TADA-a machine learning tool for functional annotation-based prioritisation of pathogenic CNVs

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Limited methods exist for prioritizing copy number variants (CNVs) based on predicted pathogenicity.
• Understanding the disease impact of genomic alterations is crucial for diagnostics.

Purpose of the Study:

• To introduce TADA, a new method for prioritizing pathogenic CNVs.
• To develop accurate classifiers for predicting CNV pathogenicity using functional annotations.

Main Methods:

• Developed TADA, a method combining assisted manual filtering and automated classification.
• Utilized an extensive catalogue of functional annotations and rigorous enrichment analysis.
• Trained and validated predictive classifiers for CNV pathogenicity.

Main Results:

• TADA accurately predicts pathogenic CNVs, outperforming existing methods.
• The method generates a well-calibrated pathogenicity score.
• Functional annotation-based prioritization proved effective.

Conclusions:

• TADA offers a promising approach for prioritizing pathogenic CNVs.
• This method can support clinical diagnostics and advance understanding of disease mechanisms.
• Functional annotation is a valuable strategy for analyzing large genomic alterations.