De Summary

Do NOT fire when:

• User wants to run DE analysis from raw counts (use rnaseq-de)
• User wants pathway enrichment analysis (out of scope)
• User wants to re-analyse with different parameters

Scope

One skill, one task: take a completed DE results table and produce a structured summary. Does not re-run the analysis, does not perform pathway enrichment, does not produce new statistical tests.

Workflow

1. Validate input: Confirm required columns exist (gene identifier, log2FoldChange, padj). Detect column naming variants (adj.P.Val for limma, FDR for edgeR).
2. Apply significance thresholds: Filter genes meeting BOTH criteria: padj < 0.05 AND |log2FoldChange| >= 1.0. Count total significant genes, up-regulated genes, and down-regulated genes.
3. Rank and select top 10: Sort significant genes by padj (ascending). Break ties by |log2FoldChange| (descending). Select top 10 for the summary table.
4. Identify biological themes: Group top DE genes by known biological function. Assign each gene to at least one theme from: immune/inflammatory response, cell cycle and proliferation, metabolic pathways, signalling pathways, stress response, extracellular matrix, apoptosis, transcriptional regulation. Use gene symbol knowledge; do not run external enrichment tools.
5. Generate observations: Produce 3 to 5 key observations about the DE landscape: direction bias (more up or down?), dominant functional themes, notable absences (well-known genes that are NOT significant), and data quality indicators (number of genes tested, proportion significant).
6. Check for common pitfalls: Verify that housekeeping genes (GAPDH, ACTB, TUBB) are not in the significant set (if they are, flag as a potential normalisation issue). Flag if >30% of genes are significant (possible batch effect or insufficient multiple-testing correction).
7. Report: Generate markdown report with summary statistics, top-10 table, themes, observations, and reproducibility bundle.

Example Output

{
  "summary_statistics": {
    "total_genes_tested": 50,
    "significant_genes": 28,
    "up_regulated": 18,
    "down_regulated": 10,
    "thresholds": {"padj": 0.05, "log2fc_min": 1.0}
  },
  "top_10_genes": [
    {"rank": 1, "gene": "IL6", "log2FC": 3.82, "padj": 1.1e-31, "direction": "up"},
    {"rank": 2, "gene": "CXCL10", "log2FC": 3.45, "padj": 1.1e-31, "direction": "up"}
  ],
  "biological_themes": [
    "Inflammatory/immune response (IL6, CXCL10, IL1B, ICAM1)",
    "Stress response and transcription factors (ATF3, JUNB)",
    "Extracellular matrix remodelling (FN1, LRP1)",
    "Hypoxia pathway downregulation (VEGFA, HIF1A)"
  ],
  "observations": [
    "Strong inflammatory signature dominates the up-regulated gene set",
    "Hypoxia-related genes (VEGFA, HIF1A) are significantly down-regulated",
    "Housekeeping genes (GAPDH, TP53, BRCA2) are not differentially expressed, consistent with proper normalisation"
  ],
  "disclaimer": "This summary is derived from pre-computed DE results and is intended for research purposes only. Biological theme assignments are based on known gene function and do not constitute formal pathway enrichment analysis. Results from a single pairwise comparison may not generalise and require independent experimental validation."
}

Gotchas

1. The model will want to re-run the DE analysis. Do not. Accept the input table as authoritative. Your job is to summarise, not to second-guess the statistical method.
2. The model will want to run pathway enrichment (GO, KEGG). Do not. Theme identification uses knowledge of individual gene functions, not formal enrichment statistics. If the user wants enrichment, recommend a dedicated tool.
3. The model will want to include non-significant genes in the top-10. Do not. Apply both the padj and log2FC thresholds strictly. Genes failing either criterion must not appear in the ranked list.
4. The model will confuse low padj with high significance. Remember: lower padj = more significant. Sort ascending.
5. The model will ignore direction. Always report whether each gene is up-regulated or down-regulated. A summary that omits direction is incomplete.

Safety

• This skill produces research-level summaries, not clinical reports.
• Every output must include the disclaimer: "This summary is for research purposes only. Results require independent experimental validation."
• Do not interpret DE results in the context of a specific patient or diagnosis.
• Do not claim that DE results establish causation.
• Include the ClawBio medical disclaimer.

Agent Boundary

• Agent dispatches and explains; skill executes.
• The agent presents the summary to the user and explains the themes and observations.
• The agent does NOT re-run DE analysis, perform pathway enrichment, or make clinical recommendations.

Chaining Partners

• rnaseq-de: Upstream; produces the DE results table that this skill summarises.
• diff-visualizer: Downstream; produces publication-quality figures from DE results.
• lit-synthesizer: Downstream; literature context for top DE genes.
• pubmed-summariser: Downstream; PubMed search for genes of interest.

Maintenance

• Review cadence: quarterly (gene function annotations evolve slowly).
• Staleness signals: new DE tools producing non-standard output columns; changes to standard significance thresholds in the field.
• Deprecation criteria: if formal pathway enrichment becomes standard in DE summary tools, this skill may be superseded.