Splicing Variant Analysis Enables Unsolved Rare Disease Diagnosis

By examining splicing variants within whole-genome sequencing data from roughly 38,000 people from the 100,000 Genomes Project, a University of Southampton-led team was able to diagnose a number of unsolved rare diseases among the project's participants. The study, which appears in Genome Medicine this week, highlights the clinical value of noncanonical splicing variant analysis. In the study, the researchers used a variant-level constraint metric to evaluate the contribution of variants at or near canonical splice sites and at predicted branchpoints to rare genetic diseases in the 100,000 Genomes Project cohort. They show that predicted splicing branchpoints harbor deleterious non-coding variants and use a gene-agnostic approach to prioritize 258 de novo variants that potentially disrupt splicing in families affected by a rare genetic disorder. Of this total, 84 were already considered to be diagnostic, with an additional 35 deemed likely to be diagnostic based on available molecular, phenotypic, and in silico data. Using their findings, the investigators confirmed a new diagnosis for six individuals.