Drug-target interaction/affinity prediction: Deep learning models and advances review.

Area of Science:

• Computational chemistry
• Pharmacology
• Artificial intelligence in drug discovery

Background:

• Drug discovery is a lengthy, costly, and safety-concerned process.
• Traditional methods struggle with complex drug-target relationships.
• Accurate prediction of drug-target interactions is crucial for faster drug development.

Purpose of the Study:

• To provide researchers with an overview of advanced methods for predicting drug-target interaction and affinity.
• To highlight promising research avenues and models in this field.
• To accelerate the development of more effective drugs.

Main Methods:

• Analysis of 180 drug-target interaction/affinity prediction methods from 2016-2025.
• Focus on machine learning, particularly deep learning and graph neural networks.
• Discussion of model novelty, architecture, and input representation.

Main Results:

• Deep learning models show significant potential in overcoming limitations of traditional methods.
• Identified various frameworks and approaches for precise and efficient interaction prediction.
• Comprehensive analysis covers a decade of research in computational drug discovery.

Conclusions:

• Deep learning and graph neural networks are key to advancing drug-target interaction prediction.
• Further research in these areas can significantly speed up the delivery of novel therapeutics.
• This review serves as a guide for researchers seeking to improve drug discovery efficiency.