Tooluniverse Rnaseq Deseq2

You've got count matrices and metadata, you need differential expression with DESeq2, and you want the answer to match what's actually in the published notebook. This skill provides audited wrapper scripts that emit every common interpretation at once: multiple filter combinations (strict padj+LFC+baseMean, padj+LFC-only, padj-only), shrunk and unshrunk fold changes, multi-contrast Venn regions with different denominators, and correlation/PCA variants across transforms. The core insight is that questions phrased as "using DESeq2, how many genes..." are usually asking you to read the executed notebook's output, not rerun the analysis. When you do need to rerun, the wrappers handle sample exclusions, covariate designs, and LFC shrinkage methods in one deterministic call so you're not debugging subtle filter mismatches between your reimplementation and the original.

skills.sh installs: 409
GitHub stars: 1.8K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add mims-harvard/tooluniverse --skill tooluniverse-rnaseq-deseq2 --agent claude-code

Installs into .claude/skills of the current project.

Files

RNA-seq Differential Expression Analysis (DESeq2)

PRIMARY SCRIPTS — use these FIRST before writing custom code

The four scripts below are deterministic, audited wrappers that handle the ambiguity in DESeq2 / correlation / PCA / ANOVA questions by emitting EVERY common interpretation in one call. Reading their output and matching the variant the published notebook used is more reliable than re-deriving the answer from scratch.

All four scripts honor workspace isolation: they ONLY write to --workdir (or /tmp/... by default). They never touch the input data folder. Always pass --workdir /tmp/<run-name> when you need intermediate files.