RNA-seq Pipeline for HPC

This repository contains a comprehensive RNA-seq analysis pipeline tailored for High-Performance Computing (HPC) environments, utilizing SLURM for job scheduling. The pipeline processes RNA-seq data from raw reads to downstream analyses, including alternative splicing using RMATS.

Features

• Download RNA-seq data from the SRA database.
• Convert SRA files to FASTQ format.
• Quality control and trimming of raw reads.
• Alignment to a reference genome (using STAR or other preferred aligners).
• BAM file sorting and indexing.
• Differential alternative splicing analysis using RMATS.
• Containerized execution using Singularity or Docker.
• Parallelized and efficient job execution designed for HPC environments.

SLURM Job Support

Each process in the pipeline can be run individually as a SLURM job, allowing for flexibility in the execution of RNA-seq tasks. Individual SLURM sbatch scripts are provided for each step, giving users control over the workflow.

Nextflow Integration

In addition to individual scripts, the repository includes a modular Nextflow script that enables users to select which processes to run. The Nextflow pipeline is designed to be executed as a SLURM job, providing:

• A streamlined and automated workflow.
• Modular execution, allowing users to choose specific processes.
• The ability to harness the power of parallel processing with Nextflow.

Usage

Running Individual Scripts

To run individual processes as SLURM jobs, follow these steps:

1. Prepare SLURM sbatch Scripts: Review and adjust the provided SLURM sbatch scripts for your HPC environment, ensuring the parameters match your computational resources and requirements. All slurm scripts are in slurm_scripts directory.

2. Submit SLURM Jobs: Use the sbatch command to submit each process to the SLURM queue. For example:

   sbatch sra_download.sh
   sbatch star_alignment.sh
   sbatch bam_sort.sh
   sbatch rmats.sh
   

Running the Nextflow Pipeline

To run the entire pipeline using Nextflow with SLURM, follow these steps:

1. Prepare Environment: Ensure you have Nextflow installed and configured, and that your HPC environment supports Nextflow and SLURM. You may also need to load or activate the Nextflow environment. For example:

   source activate nextflow_env
   

2. Edit SLURM Job Script: Customize the rnaseq_pipeline_slurm.sh script with your specific parameters and resource settings.

3. Submit Nextflow Job: Submit the job to the SLURM queue using sbatch:

   sbatch rnaseq_pipeline_slurm.sh
   

Parameters

• -entry [optional] workflow_name: Select the specific workflow to execute (e.g., do_full, do_fastq2bam, do_rmats, do_rmatsv2).

• -cartpath "path/to/cartfile": Path to the SRA cartfile for data download.

• -ngcpath "path/to/ngcfile": Path to the NGC file for SRA authentication.

• -genomedir "path/to/ref/genome/directory": Path to the reference genome directory for alignment.

• -gtfpath "path/to/annotated/gtf": Path to the GTF file for gene annotation.

• -rmats_b1 "path/to/rmats/batch1/files/list/txt": Path to the list of batch 1 files for RMATS analysis.

• -rmats_b2 "path/to/rmats/batch2/files/list/txt": Path to the list of batch 2 files for RMATS analysis.

• -singularity_image "path/to/singularity/image/for/rmats": Path to the Singularity image for RMATS analysis.

• -modulepath "path/to/modulefiles": Path to modulefiles directory.

• -sratoolkit "sratoolkit/[module version OPTIONAL]": SRA Toolkit module.

• -STAR "STAR/[module version OPTIONAL]": STAR aligner module.

• -samtools "samtools/[module version OPTIONAL]": Samtools module.

• -singularity "singularity/[module version OPTIONAL]": Singularity module.

• -intel "intel/[module version OPTIONAL]": Intel module.

• -pgi "pgi/[module version OPTIONAL]": PGI module.

• -rmats "rmats/[module version OPTIONAL]": RMATS module.

• -workingdir $(PWD): Working directory for intermediate and output files.

• -cpus ${SLURM_CPUS_PER_TASK}: Number of CPUs allocated for the job.

• -memory ${SLURM_MEM}: Amount of memory allocated for the job.

• -queue 'your_queue': SLURM queue to use for job submission.

Contribution

Feel free to contribute to this project by opening issues or submitting pull requests. Contributions are welcome to improve functionality, add features, or fix bugs.

License

This project is licensed under the MIT License. See the LICENSE file for details.