Multiscale computational genomics in Wilson disease: from atomic dynamics to clinical prediction

Wilson disease (WD) is an autosomal recessive disorder caused by pathogenic variants in the ATP7B gene, leading to toxic copper accumulation. The integration of computational genomics approaches is now essential for deciphering the complex genotype-phenotype relationships and advancing towards targeted therapies. This review synthesizes how multiscale computational strategies are transforming WD research. At the atomic level, molecular dynamics (MD) simulations reveal the conformational dynamics of the ATP7B protein, the functional impact of mutations, and the detailed copper transport cycle. At the systems level, machine learning (ML) models integrate genomic, epigenomic, transcriptomic, and clinical data to classify variant pathogenicity, predict disease subtypes, and forecast clinical outcomes such as cirrhosis or neurological deterioration. Furthermore, multi-omics network analyses uncover disease-associated regulatory modules, elucidate the role of epigenetic dysregulation, and implicate emerging pathways like cuproptosis in WD pathogenesis. Critically, these computational insights are increasingly guiding therapeutic innovation, including the in silico design of allosteric modulators (e.g., nanobodies) and pharmacological chaperones to correct ATP7B folding. By bridging scales from molecular structure to patient phenotypes, computational genomics provides a powerful, integrative framework that holds the potential to accelerate the development of dynamic, mechanism-based therapies and pave the way for personalized medicine in Wilson disease.