Predict pathogenicity of human missense SNVs with calibrated probabilities and ACMG classifications.
npx skills add https://github.com/zja2004/SNV-judge --skill snv-judge
Please help me install this Agent Skill. Skill: snv-judge Source: https://github.com/zja2004/SNV-judge/tree/main/skill Command: npx skills add https://github.com/zja2004/SNV-judge --skill snv-judge
SNV-judge provides calibrated, ACMG-aware pathogenicity predictions for human missense single-nucleotide variants (SNVs) by integrating 8 heterogeneous features from free public sources, enabling clinicians and researchers to interpret variants with quantitative probabilities and actionable classifications.
Provide a chrom/pos/ref/alt or protein change and run predict_variant to obtain a calibrated pathogenicity probability, ACMG classification, and SHAP explanations.
High-intent search queries and answers about installing and using this skill.
You can predict missense SNV pathogenicity by processing single variants or VCF batches to receive calibrated probability scores and 5-tier ACMG classifications (P/LP/VUS/LB/B) without needing API keys.
Yes, clinical genomics variant interpretation includes per-feature SHAP contributions, providing visual explanations that detail how eight public-source features influence the calibrated pathogenicity probability.
No, basic variant queries require no API keys, as the tool uses eight free public-source features to deliver calibrated pathogenicity probabilities and ACMG classifications for missense SNVs.
The best way to generate a clinical interpretation report is by using optional LLM templates, which produce Chinese, English, or summary reports based on calibrated pathogenicity predictions and ACMG classifications.
You can query missense SNV pathogenicity by providing a protein change instead of genomic coordinates, allowing the model to predict calibrated pathogenicity probabilities and ACMG classifications.
SNV pathogenicity prediction and model explainability rely on xgboost, lightgbm, scikit-learn, and shap to process eight public-source features and generate calibrated probability scores with visual explanations.
MubasherMohammedAnnotate genomic variants in VCF files using Ensembl VEP, ClinVar, and gnomAD APIs.
nahisahoAutomate ACMG-aligned clinical interpretation of genetic variants and generate standardized reports.
xjtulycBuild scikit-learn pipelines with nested cross-validation and FLAML AutoML.
weisbergBuild predictive lead scoring models from CRM data with SHAP explanations.
