tooluniverse-pharmacogenomics

Generates evidence-graded pharmacogenomic dosing recommendations from CPIC, PharmGKB, and FDA biomarker data.

1.7k|254|Updated Mar 3, 2025
One-click install
npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-pharmacogenomics
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tooluniverse-pharmacogenomics
Source: https://github.com/mims-harvard/ToolUniverse/tree/main/plugins/tooluniverse/skills/tooluniverse-pharmacogenomics
Command: npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-pharmacogenomics

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Interpreting how a patient's genetic variants affect drug response requires querying multiple pharmacogenomics databases (CPIC, PharmGKB, FDA labels) and reconciling allele functions, metabolizer phenotypes, and guideline recommendations — a slow, error-prone manual process.

Core Features & Use Cases

  • CPIC Guideline Retrieval: Resolve gene-drug pairs by evidence level and fetch phenotype-specific dosing recommendations with strength classifications.
  • Allele & Variant Annotation: Look up star allele function status, activity scores, and PharmGKB clinical annotations for rsIDs.
  • FDA Biomarker Labeling: Check regulatory PGx status of drugs, including boxed warnings and contraindications by biomarker.
  • Use Case: Given a patient who is a CYP2D6 poor metabolizer, retrieve the CPIC recommendation for codeine (avoid use due to reduced morphine activation), confirm star allele function, and verify FDA label status — producing an evidence-graded clinical summary.

Quick Start

Ask the agent to get CPIC dosing recommendations for codeine in a CYP2D6 poor metabolizer and check the FDA pharmacogenomic biomarker label.

Frequently Asked Questions about tooluniverse-pharmacogenomics

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I get CPIC dosing recommendations for a drug?

Call CPIC_get_recommendations with the drug name (e.g., drug="codeine") to auto-resolve the guideline, or pass a numeric guideline_id. Filter the returned recommendations by the patient's metabolizer phenotype to get the applicable recommendation and its strength classification.

How do I interpret CYP2D6 star alleles and metabolizer status?

Use CPIC_get_alleles with the gene symbol to retrieve each allele's clinical functional status and activity value. Combine the diplotype's activity scores to derive the phenotype (poor, intermediate, normal, or ultrarapid metabolizer), then map it to the guideline recommendation.

What is the difference between CPIC and PharmGKB evidence levels?

CPIC levels A and B are actionable for prescribing, while C and D are informational. PharmGKB level 1A means the annotation is embedded in a CPIC or DPWG guideline; levels 3 and 4 are hypothesis-generating and should not alone drive prescribing decisions.

Why does CPIC_get_recommendations return zero rows for warfarin?

Warfarin dosing uses an algorithm-based guideline rather than a lookup table, so the recommendations endpoint returns no rows. Direct users to the CPIC website or use PharmGKB_get_dosing_guidelines with the clinpgxid for warfarin guidance.

Does the FDA require pharmacogenomic testing for specific drugs?

Query fda_pharmacogenomic_biomarkers with limit=1000 to retrieve all FDA PGx label entries. The LabelingSection field indicates urgency: Boxed Warning or Contraindications implies testing may be required, while Clinical Pharmacology is typically informational.

What are the limitations of CPIC guideline coverage?

CPIC covers roughly 29 guidelines across about 130 genes, so many drug-gene pairs lack formal guidelines. For uncovered pairs, fall back to PharmGKB variant annotations, EpiGraphDB gene-drug associations, or DGIdb interaction data.