# MLM Manhattan and QQ plotting skill

Use this skill only for the PlantAI `gwas_plot` stage.

- Canonical runner: `python -m apps.cli.run_gwas_plot`
- Dependencies: `gwas_mlm`
- Inputs: `mlm_pvalues`, `cmlm_pvalues`
- Outputs: `mlm_manhattan`, `mlm_qq`, `cmlm_manhattan`, `cmlm_qq`, `manifest`, `report`, `environment`

Resolve authoritative inputs from successful Manifests before execution. Reject hash, scope, sample-order, or model-lineage mismatches. Write standard logs, report, environment, Manifest v1, and the stage-local runner-only reproducibility bundle. Do not copy raw analysis code or raw underlying tool commands into `final/`.