Genome-wide functional annotation and interpretation of splicing variants: toward RNA-targeted therapies.

Splice-disruptive variants represent an underrecognized yet critical category of disease-causing mutations. While canonical splice site disruptions have long been associated with genetic disorders, it is now increasingly evident that synonymous, deep-intronic, and regulatory variants can also perturb splicing events and contribute to diseases. As genomic diagnostics shift from phenotype-first to genome-first paradigms, there is an urgent need for systematic strategies to identify and interpret such variants-including those residing in noncoding regions that escape detection by traditional annotation pipelines. This review provides an integrative overview of current in silico approaches for the annotation and interpretation of splice-disruptive variants. We outline the mechanistic diversity of splicing aberrations and discuss recent advances in computational prediction frameworks, including both deep learning-based models and motif-oriented tools. In parallel, we summarize experimental strategies that are used to validate predicted splicing effects and assess their pathogenic relevance. Focusing on clinically relevant contexts, we discuss how splicing-aware variant interpretation enhances diagnostic yield, informs the reclassification of variants of uncertain significance, and uncovers targets for therapeutic intervention. Finally, we consider the implications of such interpretation for RNA-targeted strategies, including antisense oligonucleotides, small-molecule modulators, and emerging RNA-editing platforms, particularly in neuromuscular and other splicing-driven disorders. Together, these insights underscore the expanding role of in silico splicing prediction in precision medicine, offering new diagnostic and therapeutic avenues for rare and undiagnosed genetic diseases.