{"id":17695,"plugin_id":"plugins_6a76572d8f8081918362aa7ff90947fb","kind":"skill","collection_source":null,"comparison_source":null,"observed_at":"2026-09-30T23:14:19.516Z","digest":"2146a4de6c6b48db5f9f1929161cd62a1bd4bff45c9dcdf4b2f3c98199c8c2e6","against":null,"payload":{"description":"Route NVIDIA Parabricks pbrun tools, assess GPU/runtime readiness, and provide version-aware command guidance for FASTQ/BAM processing, RNA-seq, variant calling, BAM QC, and GVCF workflows. Do NOT use for inspecting or accelerating whole pipelines — use genomics-workflow-acceleration.","included_files":[{"relative_path":"references/parabricks-rna-validate.md","size_in_bytes":6362},{"relative_path":"references/pbrun-applybqsr.md","size_in_bytes":4440},{"relative_path":"references/pbrun-bam2fq.md","size_in_bytes":4445},{"relative_path":"references/pbrun-bammetrics.md","size_in_bytes":6876},{"relative_path":"references/pbrun-bamsort.md","size_in_bytes":4283},{"relative_path":"references/pbrun-bqsr.md","size_in_bytes":4056},{"relative_path":"references/pbrun-collectmultiplemetrics.md","size_in_bytes":8048},{"relative_path":"references/pbrun-dbsnp.md","size_in_bytes":3064},{"relative_path":"references/pbrun-deepsomatic.md","size_in_bytes":4541},{"relative_path":"references/pbrun-deepvariant.md","size_in_bytes":5214},{"relative_path":"references/pbrun-deepvariant_germline.md","size_in_bytes":5567},{"relative_path":"references/pbrun-fq2bam.md","size_in_bytes":11271},{"relative_path":"references/pbrun-fq2bam_meth.md","size_in_bytes":8370},{"relative_path":"references/pbrun-genotypegvcf.md","size_in_bytes":3324},{"relative_path":"references/pbrun-germline.md","size_in_bytes":5414},{"relative_path":"references/pbrun-giraffe.md","size_in_bytes":7299},{"relative_path":"references/pbrun-haplotypecaller.md","size_in_bytes":5655},{"relative_path":"references/pbrun-indexgvcf.md","size_in_bytes":3087},{"relative_path":"references/pbrun-markdup.md","size_in_bytes":5587},{"relative_path":"references/pbrun-minimap2.md","size_in_bytes":6147},{"relative_path":"references/pbrun-mutectcaller.md","size_in_bytes":4502},{"relative_path":"references/pbrun-ont_germline.md","size_in_bytes":4555},{"relative_path":"references/pbrun-pacbio_germline.md","size_in_bytes":4921},{"relative_path":"references/pbrun-pangenome_aware_deepvariant.md","size_in_bytes":4339},{"relative_path":"references/pbrun-pangenome_germline.md","size_in_bytes":4848},{"relative_path":"references/pbrun-postpon.md","size_in_bytes":3127},{"relative_path":"references/pbrun-prepon.md","size_in_bytes":3307},{"relative_path":"references/pbrun-rna_fq2bam.md","size_in_bytes":10724},{"relative_path":"references/pbrun-somatic.md","size_in_bytes":5033},{"relative_path":"references/pbrun-starfusion.md","size_in_bytes":10371},{"relative_path":"references/runtime-environment.md","size_in_bytes":8530},{"relative_path":"references/tool-index.md","size_in_bytes":3467},{"relative_path":"scripts/check_parabricks_runtime.py","size_in_bytes":20328}],"name":"parabricks","skill_md_contents":"---\nname: parabricks\ndescription: >-\n  Route NVIDIA Parabricks pbrun tools, assess GPU/runtime readiness, and provide\n  version-aware command guidance for FASTQ/BAM processing, RNA-seq, variant\n  calling, BAM QC, and GVCF workflows. Do NOT use for inspecting or accelerating\n  whole pipelines — use genomics-workflow-acceleration.\nlicense: CC-BY-4.0 AND Apache-2.0\nmetadata:\n  version: \"1.1.0\"\n  tags:\n    - parabricks\n    - genomics\n    - nvidia\n---\n\n# Parabricks\n\n## Purpose\n\nUse this skill to discover the right NVIDIA Parabricks `pbrun` command, assess\nruntime readiness, and generate version-aware command guidance for individual\ntools and pipelines.\n\nDo **not** use this skill for whole-workflow inspection, acceleration planning,\nor wiring optional GPU branches. For pipeline-level work, use\n`genomics-workflow-acceleration`.