RNA-sequencing improves diagnosis for neurodevelopmental disorders by identifying pathogenic non-coding variants and reinterpretation of coding variants

BackgroundFor neurodevelopmental disorders (NDD), a molecular diagnosis is key for predicting outcome, treatment and genetic counseling. Currently, in about half of NDD cases, routine DNA-based testing fails to establish a genetic diagnosis. Transcriptome analysis (RNA-seq) improves the diagnostic yield for some groups of diseases, but has not been applied to NDD in a routine diagnostic setting. MethodsHere, we explored the diagnostic potential of RNA-seq in a cohort of 96 individuals including 67 undiagnosed NDD subjects. We created a user-friendly web-application to analyze RNA-seq data from single individuals cultured skin fibroblasts for genic, exonic and intronic expression outliers, based on modified OUTRIDER Z-scores. Candidate pathogenic events were complemented/matched with genomic data and, if required, confirmed with additional functional assays. ResultsWe identified pathogenic small genomic deletions, mono-allelic expression, aberrant splicing events, deep intronic variants resulting in pseudo-exon insertion, but also synonymous and nonsynonymous variants with deleterious effects on transcription. This approach increased the diagnostic yield for NDD by 12%. Diagnostic pitfalls during transcriptome analysis include detection of splice abnormalities in putative disease genes caused by benign polymorphisms and/or absence of expression of the responsible gene in the tissue of choice. This was misleading in one case and could have led to the wrong diagnosis in the absence of appropriate phenotyping. ConclusionsNonetheless, our results demonstrate the utility of RNA-seq in molecular diagnostics and stress the importance of multidisciplinary team consultation. In particular, the approach is useful for the identification and interpretation of unexpected pathogenic changes in mRNA processing and expression in NDD.