pa-cwl: Production-ready CWL v1.2 workflows for AI-agent execution via WES
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
pa-cwl (Pretty Agentic CWL) is a collection of 16 production-ready CWL v1.2 workflows for bioinformatics analysis, designed for AI-agent execution via the GA4GH Workflow Execution Service (WES) standard. Workflows (16 pipelines) Data Retrieval: fetchngs Transcriptomics: rnaseq, scrnaseq, rnafusion Epigenomics: chipseq, atacseq, methylseq, cutandrun Variant Calling: sarek, raredisease, viralrecon Metagenomics: ampliseq, mag, taxprofiler Long-Read & 3D Genomics: nanoseq, hic Contents pa-cwl-workflows-v1.0.tar.gz — 16 CWL workflows, 124 reusable tools, AI-agent specifications (agent.yaml), 30 Workflow Run RO-Crate provenance records, schemas, documentation, and CI configuration pa-cwl-test-data-v1.0.tar.gz — Simulated test datasets covering 8 biological domains (yeast, nanopore, metagenome, viral, Hi-C, amplicon, fusion, taxonomic) plus pre-built indices for reproducible testing Features CWL v1.2 with conditional steps, scatter/gather parallelism, and ExpressionTools BioContainers Docker images with pinned versions Machine-readable agent.yaml for LLM-driven workflow execution Workflow Run RO-Crate provenance from Sapporo WES 2.2.4 35 integration tests (25 passing on macOS ARM, all expected to pass on x86_64 Linux)
Fields of Study & Subjects
Scientific Computing and Data Management
Genomics and Phylogenetic Studies
Genetics
Bioinformatics
and Biomedical Research
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 26 March 2026
- Language
- EN