Source-aware drug-disease relationship dataset integrating 11 public biomedical resources

This Zenodo release contains a source-aware drug-disease relationship dataset generated from 11 public biomedical resources: AACT/ClinicalTrials.gov, ChEMBL, Comparative Toxicogenomics Database (CTD), DrugCentral, Hetionet, MEDI, Open Targets Platform, PrimeKG, SIDER 4.1, the Broad Drug Repurposing Hub, and repoDB.The release includes the raw source-level merged dataset, the source-level schema-record deduplicated dataset, a source-collapsed representation, a data dictionary, and processing summary metadata. The raw merged dataset contains 10,870,837 source-level records. Schema-record deduplication produced 4,465,457 source-level deduplicated records. The source-collapsed representation contains 4,464,947 records.Records preserve source provenance, relationship type, evidence type, drug and disease/condition names, identifiers where available, and source-specific metadata. The dataset should be interpreted as a heterogeneous, source-aware relationship resource, not as a uniformly approved indication database or a predictive drug repurposing model.The accompanying code repository is available at: https://github.com/MuhammadMuneeb007/drug-disease-mappingUsers should consider the licences, attribution requirements, and reuse terms of the original upstream databases when reusing or redistributing derived records.