Metapath-guided topology and domain knowledge fusion for predicting miRNA-drug-disease triplet resistance and sensitivity associations.

MTDKF predicts miRNA–drug–disease triplet resistance (MRDR) and sensitivity (MRDS) associations by fusing metapath topology with domain knowledge, consistently outperforming prior methods.

• Why it matters: Drug response depends on disease context, yet most computational approaches model only miRNA–drug pairings and miss disease-specific effects on resistance and sensitivity.
• What they did: The method builds a triplet heterogeneous graph for miRNAs, drugs, and diseases, then uses a metapath topology encoder plus a domain-knowledge encoder (sequence, structure, semantics) to integrate representations.
• The result: On two manually curated benchmark datasets from experimentally validated associations, MTDKF achieves higher performance across multiple evaluation metrics and identifies disease-specific candidate miRNAs for doxorubicin and tamoxifen in breast cancer.

AI summary of the abstract · may contain errors · not medical advice · read the paper.