Drug repurposing designates the systematic identification of new therapeutic indications for existing, failed, or investigational drug molecules that have already undergone preclinical or clinical development ^[1]. Within computational drug discovery methods, the term is defined by its operational logic: rather than designing a novel chemical entity de novo, repurposing exploits the known pharmacological, structural, or transcriptional profile of a pre-existing drug to infer its potential activity against a disease target or pathway for which it was not originally developed ^[2]. The computational methodological space of drug repurposing is organized into several complementary families: structure-based approaches that dock known drugs against new target structures; ligand- or signature-based approaches that match drug-induced transcriptional or phenotypic profiles to disease signatures; network-based approaches that infer repurposing candidates from the topology of drug–target–disease interaction networks; and matrix-decomposition or machine-learning approaches that predict drug–disease associations from large-scale heterogeneous data ^[1][3]. A defining characteristic of computational drug repurposing is its reliance on in silico resources—databases, predictive platforms, and knowledge graphs—that encode prior knowledge and enable the systematic prioritization of candidates for experimental validation ^[1][2]. The term thus denotes the computationally driven inference of new indications for known drugs, together with the algorithmic and knowledge-representation frameworks that make this inference reproducible and scalable ^[3][4].