This pipeline is largely based on the nf-core/rnaseq pipeline. The main purpose of this repository is to provide a tutorial for running the nf-core/rnaseq pipeline on the MSU HPCC using SLURM as the job executor.
This is a workflow to quantify transcript-level RNASeq abundance, detect differentially expressed genes (DEGs), and perform gene set enrichment analysis ('GSEA'). This pipeline uses STAR to map FASTQ reads to the reference genome, Salmon to perform BAM-level quantification, and DESeq2 to normalize counts and detect DEGs.
STAR typically uses around 38GB of RAM. It is recommended to run this workflow on an HPC.
- Create a directory for your your analysis (ex: $HOME/rnaseq).
- Make a samplesheet table for pre-processing in csv format. (ex: samplesheet.csv)
- Make a nextflow configuration file. (ex: nextflow.config)
- Download the reference genome and gtf files into your directory.
- Write a bash script to run the pre-processing pipeline using SLURM. (ex: run_rnaseq.sb)
- Run the pre-processing pipeline from your rnaseq directory. (ex: sbatch run_rnaseq.sb)
- Make a samplesheet table for differential expression analysis in csv format. (ex: DE_samplesheet.csv)
- Write a bash script to run the differential expression pipeline using SLURM. (ex: run_differential.sb)
Login to your HPCC account using OnDemand. Navigate to your home directory by clicking 'Files' in the navigation bar. Select 'Home Directory'.
Create a directory for your analysis by clicking 'New Directory'. Name your directory (ex: rnaseq). Navigate to the newly created rnaseq directory.
Make sure to upload your data (FASTQ files) into this directory.
In your rnaseq directory, click 'New File'. Name the file 'samplesheet.csv'. Click the ⋮ symbol and select edit. Create the samplesheet table FOR YOUR DATA. Below is a template:
sample,fastq_1,fastq_2,strandedness
CONTROL_REP1,/path/to/fastq/files/CONTROL_REP1_R1_001.fastq.gz,/path/to/fastq/files/CONTROL_REP1_R2_001.fastq.gz,auto
CONTROL_REP2,/path/to/fastq/files/CONTROL_REP2_R1_001.fastq.gz,/path/to/fastq/files/CONTROL_REP2_R2_001.fastq.gz,auto
CONTROL_REP3,/path/to/fastq/files/CONTROL_REP3_R1_001.fastq.gz,/path/to/fastq/files/CONTROL_REP3_R2_001.fastq.gz,auto
TREATED_REP1,/path/to/fastq/files/TREATED_REP1_R1_001.fastq.gz,/path/to/fastq/files/TREATED_REP1_R2_001.fastq.gz,auto
TREATED_REP2,/path/to/fastq/files/TREATED_REP2_R1_001.fastq.gz,/path/to/fastq/files/TREATED_REP2_R2_001.fastq.gz,auto
TREATED_REP3,/path/to/fastq/files/TREATED_REP3_R1_001.fastq.gz,/path/to/fastq/files/TREATED_REP3_R2_001.fastq.gz,auto
Note: the samplesheet above is an example to show the required format. You will still need to make a samplesheet FOR YOUR DATA.
Save the samplesheet.csv file and return to your rnaseq directory.
In your rnaseq directory, click New File. Name the file 'nextflow.config'. Click the ⋮ symbol and select edit. Create the Nextflow config file:
process {
executor = 'slurm'
}
Save the nextflow.config file and return to your rnaseq directory.
Find your reference genome in ICER common-data OR download the reference genome and gtf files in your rnaseq directory
The following organisms have updated reference genomes and gtf/gff3 files in common-data: human, mouse, rat, zebrafish, and Arabidopsis. The paths to those files can be found here.
In your rnaseq directory, click Open in Terminal to enter a development node. Download the most recent genome primary assembly and gtf for the organism of interest from Ensembl. Follow the instructions here to download the reference genome for your organism of interest. This may take more than a few minutes. For example:
wget https://ftp.ensembl.org/pub/release-108/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz
wget https://ftp.ensembl.org/pub/release-108/gtf/homo_sapiens/Homo_sapiens.GRCh38.108.gtf.gz
Note: If you do not download the genome and gtf for the organism that the RNA-seq data is derived from, this pipeline will not work correctly.
