skills.rest
mims-harvard
mims-harvardOFFICIAL
Explore all skills by this author →
Website𝕏@marinkazitnikGitHub

tooluniverse-variant-interpretation

Classify genetic variants using ACMG guidelines with population, structural, and literature evidence.

1.7k|254|Updated Mar 3, 2025
Search Tags:#variant interpretation#clinvar#clinical genomics#gnomad#pathogenicity#ACMG classification#vus
One-click install
npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-variant-interpretation
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tooluniverse-variant-interpretation
Source: https://github.com/mims-harvard/ToolUniverse/tree/main/plugins/tooluniverse/skills/tooluniverse-variant-interpretation
Command: npx skills add https://github.com/mims-harvard/ToolUniverse --skill tooluniverse-variant-interpretation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Interpreting the clinical significance of genetic variants requires querying dozens of databases, applying complex ACMG/AMP evidence rules, and synthesizing findings into defensible reports. This Skill automates that workflow, turning raw variant calls into ACMG-classified interpretations with cited evidence and clinical recommendations.

Core Features & Use Cases

  • ACMG Classification: Applies evidence codes (PVS1, PM2, PP3, BA1, etc.) with a Bayesian point system to produce Pathogenic through Benign classifications.
  • Multi-Source Evidence Gathering: Queries ClinVar, gnomAD, OMIM, ClinGen, COSMIC, SpliceAI, CADD, AlphaMissense, and EVE for population frequencies, clinical assertions, and pathogenicity predictions.
  • Structural & Non-Coding Analysis: Uses AlphaFold/PDB structures for missense impact and deep-learning models (AlphaGenome, Enformer, Evo 2) for regulatory variant effects.
  • Use Case: A clinician asks about TP53 p.R175C reported as a VUS. The Skill checks gnomAD (absent), finds pathogenic variants at the same residue (PM5), confirms concordant damaging predictions (PP3), analyzes the DNA-binding domain structure, and reclassifies it as Likely Pathogenic with surveillance recommendations.

Quick Start

Interpret the clinical significance of BRCA1 c.5266dupC and produce an ACMG-classified report with clinical recommendations.

Frequently Asked Questions about tooluniverse-variant-interpretation

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

FAQPage Schema
How do I classify a variant using ACMG guidelines?▼

Provide the variant in HGVS notation (e.g., NM_007294.4:c.5266dupC) and the Skill gathers evidence from ClinVar, gnomAD, and predictors, then applies ACMG evidence codes using a Bayesian point system. Scores of 10+ points yield Pathogenic, 6-9 Likely Pathogenic, and negative ranges yield Benign classifications.

How to interpret a VUS missense variant with structural analysis?▼

The Skill retrieves PDB experimental structures or AlphaFold predictions for the protein, maps the residue position, and assesses domain context, solvent accessibility, and functional site proximity. This structural evidence supports PM1 or PP3 codes that can reclassify a VUS.

Can this predict effects of non-coding or intronic variants?▼

Yes, it uses SpliceAI for splice-site variants and sequence deep-learning models (AlphaGenome, Enformer, Borzoi, ChromBPNet, Evo 2) for promoter, enhancer, and UTR variants. These predict delta effect sizes on expression and chromatin tracks, providing mechanistic evidence where SIFT and PolyPhen do not apply.

What population frequency thresholds apply for gnomAD data?▼

Allele frequency above 5% triggers BA1 (stand-alone benign), above 1% supports BS1, and absence from controls supports PM2. BS1 thresholds are gene-specific, calibrated by disease prevalence and penetrance rather than a universal cutoff.

What are the limitations of computational pathogenicity predictors?▼

Predictors like CADD, AlphaMissense, and EVE provide supporting evidence only (PP3/BP4) and cannot alone classify a variant. Discordant predictions, missing REVEL scores, or absent functional studies keep variants at VUS until stronger evidence emerges.

Does the Skill handle conflicting functional and population evidence?▼

Yes, it follows ClinGen SVI guidance that epidemiological data generally outweighs in-vitro assays for clinical classification. Conflicts are documented explicitly in the report, and the Bayesian point system nets opposing codes naturally.

Related Skills

Explore More Education & Research Skills→
maziyarpanahimaziyarpanahi
5.0k

annotating-variants

Annotate and normalize genomic variants from VCF files, rsIDs, and HGVS descriptions with consequences and population frequencies.

Community
Advanced
mims-harvardmims-harvard
1.6k

Clinical Variant Interpreter

Integrate population, ClinVar, ClinGen, OMIM, and literature data to classify genetic variants with ACMG criteria.

Official
Advanced
mims-harvardmims-harvard
1.6k

tooluniverse-rare-disease-diagnosis

Match patient phenotypes and genetics to rare diseases with HPO terms.

Official
Advanced
amanhsnamanhsn

fda-consultant-specialist

Guide FDA regulatory pathways and submission requirements for medical devices.

Community
Advanced