Revealing Drug-Target Interactions with Computational Models and Algorithms.

Area of Science:

• Pharmacology
• Bioinformatics
• Computational Biology

Background:

• Identifying drug-target interactions (DTIs) is crucial for drug research and development.
• Experimental methods for DTI validation are costly, time-consuming, and have low success rates.
• Computational models offer an efficient alternative for inferring potential DTI candidates.

Purpose of the Study:

• To provide a comprehensive review of computational models for DTI identification.
• To discuss the strengths and limitations of existing DTI prediction methods.
• To suggest future directions for improving DTI prediction accuracy.

Main Methods:

• Review of network-based algorithms for DTI identification.
• Overview of machine learning-based methods, including bipartite local model, matrix factorization, regularized least squares, and deep learning.
• Discussion of relevant databases and software packages.

Main Results:

• Computational methods have significantly advanced DTI prediction.
• Existing models, including network-based and machine learning approaches, have inherent limitations.
• Potential strategies for enhancing DTI prediction accuracy were explored.

Conclusions:

• Computational approaches are vital for efficient DTI candidate identification.
• Further research is needed to overcome the limitations of current DTI prediction models.
• Optimizing DTI prediction accuracy will accelerate drug discovery and development.