The utility of transcriptomics to increase diagnostic yield in cases of rare genetic disease

Rare diseases affect less than 1 in 200,000 individuals by definition, however, collectively are common, with approximately 30 million individuals in the United States impacted. Frequently individuals experience an arduous journey towards a rare disease diagnosis, often termed a “diagnostic odyssey”. Exome sequencing (ES) has become standard of care for patients with suspected rare genetic disease, with a diagnostic yield of 30%. Here we report on preliminary results of the largest study to date integrating transcriptome sequencing (RNAseq) with ES to increase diagnostic yield for rare disease patients. A cohort of 406 patients who were consented into the Mayo Clinic Diagnostic Odyssey study were selected for study. All patients had received clinical or research exome sequencing which failed to result in a definitive diagnosis. RNA was extracted from whole blood collected in PAXgene tubes, and sequencing libraries prepared using a hybridization capture protocol with the TruSeq RNA Access Library Prep Kit (Illumina, San Diego, CA). Paired-end 101-basepair reads were sequenced using Illumina instruments. Resulting RNA sequences were analyzed for fusion using two algorithms, LeafCutterMD and Fraser, with results integrated for review. Allele specific expression was determined using AnevaDot. Gene expression was analyzed for outlier expression patterns in individual patients compared to the entire cohort using Outrider. Fusion analysis was performed using an internal workflow previously described. The major advantages RNAseq in addition to ES are the high-throughput functional characterization of DNA variation and the detection of aberrations not identifiable by ES sequencing alone. Recent work has shown a 7.5-35% yield using RNAseq in unsolved rare disease cases, dependent on tissue and methodology. In this study RNAseq was performed on blood from 406 individuals with suspected rare genetic disease for whom clinical ES was non-diagnostic. The transcriptome data was analyzed to interrogate four different effects: genes with outlier expression, aberrant RNA-splicing, gene fusions, and allele specific expression. Splicing analysis utilized two complementary algorithms, FRASER and LeafcutterMD, to identify modified intron usage, shifted splice donor and acceptor sites, intron retention events, and creation of cryptic exons. All data were studied for ultra-rare modifications with potential disease ramifications, using the remaining patient cohort as a control RNASeq population. Our preliminary results include gene fusion events, splice site variants with confirmed splicing alterations, and significantly altered outlier gene expression which were integral to the patient’s genetic diagnosis. RNA sequencing is complementary to DNA sequencing in the search for genetic diagnoses in Mendelian disease. When analyzing RNA data, care must be taken to evaluate all forms of RNA mediated events to obtain a thorough representation of the underlying genetics. Here we discuss a 406 patient cohort study using RNA and DNA sequencing to profile patient seen at the Mayo Clinic. Complete results of this study will provide valuable information regarding the clinical utility of RNAseq in tandem with ES for identifying a genetic diagnosis for rare disease patients.