PPAEDTI

PPAEDTI predicts drug–target interactions to support drug repurposing and development.

Key Features:

• Graph Personalized Propagation Technique: Applies graph personalized propagation on a known interaction network to leverage existing drug–target interaction data for prediction.
• High Prediction Accuracy: Evaluated on six benchmark datasets with comparisons to state-of-the-art methods, achieving average Area Under the Curve (AUC) scores greater than 90% across five datasets using 5-fold cross-validation.
• Manual Validation of Predictions: Top-20 predictions for proteins hsa:775 and hsa:779 and for drug D00618 were manually verified against public datasets, confirming 18, 17, and 20 items respectively.
• Research Applications: Facilitates identification of potential drug targets by generating accurate predictions based on known interactions.

Scientific Applications:

• Drug repurposing and development: Uses predicted drug–target interactions to prioritize candidate drugs for repurposing and therapeutic development.
• Target identification: Supports efficient identification of potential drug targets from interaction networks.
• Discovery of novel interactions: Predicts previously unreported drug–target interactions for follow-up validation.

Methodology:

Uses graph personalized propagation within a known interaction network; evaluation employed six benchmark datasets with 5-fold cross-validation and AUC metrics, comparisons to state-of-the-art methods, and manual verification of top-20 predictions against public datasets.