Using variant databases for variant prioritization and to detect erroneous genotype-phenotype associations.

Area of Science:

• Genetics
• Bioinformatics
• Genomic Medicine

Background:

• Variant databases are crucial for identifying causal mutations by reducing potential variants.
• Current filtering methods include the absence-approach and the 1%-approach, each with limitations.

Purpose of the Study:

• To investigate the false negative and false positive rates of existing variant filtering approaches.
• To introduce and evaluate a novel quantile-based approach for variant filtering.

Main Methods:

• Comparison of the absence-approach, 1%-approach, and the novel quantile-based approach using real-life data.
• The quantile-based approach utilizes variable frequency thresholds based on disease prevalence, inheritance models, and database characteristics.

Main Results:

• The quantile-based approach demonstrated superior performance in reducing false negatives compared to the absence-approach.
• It offers better control over false positives by appropriately handling variable allele frequencies.
• The approach also proved effective in flagging variants that deviate from theoretical expectancies, prompting reevaluation of genotype-phenotype associations.

Conclusions:

• The quantile-based approach is a highly suitable method for variant filtering and flagging.
• This novel method enhances the accuracy of identifying disease-causing variants.
• User-friendly resources, including lookup tables and R calculators, are provided.