RNAseq README

• RNAseq README
  • Flowchart
  • Pipeline overview
  • Outputs
  • Docker images
    • Naming convention [Documentation]
    • Modules and images
  • Gitlab repositories
  • Biodebian organisation
  • Submitting jobs to the cromwell server on biodebian
  • Blog posts on GeDaC webpage

Flowchart

Pipeline overview

Starting with fastQ files as input, this pipeline has the following stages:

1. Pre-alignment QC with fastp (adapter-trimming + QC) and multiQC report collation
2. Alignment with STAR
3. Post-alignment QC and analysis with:

• Infer library strandedness with RSeQC inferstrandedness.py
• Outputs Transcript Integrity number (TIN) with RSeQC tin.py for evenness of coverage assessment
• Outputs wig and normalised bigwig files for genome browser visualisations with RSeQC bam2wig and UCSC wig2BigWig binary

4. Quantification with FeatureCounts from Subread at the exon-level, grouped by gene_id into metafeatures.
5. [Conditional] If there are >=2 sample conditions and >=3 samples, we carry out the R secondary analysis which performs:

• Differential gene expression across all pairwise comparisons
• Overrepresentation analysis for pathway enrichment

Outputs

Name	Type	Extension	Description
fastpQcHtml	pre-alignment qc	.html files	fastp html reports
multiqcHtml	pre-alignment qc	.html file	multiqc html report
bamFiles	alignment results	.bam files	STAR sorted-by-coordinate bams
starLogs	post-alignment qc	log files	STAR log.final.out files
flagstat	post-alignment qc	log files	flagstat files
tin_summary	post-alignment qc	.txt files	rseqc tin summary files
tin_xls	post-alignment qc	.xls files	rseqc tin xls files
wigs	alignment results	.wig files	wig files
normalised bigwigs	alignment results	.bigwig files	bigwig files
countMatrix	quantification results	.txt	featureCounts count matrix
countsParsed	quantification results	.txt	Output of parse_counts.py which provides gene names
countsSummary	quantification results	.txt	featureCounts summary file
downstreamResDir	R secondary analysis results	.zip	R analysis results files

Docker images

Naming convention [Documentation]

• dockerBase = 026171442599.dkr.ecr.ap-southeast-1.amazonaws.com/
• dockerPrefix = species + genomeVersion
  • E.g. “human” + “grch38”
• dockerUri = dockerBase + dockerPrefix

Modules and images

For other species, replace docker image according to naming convention

Index	Module	Docker image name	Scripts/apps
1	countArrayUniqueItems.wdl	ubuntu	None
2	fastp.wdl	qc (apps_refless)	None
3	multiQc.wdl	qc (apps_refless)	None
4	pairsToR1R2.wdl	None	None
5	star.wdl	humangrch38rnaseq (fat Docker)	(ref_files/species): STAR indices, annotation.gtf
6	samtools.wdl	samtools (apps_refless)	None
7	rseqc.wdl	humangrch38rseqc (apps)	(ref_files/species): annotation.bed, houseKeepingGenes.bed, chrNameLength.txt; (ref_files/apps): wigToBigWig; (apps): processInferExperimentScript.py
8	featureCounts.wdl	humangrch38featureCounts (apps)	(ref_files/species): annotation.gtf; (apps) parse_counts.py
9	downstreamRNAseq.wdl	downstreamRNAseq (apps_refless)	(apps): main.R

Gitlab repositories

• Main: RNAseq repository
• Dockerfiles and configurations
• Modules (checkout rnaseq branch)

Biodebian organisation

See gedac documentation

Submitting jobs to the cromwell server on biodebian

The options.json file is a dummy file required for running the cromshell submit subcommand. Currently, cromwell has the following IP: 172.18.149.93:7098.

cromshell submit ./main.wdl ./tests/tests_biodebian/rnaseq_mouse_grcm39.json ./options.json ./modules.zip

Blog posts on GeDaC webpage

• GeDaC RNA-seq pipeline Walkthrough: Interpreting the Results
• GeDaC RNA-seq pipeline Walkthrough: How does the pipeline work?