tooluniverse-phewas

Queries cross-ancestry PheWAS associations for a variant or gene across four biobanks.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Determining whether a genetic variant or gene association replicates across populations requires manually querying multiple biobank portals with different coordinate systems, phenotype codings, and evidence types. This Skill orchestrates phenome-wide association lookups for one rsID or gene across UKB-TOPMed, FinnGen, BioBank Japan, TPMI, and Genebass, then judges replication versus ancestry-specificity.

Core Features & Use Cases

  • Cross-biobank variant PheWAS: Queries UKB-TOPMed, BioBank Japan, TPMI, and FinnGen for every phenotype associated with a single rsID, handling GRCh37/GRCh38 build differences and rsID-to-coordinate resolution.
  • Gene-burden layer: Runs Genebass exome-wide rare-variant burden PheWAS (e.g., pLoF for PCSK9) to detect phenotypes driven by rare coding variation rather than common variants.
  • Replication interpretation: Aligns phenotypes across differing coding systems (phecodes, ICD endpoints, FinnGen endpoints), compares effect directions and allele frequencies, and flags power-related false absences.
  • Use Case: Given rs7903146, retrieve its T2D association across European, Japanese, and Taiwanese cohorts, confirm concordant effect direction, and explain the weaker East-Asian p-value via lower allele frequency.

Quick Start

Ask the agent to run a cross-ancestry PheWAS on rs7903146 and report which associations replicate across biobanks.

Frequently Asked Questions about tooluniverse-phewas

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

FAQPage Schema
How do I run a PheWAS for a variant across multiple biobanks?

Pass the rsID directly to the UKB-TOPMed, BioBank Japan, and TPMI phewas_by_variant tools, which resolve coordinates internally. For FinnGen, first resolve the rsID to a GRCh38 chr:pos:ref:alt string from the UKB-TOPMed result, then call FinnGen_get_variant_finemapping.

What is the difference between PheWAS and GWAS?

A GWAS fixes a phenotype and scans variants, while a PheWAS fixes one variant or gene and scans the entire phenome. Use this Skill for the PheWAS direction; for trait-to-SNP queries use the gwas-* skills instead.

Can I use an rsID directly with FinnGen fine-mapping?

No. FinnGen_get_variant_finemapping requires an explicit GRCh38 variant string in chr:pos:ref:alt format. Resolve the rsID first, typically by reading the variant field returned by the UKB-TOPMed call.

Why does an association appear in Europeans but not East Asians?

Check allele frequency and case counts before concluding biology. A variant common in Europeans but rare in East Asians may simply lack power there; the results include af, num_cases, and num_controls for this assessment.

What does a p-value of 0.0 mean in PheWAS results?

A returned pval of exactly 0.0 is floating-point underflow on an extremely significant association, not a null result. Treat it as the strongest possible hit rather than absence of association.

What are the limitations of cross-biobank PheWAS?

Results are limited to each biobank's pre-computed summary statistics and phenotype catalogues, Genebass covers only European exomes, and FinnGen reports credible-set membership rather than p-values. No formal cross-ancestry meta-analysis is performed.