Likelihood-based calibration improves the clinical utility of JAG1 functional data for variant classification

Area of Science:

• Genomic Medicine
• Molecular Biology
• Clinical Genomics

Background:

• Multiplexed assays of variant effects (MAVEs) generate large-scale functional variant data.
• Translating MAVE readouts into clinical genomics standards is crucial for utility.
• JAG1 variants are the primary cause of Alagille syndrome.

Purpose of the Study:

• To translate MAVE data for JAG1 variants into clinical variant classification evidence.
• To assess the impact of calibrated MAVE data on diagnosing Alagille syndrome.
• To evaluate the broader applicability of MAVE data calibration in clinical genomics.

Main Methods:

• Calculated log-likelihood ratios of pathogenicity for JAG1 variant scores from a previous MAVE.
• Translated MAVE scores into American College of Medical Genetics and Genomics (ACMG) and Association for Molecular Pathology (AMP) evidence weights.
• Retrospectively applied calibrated MAVE evidence to a cohort of Alagille syndrome patients with JAG1 variants of uncertain significance (VUS).

Main Results:

• Calibration improved separation between known benign and pathogenic variants.
• Increased classification rate of abnormal missense variants, providing clear evidence bins (strong, moderate, supporting).
• Reclassified 31% of VUS cases, upgrading 21% to likely pathogenic or pathogenic, improving diagnostic yield.

Conclusions:

• Calibrating MAVE data to ACMG/AMP standards enhances diagnostic yield for JAG1 variants in Alagille syndrome.
• This approach facilitates the clinical interpretation of MAVE data.
• The findings support the widespread adoption of calibrated MAVE models for variant classification across various genes.