omics-dataset-retrievallisted

# Omics Dataset Retrieval Build a deduplicated, relevance-audited catalog of public omics datasets for a disease, phenotype, gene, or biological process. Query every reachable public repository, classify each hit by omics type, score its relevance, and emit a CSV catalog + Markdown summary (+ optional landscape figure). **Boundary:** catalog metadata only. This skill does not download raw data files and does not perform downstream analysis. Worked examples (Alzheimer's, BCL11A/erythroid, PDAC) are in `references/examples.md`. ## Inputs | Parameter | Type | Description | |-----------|------|-------------| | `disease_or_topic` | string | Disease, phenotype, gene, or process (e.g. `"sickle cell disease"`, `"BCL11A"`) | | `synonyms` | list[str] | Alternative names, abbreviations, gene symbols (e.g. `["SCD","SCA","HbSS"]`) | | `omics_types` | list[str] or `"all"` | Restrict to omics types, or `"all"` (default) | | `organism` | string | `"all"` (default) — human, mouse, and all others. Restrict to `"human"`/`"mouse"` only if explicitly requested | | `year_min` | int | Earliest publication year (default: no limit) | | `output_dir` | path | Where to write outputs (default: `/mnt/results/`) | ## Outputs | File | Description | |------|-------------| | `<disease>_omics_datasets_MASTER.csv` | Full catalog with all metadata + relevance labels | | `<disease>_omics_datasets_VALIDATED.csv` | Filtered to `CORE/DIRECT` + `ADJACENT` only | | `<disease>_omics_summary.md` | Counts by omics