Tooluniverse Variant Analysis

This handles the full VCF analysis stack: parsing variant calls, annotating consequence types (synonymous, missense, frameshift), filtering by VAF, and splitting coding from non-coding variants. Ships with three battle-tested scripts that encode the gotchas you'd otherwise spend hours debugging. The GATK HaplotypeCaller pipeline script runs BWA alignment through variant calling or counts existing VCFs with proper multi-allelic split handling. The coding variant filter knows about two-row VarSeq Excel headers and correctly excludes intronic, intergenic, and UTR junk. The fraction calculator uses coding variants as the denominator, not all variants, which matters when clinical papers report mutation burdens. One heads-up: it defaults to diploid calling because that's what most published HaplotypeCaller runs used, even on prokaryotes.

skills.sh installs: 373
GitHub stars: 1.8K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add mims-harvard/tooluniverse --skill tooluniverse-variant-analysis --agent claude-code

Installs into .claude/skills of the current project.

Files

Variant Analysis and Annotation

RULE ZERO — Check for pre-computed results FIRST

Before following any instruction below, scan the data folder for:

*_executed.ipynb → read with tu run read_executed_notebook '{"data_folder":"<path>","search":"<keyword>"}' and cite its cell outputs as the authoritative answer
Pre-computed result files (CSV/TSV with names like *results*, *deseq*, *enrich*, *stats*, *_simplified.csv) → read directly and report the requested value
Canonical analysis scripts (analysis.R, run_*.py, find_*.R, *.Rmd) → execute as-is and read the output

Only follow this skill's re-analysis recipe below if none of the above exist. Re-running from raw data produces different numbers than the published answer and is much slower (often 5-10× turn count).

PRIMARY SCRIPTS — use these FIRST

These bundled scripts encode the question-specific gotchas (denominator choices, ploidy defaults, multi-allelic split, multi-row Excel headers, non-coding allowlist). They emit labelled KEY=VALUE lines that are easier to parse than ad-hoc pandas/awk output. Prefer them over writing new code.

SKILL.md on GitHub