Transcriptome-based variant calling and aberrant mRNA discovery enhance diagnostic efficiency for neuromuscular diseases

BACKGROUND: Whole-exome sequencing-based diagnosis of rare diseases typically yields 40%–50% of success rate. Precise diagnosis of the patients with neuromuscular disorders (NMDs) has been hampered by locus heterogeneity or phenotypic heterogeneity. We evaluated the utility of transcriptome sequencing as an independent approach in diagnosing NMDs. METHODS: The RNA sequencing (RNA-Seq) of muscle tissues from 117 Korean patients with suspected Mendelian NMD was performed to evaluate the ability to detect pathogenic variants. Aberrant splicing and CNVs were inspected to identify additional causal genetic factors for NMD. Aberrant splicing events in Dystrophin (DMD) were investigated by using antisense oligonucleotides (ASOs). A non-negative matrix factorisation analysis of the transcriptome data followed by cell type deconvolution was performed to cluster samples by expression-based signatures and identify cluster-specific gene ontologies. RESULTS: Our pipeline called 38.1% of pathogenic variants exclusively from the muscle transcriptomes, demonstrating a higher diagnostic rate than that achieved via exome analysis (34.9%). The discovery of variants causing aberrant splicing allowed the application of ASOs to the patient-derived cells, providing a therapeutic approach tailored to individual patients. RNA-Seq data further enabled sample clustering by distinct gene expression profiles that corresponded to clinical parameters, conferring additional advantages over exome sequencing. CONCLUSION: The RNA-Seq-based diagnosis of NMDs achieves an increased diagnostic rate and provided pathogenic status information, which is not easily accessible through exome analysis.