.config file for SLURM:

process {
    executor = 'slurm'
}

5. Run nf-core/cutandrun

Example SLURM Job Submission Script

Below is a shell script for submitting an nf-core/cutandrun job to SLURM:

#!/bin/bash

#SBATCH --job-name=cutandrun_job
#SBATCH --time=48:00:00
#SBATCH --mem=48GB
#SBATCH --cpus-per-task=12

cd $HOME/cutandrun_project
module load Nextflow/23.10.0

nextflow pull nf-core/cutandrun
nextflow run nf-core/cutandrun -r 3.14.0 --input ./samplesheet.csv -profile singularity --outdir ./cutandrun_results --genome GRCh38 -work-dir $SCRATCH/cutandrun_work -c ./nextflow.config

• Modify --outdir and --genome to match your paths and reference genome.

Note on Reference Genomes

Common reference genomes are located in the research common-data space on the HPCC. Refer to the README file for details. For more guidance on downloading reference genomes from Ensembl, see this GitHub repository.

Execute the pipeline with the following command, including the -w flag for a separate working directory:

nextflow run nf-core/cutandrun -profile singularity --input samplesheet.csv --genome GRCh38 -c icer.config -w $SCRATCH/cutandrun_project

• Modify -w $SCRATCH/cutandrun_project as needed.

6. Monitor and Manage the Run

• Use squeue or sacct to track job status.
• Check the output directory for results.

Best Practices

• Review Logs: Regularly check log files for warnings or errors.
• Optimize Resource Usage: Adjust icer.config to match your dataset requirements.
• Manage Storage: Ensure sufficient storage for intermediate and final results.

Getting Help

If you encounter issues running nf-core/cutandrun on the HPCC, consider these resources:

• nf-core Community: Visit the nf-core website for documentation and support.
• ICER Support: Contact ICER via the MSU ICER support page.
• Slack Channel: Join the nf-core Slack for real-time assistance.
• Nextflow Documentation: See the Nextflow documentation for further details.

Conclusion

Running nf-core/cutandrun on the MSU HPCC is simplified using Singularity and Nextflow. This guide ensures reproducible and efficient CUT&RUN and CUT&Tag analysis, leveraging the HPCC’s computational capabilities for bioinformatics research.