Computational Discovery of Putative Leads for Drug Repositioning through Drug-Target Interaction Prediction

Area of Science:

• Computational biology
• Drug discovery
• Bioinformatics

Background:

• De novo drug discovery is costly and time-consuming.
• Identifying drug-target interactions (DTIs) is essential for drug development.
• Computational models can accelerate DTI identification and reduce costs.

Purpose of the Study:

• To present a computational pipeline for discovering drug repositioning leads.
• To apply the pipeline to any microbial proteome with a partially known interactome.
• To identify novel drug-target interactions and score potential candidates.

Main Methods:

• Utilized network metrics to identify putative drug targets in microbial interactomes.
• Developed a random forest classification model for DTI prediction, achieving an AUC of 0.91.
• Constructed a drug-target network to predict and score new DTIs.
• Performed molecular docking on predicted DTI pairs.

Main Results:

• The DTI prediction model demonstrated high accuracy on out-of-sampling data.
• A network of 3,081 ligands and top drug targets was created for prediction.
• Molecular docking validated predicted drug-target interactions.
• The pipeline successfully identified potential leads for drug repositioning.

Conclusions:

• The proposed computational pipeline is effective for identifying new leads for drug repositioning.
• The developed DTI prediction model and pipeline can accelerate drug discovery efforts.
• The classification model is publicly available for research use.