Technical Architecture

The CLI is implemented in Python under src/aree.

• validation: JSON Schema validation.
• intake: registry ingestion.
• harmonize: processed-result conversion and identifier mapping.
• meta_analysis: random-effects synthesis.
• prioritize: candidate scoring.
• reporting: evidence cards and demo reports.
• raw: shared helpers for raw-data reanalysis scripts: ENA run manifests, checksum-verified resumable FASTQ downloads, sample sheets, RefSeq GFF parsing, and DESeq2-to-evidence export.

Study-specific scripts in scripts/ keep each study’s contrasts, QC gates and report text, and import the shared machinery from aree.raw, so they work both as python scripts/NAME.py and when imported.

The reusable Nextflow pipeline in workflows/rnaseq/ supports paired-end RNA-seq reanalysis (see its README). Committed public-study analyses also use the study-specific scripts and shell wrappers in scripts/; these preserve design-specific choices, including QuantSeq processing. Methylation, proteomics and metabolomics studies currently enter through processed-result harmonization; raw-data pipelines for them are on the roadmap.