Genome-Wide Analysis Reveals a Role of 3’-untranslated Region Variants Affecting Cleavage and Polyadenylation in Undiagnosed Rare Disorders

Cleavage and polyadenylation of pre-mRNAs are essential for transcription termination and the normal expression of eukaryotic genes. However, the extent to which mutations in cleavage and polyadenylation signals act as causal variants in rare Mendelian disorders remains unknown. Using deep learning models, we identified enrichment of alleles predicted to disrupt polyadenylation in a large cohort of undiagnosed probands with rare disease disorders from the Genomics England 100kGP dataset. These alleles were predominantly located in the 3’ UTRs of genes known to be causally associated with rare diseases. Using the SLC16A2 gene linked to Allan-Herndon-Dudley syndrome as an example, we show that a putative causal variant predicted to disrupt cleavage and polyadenylation in a proband with intellectual disability has a deleterious impact on SLC16A2 expression. Our findings highlight the underappreciated contribution of cleavage and polyadenylation-disrupting variants to the genetic basis of rare diseases.