Running Juicer on the HPC

Setting up your juiceDir

To get started with Juicer, you will need to have a directory where your input files exist. The juiceDir should contain a folder called fastq, a folder called references, and a folder called restriction_sites. The names should be exactly as stated here as juicer internally looks for folders of these names.

Once you have created these folders and put your data into each one respectively (raw fastq into the fastq folder, reference genomes into the references folder etc.) you are ready to load the juicer module.

$ module load aidenlab/juicer/create_sbatch

Loading this module will ask you to provide the location to your juicer directory, for example:

Please provide the path to your juiceDir:
/hpc/faculty/azeezoe/new_juicer_directory

This will place two .sh files in your juiceDir and a scripts symlink to the juicer scripts.

(juicer) [azeezoe@hpc1 new_juicer_directory]$ ls
sbatch_hiccups_cpu.sh  sbatch_juicer_cpu.sh  scripts

Running juicer

To launch the juicer pipeline modify sbatch_juicer_cpu.sh and complete the variables and sbatch parameters to match your experiment

#!/bin/bash
#SBATCH --job-name=juicer_cpu_slurm_submit
#SBATCH --output=%x_%j.out
#SBATCH --error=%x_%j.err
#SBATCH --account=
#SBATCH --partition=
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task=8
#SBATCH --mem=100
#SBATCH --mail-type=END,FAIL
#SBATCH --mail-user=
#SBATCH --time=7-00:00:00


## NOTE ##
## Don't forget to update the account, partition, mail-user, cpus, and memory variables above ^^ !!!

module load aidenlab/juicer/juicer

juiceDir=/hpc/faculty/azeezoe/new_juicer_directory
genomeID=
threads=
reference_genome_file=
chrom_sizes_file=
restriction_sites_file=
site=

You can then launch the juicer pipeline using sbatch sbatch_juicer_cpu.sh and while it's running, modify the sbatch_hiccups_cpu.sh script. Only submit that script with sbatch sbatch_hiccups_cpu.sh after the juicer pipeline has finished.

Loading dependencies

If you have already set up your directory and run module load aidenlab/juicer/create_sbatch once and all you want to do is load the Juicer dependencies so you can continue an experiment, you can run module load aidenlab/juicer/juicer