This pipeline implements a comprehensive RNA-seq workflow with four main stages:

Per-sample stages:

• TrimAlignRNA: Trims paired FASTQ files using fastp and aligns reads to the genome with STAR. Outputs a CRAM file with index (<dataset_prefix>/cram/<sequencing_group_id>.cram and .crai) and temporary BAM file with index (<tmp_prefix>/bam/<sequencing_group_id>.bam and .bai).

• Count: Quantifies gene/transcript read counts from aligned reads using featureCounts. Produces count files (<dataset_prefix>/count/<sequencing_group_id>.count) and summary statistics (<dataset_prefix>/count/<sequencing_group_id>.count.summary).

Cohort-level stages:

• Fraser: Performs aberrant splicing analysis across samples in a cohort. Consumes BAM files (preferred) or CRAM files and generates an FDS archive (<dataset_prefix>/fraser/<cohort_id>.fds.tar.gz).

• Outrider: Conducts outlier gene expression analysis using count data from all samples in a cohort. Outputs results as R data files (<dataset_prefix>/outrider/<cohort_id>.outrider.RData).

• Integration of additional QC metrics and visualization tools. (Integrating PICARD.)
• Updating Fraser to its latest version for improved splicing analysis.
• Optimization to improve scalability and efficiency.

analysis-runner \
    --dataset seqr \
    --image australia-southeast1-docker.pkg.dev/cpg-common/images-dev/rdrnaseq:0.2.0-1 \
    --skip-repo-checkout \
    --description "RNA-seq analysis" \
    -o "output-description" \
    --access-level full \
    --config src/rdrnaseq/config_template.toml \
    python3 src/rdrnaseq/run_workflow.py