RNA sequencing uplifts diagnostic rate in undiagnosed rare disease patients

BackgroundRNA-sequencing is increasingly being used as a complementary tool to DNA sequencing in diagnostics where DNA analysis has been uninformative. RNA-sequencing allows us to identify alternative splicing and aberrant gene expression allowing for improved interpretation of variants of unknown significance (VUS). Additionally, RNA-sequencing provides the opportunity not only to look at the splicing effects of known VUSs but also to scan the transcriptome for abnormal splicing events and expression abnormalities in other relevant genes that may be the cause of a patients phenotype. MethodsUsing RNA from patient blood, we have systematically assessed transcriptomic profiles of 87 patients with suspected Mendelian disorders, 38% of which did not have a candidate sequence variant. Cases with VUSs and known events were assessed first followed by assessment of cases with no VUS. Each VUS was visually inspected using the Integrative Genomics Viewer (IGV) to search for splicing abnormalities. Once aberrant splicing was identified in cases with VUS, multiple open-source alternative splicing tools (MAJIQ, rMATS-turbo, FRASER2 and LeafCutterMD) were used to investigate if they would identify what was observed in IGV. Expression outliers were detected using OUTRIDER. To find diagnoses in cases without a VUS or gene of interest, two separate strategies were used. The first was a genotype to phenotype approach using variant calls obtained from the RNA-sequencing and overlapping those calls with results from splicing tools. The second strategy involved using phenotype information available to filter results from splicing tools. ResultsUsing RNA-sequencing only, we were able to assess 71% of VUSs and detect aberrant splicing in 14/48 patients with a VUS. Furthermore, we identified four new diagnoses by detecting novel aberrant splicing events in patients with no candidate sequence variants from prior genomic DNA testing (n=33) or those in which the candidate VUS did not affect splicing (n=23) and identified one additional diagnosis through detection of skewed X-inactivation. ConclusionWe demonstrate the identification of novel diagnoses using an RNA-sequencing first approach in patients without candidate VUSs. Furthermore, we demonstrate the utility of blood-based RNA analysis in improving diagnostic yields and highlight optimal approaches for such analysis.