When Two plus Four Does Not Equal Six: Combining Computational and Functional Evidence to Classify BRCA1 Key Domain...

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Accurate genetic variant classification is crucial for clinical genetic sequencing.
• Multiplexed assays of variant effect (MAVEs) and computational tools are promising for addressing variant classification uncertainty.
• The joint performance of MAVEs and computational tools on novel variants remains unestablished.

Purpose of the Study:

• To empirically validate the strength of evidence for genetic variant classification criteria.
• To assess the performance of MAVEs and computational tools, individually and combined, for BRCA1 variants.
• To investigate the utility of exceptionally conserved ancestral residues (ECARs) in predicting variant pathogenicity.

Main Methods:

• Developed a maximum likelihood estimate (MLE) model to convert frequentist odds ratios to proportions pathogenic.
• Applied the MLE model to functional assay data and computational tool predictions for BRCA1 variants.
• Defined and analyzed exceptionally conserved ancestral residues (ECARs) in BRCA1.

Main Results:

• Missense substitutions at ECARs in BRCA1 are disproportionately pathogenic, with effect sizes comparable to protein-truncating variants.
• For non-ECAR positions, concordant predictions from assays and tools often do not meet the additive assumptions of American College of Medical Genetics and Genomics (ACMG) guidelines.
• The performance of assays and tools varied significantly at ECARs versus non-ECAR positions.

Conclusions:

• The strength of evidence assigned by ACMG guidelines is not universally applicable and requires empirical validation.
• MAVEs and computational tools show promise but their combined application needs refinement based on empirical data.
• ECARs represent a valuable feature for identifying potentially pathogenic variants in genes like BRCA1.