HerediVar

HerediVar streamlines the classification of genetic variants in hereditary breast and ovarian cancer through modular annotation services and collaborative classification workflows.

Key Features:

• Modular Annotation Services: Provides flexible, modular annotation capabilities for genetic variant data.
• Collaborative Classification: Enables multiple experts to collaborate on variant classification.
• Automated Workflows: Supports automated processing through integrated workflows.
• HerediClassify Integration: Incorporates a variant classification algorithm that implements 19 of 28 American College of Medical Genetics variant classification criteria and gene-specific recommendations for ATM, BRCA1, BRCA2, CDH1, PALB2, PTEN, and TP53.
• Performance Validation: Achieves an average F1-Score of 93% across all criteria on validation datasets, outperforming vaRHC and Cancer SIGVAR.
• Extensible Design: Supports extension to other use cases and human genetic diagnostics through modular architecture.

Scientific Applications:

• Hereditary Breast and Ovarian Cancer Variant Classification: Supports expert classification of genetic variants in HBOC by integrating multiple evidence types and gene-specific guidelines.
• Gene-Specific Variant Analysis: Applies tailored classification recommendations for ATM, BRCA1, BRCA2, CDH1, PALB2, PTEN, and TP53.
• Automated Variant Classification Benchmarking: Enables comparison and validation of automated classification performance against other tools.

Methodology:

HerediClassify implements 19 of 28 American College of Medical Genetics variant classification criteria together with gene-specific recommendations for ATM, BRCA1, BRCA2, CDH1, PALB2, PTEN, and TP53; performance was validated on a dataset yielding an average F1-Score of 93% across all criteria and compared to vaRHC and Cancer SIGVAR.