Machine learning models for drug-target interactions: current knowledge and future directions.

Area of Science:

• Computational chemistry
• Pharmacology
• Artificial intelligence in drug discovery

Background:

• Accurate prediction of binding affinity is essential for efficient drug discovery.
• Computational methods reduce the need for extensive wet-lab experiments.
• Machine learning (ML) and deep learning (DL) are increasingly used to predict drug-target interactions (DTIs).

Purpose of the Study:

• To review ML and DL models for predicting binding affinities and DTIs.
• To highlight current advancements in computational drug discovery.
• To identify future research directions in DTI prediction.

Main Methods:

• Review of ligand-based and target-based ML/DL approaches.
• Analysis of models used in virtual screening for DTI prediction.
• Discussion of computational strategies for enhancing drug discovery pipelines.

Main Results:

• ML and DL models show significant promise in predicting binding affinities.
• These computational techniques enhance the efficiency of identifying lead compounds.
• Virtual screening using advanced models improves DTI prediction accuracy.

Conclusions:

• ML and DL are powerful tools for accelerating drug discovery.
• Further development in DTI prediction models is crucial.
• Computational approaches offer a cost-effective and time-saving alternative to traditional methods.