Virtual Gene Panels Have a Superior Diagnostic Yield for Inherited Rare Diseases Relative to Static Panels.

Abstract

Exome- or genome-based panels-also known as slices or virtual panels-are now a popular approach that involves comprehensive genomic sequencing while restricting analysis to subsets of genes based on patients' phenotypes. This flexible strategy enables frequent gene updates based on novel disease associations as well as reflexing to analyzing other genes up to the whole exome or genome. With recent improvements addressing limitations associated with virtual panels, the advantages of this approach, relative to static custom-based panels, remain to be systematically characterized. Here we perform slice testing on 1014 patients (50.5% females; average age 17 years) referred from multiple pediatric clinics within a single center in the Middle East (83% Arab population). Initial analysis uncovered molecular diagnoses for 235 patients for a diagnostic yield of 23% (235/1014). "On the fly" focused analysis in most negative cases (N = 779) identified clinically significant variants correlating with patients' presentations in genes outside the originally ordered panel for another 35 patients (3.5% or 35/1024) increasing the overall diagnostic yield to 27%. The pathogenic variants underlying the additional cases (13% of all positive cases) were excluded from the original "panel" gene list, mainly as result of issues related to panel selection, novel gene-disease associations, phenotype spectrum broadening, or gene lists variability. The additional findings led to changes in clinical management in most patients (94%). Our findings support slice testing as an efficient and flexible platform that facilitates updates to gene lists to achieve high clinical sensitivity and utility.