DeMAG.webservice

DeMAG.webservice predicts the pathogenicity of missense variants in actionable disease genes to improve variant interpretation for clinical and research genomics.

Key Features:

• Supervised Classification: Trained on comprehensive diagnostic datasets from 59 actionable disease genes as outlined in the American College of Medical Genetics and Genomics Secondary Findings v2.0 (ACMG SF v2.0) to distinguish benign and pathogenic variants.
• Enhanced Performance: Demonstrates balanced specificity (82%) and sensitivity (94%) compared with existing Variant Effect Predictors (VEPs).
• Novel Epistatic Feature - 'Partners Score': Incorporates a partners score that models evolutionary and structural partnerships between residues and integrates clinical and functional information.
• Broad Application: Extends predictions to all missense variants in 316 clinically actionable disease genes.

Scientific Applications:

• Clinical variant interpretation: Reduces variants of uncertain significance (VUSs) and supports diagnostic decision-making in precision medicine.
• Gene-disease research: Integrates evolutionary and structural data to enable investigation of gene-disease associations and mechanisms of genetic disorders.

Methodology:

Implements a supervised Variant Effect Predictor trained on diagnostic datasets from 59 ACMG SF v2.0 genes and uses a partners score that models evolutionary and structural residue partnerships while integrating clinical and functional information to classify missense variants, with predictions extended to 316 clinically actionable genes.