Drug Target Validation (ToolUniverse Claude Skill)

A ToolUniverse agent skill that validates a drug-target hypothesis with multi-dimensional computational evidence and returns a quantitative Target Validation Score, priority tier, and GO/NO-GO recommendation before any experimental commitment.

   
Type Claude Skill (one of ToolUniverse’s pre-built agent skills)
Supplier Zitnik Lab, Harvard Medical School
Availability GA — part of the ToolUniverse skills collection (skills/tooluniverse-drug-target-validation/)
Pricing Free / OSS (Apache-2.0); wraps public APIs (Open Targets, ChEMBL, UniProt, Ensembl, HGNC, expression atlases)
Capabilities Read-only — drives ToolUniverse tool calls; no data writes
Verified works · 2026-07-20
Security cleared · 2026-07-20 — provenance matches Zitnik Lab, Apache-2.0, skill dir confirmed, no OSV advisories

How to install

This skill calls ToolUniverse tools, so the ToolUniverse MCP server must be installed first (see the ToolUniverse page). Simplest registration:

claude mcp add --transport stdio tooluniverse -- uvx tooluniverse

Then add the skills:

  • Claude Code — install the whole skill collection (the skill resolves as tooluniverse-drug-target-validation): 
    npx skills add mims-harvard/ToolUniverse
  • Manual / other agents — copy just this skill directory into your skills folder: 
    git clone https://github.com/mims-harvard/ToolUniverse
cp -r ToolUniverse/skills/tooluniverse-drug-target-validation ~/.claude/skills/

The skill sets disable-model-invocation: true upstream, so invoke it explicitly (e.g. ask Claude to “use the drug-target-validation skill”) rather than relying on automatic dispatch.

What it does

Runs a four-gate sequential validation model — failure at an early gate makes later gates irrelevant, and negative results are documented rather than hidden:

  1. Disease association — genetic evidence (GWAS, rare/Mendelian variants), literature, and pathway support.
  2. Druggability — structure and binding-pocket assessment, available chemical matter, and target class.
  3. Safety in normal tissue — expression across critical organs, mouse-knockout lethality, known ADRs.
  4. Competitive landscape — approved drugs, late-stage trials, and the differentiation bar.

It produces a 0–100 score (Disease Association 30 + Druggability 25 + Safety 20 + Clinical Precedent 15 + Validation Evidence 10) mapped to priority tiers: 80–100 Tier 1 (GO), 60–79 Tier 2 (conditional GO), 40–59 Tier 3 (caution), 0–39 Tier 4 (NO-GO). It resolves all identifiers (Ensembl, UniProt, ChEMBL, HGNC) first and computes statistics in Python rather than narrating hypothetical steps.

Primary use cases: target prioritisation before experimental work, GO/NO-GO triage of candidate targets, portfolio ranking by genetic and druggability evidence.

Notes

It is a reasoning layer over ToolUniverse; without the MCP server registered, the tool calls fail. The genetic-evidence and competitive-landscape gates lean on Open Targets OpenTargets_* tools — if the ToolUniverse Open Targets surface is degraded (see the Open Targets flag), scores in those dimensions may be incomplete. ToolUniverse ships ~68 such skills; the research, repurposing, and synergy workflows are catalogued separately.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.