In your rnaseq directory, click New File. Name the file 'run_rnaseq.sb;. Write the bash script, using #SBATCH directives to set resources, for example:
#!/bin/bash
#SBATCH --job-name=$jobname_rnaseq
#SBATCH --time=24:00:00
#SBATCH --mem=24GB
#SBATCH --cpus-per-task=8
cd $HOME/rnaseq
module load Nextflow/23.10.0
nextflow pull nf-core/rnaseq
nextflow run nf-core/rnaseq -r 3.14.0 --input ./samplesheet.csv -profile singularity --outdir ./rnaseq_results --fasta ./Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz --gtf ./Homo_sapiens.GRCh38.108.gtf.gz -work-dir $SCRATCH/rnaseq_work -c ./nextflow.config
Note: above is an example script. You will need to write one specific to YOUR DATA using this template.
Save the run_rnaseq.sb file and return to your rnaseq directory.
In your rnaseq directory, click Open in Terminal to enter a development node. Run jobs on the SLURM cluster:
sbatch run_rnaseq.sb
Check the status of your pipeline job:
squeue -u $username
Note: replace $username with your username
In your rnaseq directory, click 'New File'. Name the file 'DE_samplesheet.csv'. Click the ⋮ symbol and select edit. Create the samplesheet table FOR YOUR DATA. Below is a template:
sample,fastq_1,fastq_2,condition,replicate,batch
CONTROL_REP1,/path/to/fastq/files/CONTROL_REP1_R1_001.fastq.gz,/path/to/fastq/files/CONTROL_REP1_R2_001.fastq.gz,control,1,
CONTROL_REP2,/path/to/fastq/files/CONTROL_REP2_R1_001.fastq.gz,/path/to/fastq/files/CONTROL_REP2_R2_001.fastq.gz,control,2,
CONTROL_REP3,/path/to/fastq/files/CONTROL_REP3_R1_001.fastq.gz,/path/to/fastq/files/CONTROL_REP3_R2_001.fastq.gz,control,3,
TREATED_REP1,/path/to/fastq/files/TREATED_REP1_R1_001.fastq.gz,/path/to/fastq/files/TREATED_REP1_R2_001.fastq.gz,treated,1,
TREATED_REP2,/path/to/fastq/files/TREATED_REP2_R1_001.fastq.gz,/path/to/fastq/files/TREATED_REP2_R2_001.fastq.gz,treated,2,
TREATED_REP3,/path/to/fastq/files/TREATED_REP3_R1_001.fastq.gz,/path/to/fastq/files/TREATED_REP3_R2_001.fastq.gz,treated,3,
Note: the samplesheet above is an example to show the required format. You will still need to make a samplesheet FOR YOUR DATA.
Save the samplesheet.csv file and return to your rnaseq directory.
In your rnaseq directory, click 'New File'. Name the file 'contrasts.csv'. Click the ⋮ symbol and select edit. Create the contrasts file FOR YOUR DATA. Below is a template:
id,variable,reference,target,blocking
condition_control_treated,condition,control,treated,
Note: the contrasts file above is an example to show the required format. You will still need to make a contrasts file FOR YOUR DATA.
Save the contrasts.csv file and return to your rnaseq directory.
In your rnaseq directory, click New File. Name the file 'run_differential.sb;. Write the bash script, using #SBATCH directives to set resources, for example:
#!/bin/bash
#SBATCH --job-name=$jobname_differential
#SBATCH --time=24:00:00
#SBATCH --mem=24GB
#SBATCH --cpus-per-task=8
cd $HOME/rnaseq
module load Nextflow/23.10.0
nextflow pull nf-core/differentialabundance
nextflow run nf-core/differentialabundance -r 1.5.0 --input ./samplesheet.csv --contrasts ./contrasts.csv --matrix ./rnaseq_results/star_salmon/salmon.merged.gene_counts_length_scaled.tsv --gtf ./Homo_sapiens.GRCh38.108.gtf.gz --outdir ./differential_results -profile singularity -work-dir $SCRATCH/differential_work -c ./nextflow.config
Note: above is an example script. You will need to write one specific to YOUR DATA using this template.
Save the run_rnaseq.sb file and return to your rnaseq directory.
In your rnaseq directory, click Open in Terminal to enter a development node. Run jobs on the SLURM cluster:
sbatch run_differential.sb
Check the status of your pipeline job:
squeue -u $username
Note: replace $username with your username