Simple RNA-Seq pipeline (rnaseq-pipeline-2)

Differential gene expression pipeline utilising Salmon and Bioconductor DESeq2 for Illumina RNA-Seq sequencing results.

For detail see the following publication:

Bukowski M, Kosecka-Strojek M, Wladyka B. Disruption of saoB and saoC genes in Staphylococcus aureus alters transcription of genes involved in amino acid metabolism and virulence. [awaiting publication]

If you find the pipeline useful, please cite it.

1. Environment

Miniconda/Anaconda installation is a prerequisite for running the pipeline. The pipeline was created and tested using the following conda environments and crucial packages:

• rnaseq-base:
  • sra-tools (2.8.0), only for test data acquisition
  • nextflow (22.10.0)
• rnaseq-salmon:
  • salmon (1.10.2)
• rnaseq-data:
  • python (3.12.0)
  • r-base (4.3.2)
  • bioconductor-deseq2 (1.42.0)
  • bioconductor-apeglm (1.24.0)

The first environment is required for obtaining the test data and running the pipeline. Remaining environemnts (see YML files in envs/ directory) are installed by the pipeline. Run the following commands from the pipeline directory to recreate and activate the rnaseq-base environment:

conda env create -f envs/rnaseq-base.yml
conda activate rnaseq-base

2. Test data acquisition

To download test data (read archives) from NCBI SRA run the test_data.sh script. This script will fetch 12 read archives to reads/ directory (~3.7 GB). It may take a while. Any problems are usually related to the accesibility of NCBI servers. If something does not work, try again later.

./test_data.sh

3. Directory structure and pipeline files

Names of the FASTQ files with reads in NCBI BioProject PRJNA798259 follow the pattern, which is required by the pipeline: {strain}_{group}_{replica}_{reads}.fastq, e.g. rn_wt_lg_1_R1.fastq. Regarding the strain, in the project data there are files only for rn (Staphylococcus aureus RN4220). There are two genotypes: wt (wild type, the reference group), mt (mutant, saoC gene mutant), sampled in two settings/conditions: lg (logarithmic growth phase) and lt (late growth phase). That two pairs of groups (wt_lg, mt_lg and wt_lt, mt_lt). For each there are 3 replicas (1 -- 3). Reads of each are written to two files: R1 and R2. All in all, there are 24 files.

The pipeline utilises the following directory structure:

your_pipeline_location/
├── envs/
│   ├── rnaseq-base.yml
│   ├── rnaseq-data.yml
│   └── rnaseq-salmon.yml
├── refs/
│   └── rn.fna
├── templates/
│   ├── dge.r
│   ├── index.sh
│   ├── merge.sh
│   └── quant.sh
├── reads/
├── work/
├── output/
├── nextflow.config
└── main.nf

In the working directory you can find the main.nf file that describes the pipeline. The experimental design is described in nextflow.config file. Template scripts are in template/. In input/refs/ you will have rn.fna multiple FASTA file with reference transcript sequences for Staphylococcus aureus RN4220. In input/reads/ you will find 24 read archives downloaded with test_data.sh script. Directories work/ and output/ will be created automatically once the pipeline is run. The latter will contain final output from each stage of the analysis.

4. Pipeline architecture

The pipline described in the main.nf Nexflow file implements the following processes:

1. salmonIndex -- preparation of an index of reference transcript sequences with Salmon.
2. salmonQuant -- read mapping and counts with Salmon.
3. salmonQuantMerge -- gathering TPM and NumRead values for replicas assigned to one experiment.
4. calculateDGE -- differential gene expression with utilising DESeq2 library for each experiment.

More detailed description on how the pipeline works you will find in comments included in the main.nf and nexflow.config files.

5. Running the pipeline

To start the pipeline, once you have read data and experiment design description done as well as rnaseq-base conda environment activated, run the following command from the pipeline directory:

nextflow main.nf