rnaseq_workflows's Introduction

RNAseq Workflows:

1. STAR 2-pass

Overview

This pipeline was developed to map RNAseq data using the STAR (Spliced Transcripts Alignment to a Reference) 2-pass methodology. This pipeline is paired with variant calling using GATK in the repository dthorburn/Genomic_Read_Processing. The pipeline was developed using Nextfow version 22.04.4 (version on Imperial HPC; date: 19/07/2022).

EDIT: The Imperial HPC long nodes are no longer available. STAR is now more efficient to use as an array job. Please use the starAlign.sh script instead. The 15 samples listed in 00_File_List.txt are a mix of paired and single end sequencing. You will need to change the values to define chromosomes if the number of samples change.

Usage

Update the paths in 00_File_List.txt, update the working directory path in starAlign.sh, alter the PBS_ARRAY_INDEX statements to handle the chromosomes and samples correctly, and submit using qsub starAlign.sh

2. New Tuxedo Protocol

Overview

This pipeline was developed to map RNAseq data using the "new Tuxedo protocol" (i.e., using HISAT2, StringTie, and Ballgown). This pipeline was developed using BASH since for a high number of small jobs, a BASH array is more efficient at processing the samples than a nextflow script without the use of the deprecated Imperial HPC long nodes. On request I can update this to a nextflow workflow.

Usage

First step is to create the RNAseq conda environment: conda env create -f /path/to/RNAseq.yml

Update the paths in 00_R1_List.txt and update the Reference, Annotations, and Assembly variables to reflect the reference assembly in use. Next, update the number of jobs in the array #PBS -J 1-X where X is the number of samples in use in script numbers 01, 02, and 04. Then submit each of the jobs in order from 00 to 04 using qsub. The job paramaters should be sufficient for most use cases.

Recommend Projects

dthorburn / rnaseq_workflows Goto Github PK