bioresearch-biomarker-discoverylisted

# BioResearch Agent — Biomarker Discovery Skill ## Capability End-to-end biomarker discovery: dataset loading (GEO GSE7621 or synthetic) → differential expression (t-test + Bonferroni) → hypergeometric pathway enrichment (KEGG/GO) → candidate ranking → volcano plot. Returns ranked candidate biomarkers, not just a narrative. ## Run ```bash bioresearch run biomarker --disease "Parkinson's disease" ``` ## Outputs (in `outputs/biomarker/`) - `biomarker_deg_table.csv` — differentially expressed genes - `biomarker_top_candidates.csv` — ranked candidates - `biomarker_pathway_enrichment.csv` — enriched pathways / GO terms - `biomarker_volcano_plot.png` — volcano plot - `biomarker_report.txt` — structured report ## Note This skill dispatches to the framework's `biomarker` workflow. It adds **no analysis of its own**; all statistics run in the workflow modules. Requires no LLM key.