Computational drug repositioning through heterogeneous network clustering.

Area of Science:

• Computational biology
• Pharmacology
• Drug discovery

Background:

• Drug discovery is expensive and slow, with high failure rates.
• Drug repositioning (repurposing) offers a viable strategy to address these challenges.
• Integrating mechanistic relationships and heterogeneous networks is crucial but underutilized in computational drug repositioning.

Purpose of the Study:

• To develop a computational approach for identifying drug repositioning candidates by integrating heterogeneous network information.
• To explore drug-disease relationships beyond simple drug-drug or disease-disease similarities.

Main Methods:

• Constructed a weighted disease and drug heterogeneous network using KEGG database relationships (disease-gene, drug-target).
• Network nodes represent drugs and diseases; edges represent shared genes, pathways, phenotypes, or other features.
• Clustered the network to identify modules and generated drug-disease pairs as potential repositioning candidates, validated through robustness analysis and literature/clinical trial overlap.

Main Results:

• Successfully built a weighted heterogeneous network of diseases and drugs.
• Identified modules within the network to generate putative drug repositioning candidates.
• Validated predictions by assessing robustness and overlap with known drug indications.

Conclusions:

• The developed approach offers a holistic perspective on drug-disease relationships, integrating multiple features beyond genes.
• This method complements existing computational drug repositioning strategies.
• The approach's simplicity and validated predictions suggest its utility in discovering new therapeutic applications for existing drugs.