AUTHOR=Alimohamed Mohamed Z. , Westers Helga , Vos Yvonne J. , Van der Velde K. Joeri , Sijmons Rolf H. , Van der Zwaag Paul A. , Sikkema-Raddatz Birgit , Jongbloed Jan D. H.
TITLE=Validation of New Gene Variant Classification Methods: a Field-Test in Diagnostic Cardiogenetics
JOURNAL=Frontiers in Genetics
VOLUME=13
YEAR=2022
URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2022.824510
DOI=10.3389/fgene.2022.824510
ISSN=1664-8021
ABSTRACT=Background: In the molecular genetic diagnostics of Mendelian disorders, solutions are needed for the major challenge of dealing with the large number of variants of uncertain significance (VUSs) identified using next-generation sequencing (NGS). Recently, promising approaches using constraint metrics to calculate case excess scores (CE), etiological fractions (EF), and gnomAD-derived constraint scores have been reported that estimate the likelihood of rare variants in specific genes or regions that are pathogenic. Our objective is to study the usability of these constraint data into variant interpretation in a diagnostic setting, using our cardiomyopathy cohort.
Methods and Results: Patients (N = 2002) referred for clinical genetic diagnostics underwent NGS testing of 55–61 genes associated with cardiomyopathies. Previously classified likely pathogenic (LP) and pathogenic (P) variants were used to validate the use of data from CE, EF, and gnomAD constraint analyses for (re)classification of associated variant types in specific cardiomyopathy subtype-related genes. The classifications corroborated in 94% (354/378) of cases. Next, we reclassified 23 unique VUSs to LP, increasing the diagnostic yield by 1.2%. In addition, 106 unique VUSs (5.3% of patients) were prioritized for co-segregation or functional analyses.
Conclusions: Our analysis confirms that the use of constraint metrics data can improve variant interpretation, and we, therefore, recommend using constraint scores on other cohorts and disorders and its inclusion in variant interpretation protocols.