Computational drug repositioning using meta-path-based semantic network analysis.

Area of Science:

• Computational biology
• Bioinformatics
• Network science

Background:

• Drug repositioning accelerates the discovery of new therapeutic uses for existing medications.
• Current network-based methods often overlook the semantic nuances of meta-paths in heterogeneous networks.
• There is a need for advanced methods to infer novel drug-disease associations.

Purpose of the Study:

• To propose a novel methodology, HeteSim_DrugDisease (HSDD), for effective drug repositioning.
• To enhance the prediction of new indications for approved drugs by leveraging semantic network analysis.
• To address limitations in existing network-based approaches for drug repositioning.

Main Methods:

• Construction of drug-drug and disease-disease similarity networks.
• Integration of these networks with known drug-disease associations into a heterogeneous network.
• Prediction of novel drug-disease associations using HeteSim scores derived from various meta-paths.

Main Results:

• HSDD demonstrated superior performance compared to state-of-the-art methods in predicting drug-disease associations.
• Achieved a high Area Under the Curve (AUC) score of 0.8994 in leave-one-out cross-validation.
• Case studies confirmed the practical utility and potential of HSDD in identifying new drug indications.

Conclusions:

• HeteSim_DrugDisease (HSDD) provides an effective and feasible approach for inferring drug-disease associations.
• Meta-path-based semantic network analysis is a powerful tool for drug repositioning.
• The HSDD methodology holds significant promise for advancing precision medicine.