\n\n## When to Use This Skill\n\n- Which `pbrun` tool fits the user's data and goal\n- GPU, driver, Docker, container, storage, or installation readiness\n- Command shape, flags, and validation for a specific Parabricks tool\n- Troubleshooting a single Parabricks command or tool family\n\n## Prerequisites\n\nAsk for input data type, sequencing technology, reference build, sample\nstructure, desired output, target Parabricks version/container tag, and runtime\ntarget before recommending commands.\n\nIf the user is unsure which tool applies, read\n[tool-index.md](references/tool-index.md) first, then load the matching\n`references/pbrun-<tool>.md` file.\n\n## Limitations\n\nThis skill routes and guides Parabricks commands. It does not install\nParabricks, infer missing sample metadata, guarantee output parity, provide\nclinical interpretation, or promise exact runtime without benchmark data.\n\n## Workflow\n\n1. Confirm the Parabricks version or container tag. Verify the current NVIDIA\n   docs when the user asks for the latest tool list or version-sensitive flags.\n2. Classify the request:\n   - **Runtime** → [runtime-environment.md](references/runtime-environment.md)\n   - **Tool discovery** → [tool-index.md](references/tool-index.md)\n   - **Specific command** → matching `references/pbrun-<tool>.md`\n3. Collect missing biological and filesystem context before generating commands.\n4. Generate conservative Docker commands with explicit mounts, workdir, and\n   placeholders. Validate paths, indexes, and outputs after command generation.\n\n## Tool Reference Index\n\nLoad only the reference file for the selected tool.\n\n| Tool | Reference | Use when |\n|------|-----------|----------|\n| `applybqsr` | [pbrun-applybqsr.md](references/pbrun-applybqsr.md) | Apply BQSR table to aligned BAM |\n| `bam2fq` | [pbrun-bam2fq.md](references/pbrun-bam2fq.md) | BAM → FASTQ conversion |\n| `bamsort` | [pbrun-bamsort.md](references/pbrun-bamsort.md) | Standalone BAM sort |\n| `bqsr` | [pbrun-bqsr.md](references/pbrun-bqsr.md) | Generate BQSR recalibration table |\n| `fq2bam` | [pbrun-fq2bam.md](references/pbrun-fq2bam.md) | Short-read DNA paired FASTQ → BAM/CRAM |\n| `fq2bam_meth` | [pbrun-fq2bam_meth.md](references/pbrun-fq2bam_meth.md) | Bisulfite/methylation FASTQ → BAM/CRAM |\n| `giraffe` | [pbrun-giraffe.md](references/pbrun-giraffe.md) | Pangenome graph alignment |\n| `markdup` | [pbrun-markdup.md](references/pbrun-markdup.md) | Standalone duplicate marking |\n| `minimap2` | [pbrun-minimap2.md](references/pbrun-minimap2.md) | Long-read FASTQ alignment |\n| `rna_fq2bam` | [pbrun-rna_fq2bam.md](references/pbrun-rna_fq2bam.md) | RNA-seq FASTQ(s) → splice-aware BAM (STAR alignment) |\n| `starfusion` | [pbrun-starfusion.md](references/pbrun-starfusion.md) | Fusion detection from chimeric junction input + STAR-Fusion genome library |\n| `germline` | [pbrun-germline.md](references/pbrun-germline.md) | GATK-style germline pipeline from FASTQ |\n| `deepvariant_germline` | [pbrun-deepvariant_germline.md](references/pbrun-deepvariant_germline.md) | DeepVariant germline pipeline from FASTQ |\n| `haplotypecaller` | [pbrun-haplotypecaller.md](references/pbrun-haplotypecaller.md) | Standalone HaplotypeCaller from BAM/CRAM |\n| `deepvariant` | [pbrun-deepvariant.md](references/pbrun-deepvariant.md) | Standalone DeepVariant from BAM/CRAM |\n| `somatic` | [pbrun-somatic.md](references/pbrun-somatic.md) | Tumor-normal somatic pipeline |\n| `mutectcaller` | [pbrun-mutectcaller.md](references/pbrun-mutectcaller.md) | Mutect2-compatible somatic calling |\n| `deepsomatic` | [pbrun-deepsomatic.md](references/pbrun-deepsomatic.md) | DeepSomatic-based somatic calling |\n| `pacbio_germline` | [pbrun-pacbio_germline.md](references/pbrun-pacbio_germline.md) | PacBio long-read germline |\n| `ont_germline` | [pbrun-ont_germline.md](references/pbrun-ont_germline.md) | Oxford Nanopore long-read germline |\n| `pangenome_germline` | [pbrun-pangenome_germline.md](references/pbrun-pangenome_germline.md) | Pangenome-aware germline |\n| `pangenome_aware_deepvariant` | [pbrun-pangenome_aware_deepvariant.md](references/pbrun-pangenome_aware_deepvariant.md) | Pangenome-aware DeepVariant |\n| `prepon` | [pbrun-prepon.md](references/pbrun-prepon.md) | Pangenome-aware preprocessing |\n| `postpon` | [pbrun-postpon.md](references/pbrun-postpon.md) | Pangenome-aware post-processing |\n| `bammetrics` | [pbrun-bammetrics.md](references/pbrun-bammetrics.md) | Whole-genome coverage/depth metrics |\n| `collectmultiplemetrics` | [pbrun-collectmultiplemetrics.md](references/pbrun-collectmultiplemetrics.md) | Multiple Picard/GATK-style alignment metrics |\n| `genotypegvcf` | [pbrun-genotypegvcf.md](references/pbrun-genotypegvcf.md) | Joint-genotype GVCF input(s) into VCF |\n| `indexgvcf` | [pbrun-indexgvcf.md](references/pbrun-indexgvcf.md) | Index GVCF input |\n| `dbsnp` | [pbrun-dbsnp.md](references/pbrun-dbsnp.md) | dbSNP annotation on variant files |\n\nFor routing heuristics when multiple tools could apply, see\n[tool-index.md](references/tool-index.md).\n\n## Runtime Readiness\n\nFor GPU, driver, Docker, container, storage, or installation questions, read\n[runtime-environment.md](references/runtime-environment.md) and prefer:\n\n```bash\npython3 skills/parabricks/scripts/check_parabricks_runtime.py\n```\n\nAdd `--path <dir>` for known input/output/tmp paths. Run container probes only\nwith user consent.\n\n## Command Shape\n\n```bash\ndocker run --rm --gpus all \\\n  --volume /host/input:/workdir \\\n  --volume /host/output:/outputdir \\\n  --workdir /workdir \\\n  nvcr.io/nvidia/clara/clara-parabricks:<version> \\\n  pbrun <selected-tool> \\\n  <tool-specific-options>\n```\n\nCheck the version-specific tool reference before finalizing flags.\n\n## Troubleshooting\n\n| Error | Cause | Solution |\n|-------|-------|----------|\n| Multiple plausible tools | Data type or goal underspecified | Ask for assay, inputs, caller preference, desired output; use tool-index |\n| Exact flag requested | Options are version-sensitive | Check the selected tool reference and NVIDIA docs |\n| Runtime question | GPU, Docker, drivers, or storage | Use runtime-environment reference and diagnostic script |\n| Wrong tool family | Assay or input type unclear | Confirm DNA/RNA/methylation/long-read/pangenome before routing |\n| CUDA or memory failure | Runtime not ready or GPU memory constrained | Assess runtime before tuning command flags |\n\n## Guardrails\n\n- Treat command availability and options as version-sensitive.\n- Do not infer exact flags from command names alone.\n- Do not collapse standalone tools and full pipelines when explaining tradeoffs.\n- Do not substitute DNA `fq2bam` for RNA, or germline for somatic callers.\n- Do not invent sample names, read groups, reference builds, known-sites files,\n  model files, graph resources, container tags, or output paths.\n- Do not install, upgrade, or modify packages. Label setup commands as user-run.\n- Do not claim CPU execution of Parabricks tools.\n- Do not claim biological or VCF parity without a comparison run.\n- Prefer official NVIDIA docs for exact command syntax and option defaults.\n\n## Key References\n\n- Parabricks tool index:\n  <https://docs.nvidia.com/clara/parabricks/latest/toolreference.html>\n- Output accuracy and compatible CPU software versions:\n  <https://docs.nvidia.com/clara/parabricks/latest/documentation/tooldocs/outputaccuracyandcompatiblecpusoftwareversions.html>\n- Getting started:\n  <https://docs.nvidia.com/clara/parabricks/latest/gettingstarted.html>\n- Overview:\n  <https://docs.nvidia.com/clara/parabricks/latest/overview.html>\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}