Prioritizing Clinically Relevant Copy Number Variation from Genetic Interactions and Gene Function Data.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Accurate identification of disease-causing genetic variants is essential for patient diagnosis.
• Copy Number Variants (CNVs) are significant contributors to genetic diseases but challenging to interpret.
• Existing methods struggle to efficiently filter pathogenic variants from large datasets.

Purpose of the Study:

• To develop a computational method for prioritizing disease-relevant Copy Number Variants (CNVs).
• To integrate functional genomics and clinical phenotype data for variant interpretation.
• To identify causative genes within patient CNVs.

Main Methods:

• Developed a prioritization method combining functional context and clinical phenotype data.
• Evaluated various feature and gene weighting systems for CNV and gene classification.
• Applied and optimized the best methodologies on a dataset of 2,500 CNVs from 140 patients using Agilent CGH 180k Microarrays.

Main Results:

• Achieved a high F-score of 91.59% in predicting disease relevance of CNVs.
• Demonstrated strong performance with 87.08% precision and 97.00% recall.
• Successfully identified clinically harmful CNVs and likely causative genes across diverse phenotypes.

Conclusions:

• The developed method effectively prioritizes disease-causing CNVs by integrating diverse biological data.
• This approach enhances the diagnostic yield of CNV analysis in clinical settings.
• The methodology and dataset are publicly available to facilitate further research.