Prediction of Drug-Target Interactions by Combining Dual-Tree Complex Wavelet Transform with Ensemble Learning...

Area of Science:

• Computational biology
• Drug discovery
• Bioinformatics

Background:

• Drug-target interactions (DTIs) are crucial for developing new pharmaceuticals.
• Traditional experimental methods for identifying DTIs are often time-consuming and costly.
• There is a need for efficient computational approaches to accelerate drug discovery.

Purpose of the Study:

• To develop a novel computational method for predicting drug-target interactions (DTIs).
• To utilize protein sequence information and the dual-tree complex wavelet transform (DTCWT) for feature extraction.
• To enhance the efficiency and accuracy of drug discovery processes.

Main Methods:

• Extracted evolutionary information from target protein sequences using position-specific scoring matrices (PSSMs).
• Applied DTCWT to PSSM data for feature extraction, combined with drug fingerprint features.
• Employed the Rotation Forest (RoF) model for classification, validated using 5-fold cross-validation on four diverse datasets.

Main Results:

• Achieved high average accuracies: 89.21% (Enzyme), 85.49% (Ion Channel), 81.02% (GPCRs), and 74.44% (NRs).
• Demonstrated superior performance compared to Support Vector Machine (SVM) and k-nearest neighbor (KNN) classifiers.
• Validation on an independent dataset confirmed the method's effectiveness in predicting potential DTIs.

Conclusions:

• The proposed computational approach effectively predicts drug-target interactions.
• This method offers a valuable tool for accelerating drug discovery and development.
• The integration of PSSM, DTCWT, and RoF provides a robust framework for DTI prediction.