Drug-target interaction prediction based on protein features, using wrapper feature selection.

Area of Science:

• Bioinformatics
• Computational Biology
• Drug Discovery

Background:

• Drug-target interaction (DTI) prediction is crucial for drug development.
• Experimental DTI identification methods are time-consuming, costly, and complex.
• Computational methods offer a more efficient alternative to experimental approaches.

Purpose of the Study:

• To propose a novel computational model for predicting drug-target interactions (DTI).
• To enhance the accuracy and efficiency of DTI prediction through advanced computational techniques.
• To reduce the time and cost associated with traditional experimental methods in drug discovery.

Main Methods:

• A three-phase computational model: feature extraction, feature selection, and classification.
• Feature extraction includes EAAC and PSSM from protein sequences and drug fingerprint features.
• Feature selection employs the IWSSR wrapper method, followed by Rotation Forest classification.

Main Results:

• The proposed model achieved high accuracy rates on golden standard datasets: 98.12% (enzyme), 98.07% (ion channels), 96.82% (G-protein-coupled receptors), and 95.64% (nuclear receptors).
• The IWSSR feature selection method effectively handled large amounts of extracted data.
• Rotation Forest classification demonstrated efficient DTI prediction capabilities.

Conclusions:

• The developed computational model shows acceptable performance in DTI prediction.
• The proposed model is compatible with existing methods and offers a promising approach for drug discovery.
• This computational strategy can accelerate the identification of drug-target relationships.