Blood-based RNA-Seq of 5412 individuals with rare disease identifies new candidate diagnoses in the National Genomic Research Library

Abstract RNA sequencing (RNA-Seq) is increasingly used alongside exome and genome sequencing to identify causal variants underlying rare Mendelian disorders. We present short-read RNA-Seq data from 5,412 individuals with a diverse range of rare disorders recruited to Genomics England’s 100,000 Genomes Project. We show that the proportion of genes from gene panels applied to different disorders which are well captured (transcripts per million (TPM) ≥ 5) from blood RNA varies widely, highlighting differences in applicability across disorder types. Using OUTRIDER and FRASER2 to identify gene expression and splicing outliers respectively, we identify at least one outlier event in a disorder relevant gene in 20% of the cohort. To prioritise likely diagnostic candidates, we apply multiple strategies including focussing on outlier events in known haploinsufficient genes (n=78), integrating outliers with structural variant calls (n=19), and using strategies integrating phenotypic presentation (Exomiser, n=39). We present a series of candidate diagnoses involving diverse variant types and disease mechanisms, demonstrating the broad utility of RNA-Seq in identifying and prioritising diagnostic candidates in individuals with a variety of different rare conditions and no known genetic diagnosis. Our findings demonstrate that blood-based RNA-Seq can deliver clinically relevant findings across a broad range of rare disorders.