A Standardized DNA Variant Scoring System for Pathogenicity Assessments in Mendelian Disorders.

Area of Science:

• Genetics
• Bioinformatics
• Clinical Diagnostics

Background:

• Accurate classification of DNA variants is crucial for genetic testing and clinical decision-making.
• Existing methods for variant classification can be subjective and lack standardization.
• A robust system is needed to consistently categorize variants by their pathogenicity.

Purpose of the Study:

• To develop and validate a rules-based scoring system for classifying DNA variants.
• To establish a standardized quantitative framework for assessing variant pathogenicity.
• To evaluate the reliability and consistency of the scoring system in a clinical laboratory setting.

Main Methods:

• A rules-based scoring system was developed to categorize DNA variants into five classes: pathogenic, likely pathogenic, variant of uncertain significance (VUS), likely benign, and benign.
• Over 16,500 pathogenicity assessments of 11,894 variants across 338 genes were analyzed using prediction tools, population frequency, co-occurrence, segregation data, and functional studies.
• Scores were calculated by trained scientists using a quantitative framework with differential weighting of data types, and interobserver concordance was tested.

Main Results:

• The system analyzed over 16,500 variant pathogenicity assessments from 11,894 variants in 338 genes.
• Private variants were predominantly VUS (80.5%) or likely pathogenic (19.5%).
• Interobserver agreement was high, with 98.5% exact agreement and 97% interobserver consistency, indicating a reliable and reproducible scoring system.

Conclusions:

• The developed scoring system provides reliable pathogenicity scores for DNA variants in a clinical setting.
• The quantitative framework and standardized approach enhance consistency and accuracy in variant classification.
• The system's stability and ability to incorporate new data ensure its ongoing utility for genetic testing.