Mining integrated semantic networks for drug repositioning opportunities

Area of Science:

• Computational biology
• Pharmacology
• Bioinformatics

Background:

• Traditional drug discovery is increasingly costly and less fruitful.
• Drug repositioning, identifying new uses for existing drugs, is a promising alternative.
• Current methods for drug repositioning lack systematic, scalable approaches.

Purpose of the Study:

• To develop a systematic and scalable computational methodology for drug repositioning.
• To introduce a formal framework for integrated networks and semantic subgraphs in drug interaction analysis.
• To present DReSMin, an algorithm for mining these networks to identify drug repositioning opportunities.

Main Methods:

• Developed a formal framework for integrated networks and semantic subgraphs.
• Introduced DReSMin, an algorithm for mining semantically-rich networks.
• Applied DReSMin to an integrated drug interaction network from 11 sources.

Main Results:

• Identified and ranked 9,643,061 putative drug-target interactions.
• Demonstrated a strong correlation between high-scoring associations and literature support.
• Top-ranked associations included 14 novel opportunities and 6 literature-supported ones.

Conclusions:

• DReSMin provides a computationally tractable and scalable method for drug repositioning.
• The approach effectively prioritizes known drug-target interactions over existing methods.
• This systematic mining of integrated networks can accelerate the discovery of novel therapeutic applications for existing drugs.