AI agent skill · by @ClawBio

Rnaseq De

Differential expression analysis for bulk RNA-seq and pseudo-bulk count matrices with QC, PCA, and contrast testing.

Archived 2026-10-23 · Source updated 2026-10-04 · License: MIT. Check the original repository for the latest version.

SKILL.md instructions

# 🧬 RNA-seq Differential Expression

This skill performs differential expression on bulk RNA-seq or pseudo-bulk count matrices.

## Core Capabilities

1. Input validation for count matrix and sample metadata
2. Pre-DE QC (library size, detected genes, low-count filtering)
3. PCA visualisation on normalized expression
4. Differential expression from formula + contrast
5. Volcano and MA plots
6. Markdown report with reproducibility files

## Input Contract

- Count matrix (`.csv` or `.tsv`): rows are genes, columns are samples, first column is gene identifier
- Metadata table (`.csv` or `.tsv`): one row per sample, must include `sample_id`
- Formula: e.g. `~ condition` or `~ batch + condition`
- Contrast: `factor,numerator,denominator` (e.g. `condition,treated,control`)

## Output Structure

```
rnaseq_de_report/
├── report.md
├── figures/
│   ├── pca.png
│   ├── volcano.png
│   └── ma_plot.png
├── tables/
│   ├── qc_summary.csv
│   ├── normalized_counts.csv
│   └── de_results.csv
└── reproducibility/
    ├── commands.sh
    ├── environment.yml
    └── checksums.sha256
```

## Usage

```bash
python rnaseq_de.py \
  --counts counts.csv \
  --metadata metadata.csv \
  --formula "~ batch + condition" \
  --contrast "condition,treated,control" \
  --output report_dir
```

## Safety

- Local-only processing
- Warn before overwriting existing output
- Report-level disclaimer required