Improved Prediction of Drug-Target Interactions Using Self-Paced Learning with Collaborative Matrix Factorization.

Area of Science:

• Computational biology
• Pharmacology
• Bioinformatics

Background:

• Identifying drug-target interactions (DTIs) is crucial for drug discovery, but experimental methods are costly and slow.
• Existing computational methods for DTI prediction have limitations.
• Chemogenomic approaches offer efficient alternatives for DTI prediction.

Purpose of the Study:

• To develop an efficient computational method for predicting drug-target interactions (DTIs).
• To introduce a novel self-paced learning with collaborative matrix factorization (SPLCMF) framework for DTI prediction.
• To improve the accuracy and efficiency of identifying potential drug candidates.

Main Methods:

• Utilized chemogenomic-based approaches for DTI prediction.
• Developed a self-paced learning with collaborative matrix factorization (SPLCMF) framework.
• Integrated multiple drug and target networks into regularized least-squares, focusing on low-dimensional feature representation.
• Employed soft weighting for sample selection from easy to complex during training.

Main Results:

• The proposed SPLCMF method demonstrated superior performance compared to existing state-of-the-art approaches on synthetic and benchmark datasets.
• SPLCMF effectively predicts unknown drug-target interactions.
• The framework accurately reflects the latent importance of training samples.

Conclusions:

• The SPLCMF framework provides a valuable tool for predicting unknown drug-target interactions.
• This method can offer new insights for drug discovery, side-effect prediction, and drug repositioning.
• The approach enhances the efficiency and accuracy of computational drug discovery pipelines.