An optimized variant prioritization process for rare disease diagnostics: recommendations for Exomiser and Genomiser.

Area of Science:

• Genomics
• Bioinformatics
• Rare disease diagnostics

Background:

• Exome sequencing (ES) and genome sequencing (GS) are crucial for identifying diagnostic variants in rare diseases.
• Variant prioritization remains a challenge, hindering efficient interpretation of genetic data.
• The Exomiser/Genomiser suite is widely used but lacks data-driven optimization guidelines.

Purpose of the Study:

• To provide optimized parameters and practical recommendations for Exomiser and Genomiser tools.
• To improve diagnostic variant prioritization in rare disease cases.
• To propose alternative workflows for complex cases where diagnostic variants may be missed.

Main Methods:

• Analysis of 386 diagnosed probands from the Undiagnosed Diseases Network (UDN).
• Systematic evaluation of parameters affecting tool performance: gene:phenotype data, pathogenicity predictors, phenotype terms, and family variant data.
• Assessment of coding and noncoding variants.

Main Results:

• Parameter optimization substantially improved Exomiser performance for GS (49.7% to 85.5%) and ES (67.3% to 88.2%) coding variants.
• Genomiser performance for noncoding variants improved from 15.0% to 40.0% in top 10 rankings.
• Refinement strategies, including p-value thresholds, were explored.

Conclusions:

• An evidence-based framework for variant prioritization using Exomiser and Genomiser in ES/GS data.
• Recommendations implemented in the Mosaic platform to enhance diagnostic yield for undiagnosed participants.
• Highlights the need for tracking solved cases to benchmark bioinformatics tools.