Interprétation de variants (ACMG)

Classe les variants génétiques selon les directives de l'ACMG à l'aide d'un moteur basé sur des règles.

Spar Skills Guide Bot
Data & IAIntermédiaire
1026/07/2026
#acmg#genomics#variant-classification#precision-medicine#genetics

Recommandé pour

---name: variant-interpretation-acmg description: Classifies genetic variants according to ACMG (American College of Medical Genetics) guidelines. keywords:

  • acmg
  • genomics
  • variant-classification
  • precision-medicine
  • genetics measurable_outcome: Correctly classifies >95% of variants when provided with accurate ACMG evidence codes. license: MIT metadata: author: AI Group version: "1.0.0" compatibility:
  • system: Python 3.10+ allowed-tools:
  • run_shell_command
  • read_file ---"

Variant Interpretation (ACMG)

The Variant Interpretation Skill automates the classification of genetic variants (Pathogenic, Benign, VUS) using a rules-based engine derived from ACMG guidelines.

When to Use This Skill

  • When analyzing a VCF file for clinical reporting.
  • To determine the clinical significance of a specific mutation (e.g., BRCA1 c.123A>G).
  • To aggregate evidence (population freq, computational predictions) into a final verdict.

Core Capabilities

  1. Rule Scoring: Applies codes like PVS1 (Null variant), PM2 (Rare), PP3 (In silico).
  2. Classification: Combines scores to reach a verdict (Pathogenic, Likely Pathogenic, VUS, etc.).
  3. Explanation: Provides the logic/evidence used for the classification.

Workflow

  1. Input: Variant details (Gene, HGVS, Consequence) or Evidence codes directly.
  2. Process: Sums weights of applied ACMG criteria.
  3. Output: Final classification and score breakdown.

Example Usage

User: "Classify a variant with evidence PVS1 and PM2."

Agent Action:

python3 Skills/Genomics/Variant_Interpretation/acmg_classifier.py \
    --evidence "PVS1,PM2"
Skills similaires