Fibroblast transcriptomics in molecular diagnostics of a comprehensive dystonia cohort

Exome and genome sequencing leave >50% of dystonia-affected individuals without a molecular diagnosis. Where DNA-oriented approaches remain insufficient, integrating multiomics methods and bioinformatics is essential to advance genome interpretation. Herein, we incorporated RNA sequencing (RNA-seq) from a collection of 167 fibroblast samples from individuals affected with dystonic diseases. We leveraged an RNA-seq analysis pipeline, focused on the identification of expression and splicing aberrations, on RNA-seq from skin biopsies. We evaluated a "variant-positive" group of patient samples with preexisting information on variants (36/167, 21.6%), and a "variant-negative" group in which genomic sequencing alone had been unsuccessful in yielding a diagnostic candidate (78.4%). We found that at least 80% of dystonia-associated genes from databases were sufficiently detected by RNA-seq in fibroblasts, highlighting broad applicability. Expression and splicing aberration analyses then produced a manageable number of statistically significant RNA defects affecting dystonia-associated genes for effective case-by-case review. Our approach successfully detected RNA underexpression and mis-splicing for different types of pre-identified dystonia-related variants, providing both benchmarks and insights into mutational mechanisms. Applied to 131 samples from patients without candidate variants from exome and genome sequencing, RNA-seq aided the identification of previously unprioritized causative intronic alterations on reanalysis, providing an added diagnostic yield of 6.9% (9/131). For observed events, we also report the integration of new machine-learning scores predicting corresponding aberrant gene expression in the brain. Fibroblast-based RNA-seq in our selected cohort improved variant interpretation and enabled diagnoses missed by genomic analysis alone, suggesting this framework could be generalized to other dystonias.