Drug repositioning via matrix completion with multi-view side information

Area of Science:

• Pharmacology and Bioinformatics
• Computational Drug Discovery

Background:

• Drug repositioning is crucial for identifying new therapeutic uses for existing drugs.
• Existing drug-target interaction prediction methods often rely on limited data, impacting accuracy.
• Integrating multiple data sources (multi-view information) can enhance prediction capabilities.

Purpose of the Study:

• To develop an advanced drug-target interaction prediction approach.
• To improve the accuracy of identifying biological targets for existing drugs using multi-view side information.
• To facilitate drug repositioning by predicting novel drug-target interactions.

Main Methods:

• A novel matrix completion approach named MCM (Multi-view drug-target interaction prediction by matrix completion) was developed.
• The MCM method integrates multi-view side information of drugs and proteins, including structural and chemical views.
• Drug-target interactions are predicted by directly completing the interaction matrix.

Main Results:

• The MCM method demonstrated significantly higher prediction accuracies compared to existing methods.
• The approach successfully identified new drug-target interactions for 26 FDA-approved drugs.
• Experimental results validate the effectiveness of integrating multi-view information for improved prediction.

Conclusions:

• The MCM approach offers a robust and accurate method for drug-target interaction prediction.
• This study advances drug repositioning by providing reliable predictions of novel drug-target associations.
• The identified drug-target interactions warrant further biological investigation for therapeutic development.