Machine learning methods, databases and tools for drug combination prediction.

Area of Science:

• Computational biology
• Pharmacology
• Artificial intelligence

Background:

• Combination therapy shows efficacy for complex diseases and reduces drug resistance.
• Experimental methods for discovering novel drug combinations are insufficient.
• Efficient computational methods are needed to predict drug combinations and reduce search space.

Purpose of the Study:

• To introduce and discuss recent applications of machine learning (ML) methods in drug combination prediction.
• To review widely used databases and tools for drug combination prediction.
• To summarize challenges and future directions for ML in this field.

Main Methods:

• Literature review of machine learning algorithms (classic ML and deep learning).
• Investigation of publicly available databases and prediction tools.
• Discussion on the concept and controversy of drug combination synergism.

Main Results:

• Machine learning methods are increasingly applied to improve drug combination prediction.
• Various data resources and computational tools are available for prediction tasks.
• Deep learning and classic ML approaches show promise in predicting synergistic drug combinations.

Conclusions:

• Machine learning offers a powerful approach to accelerate the discovery of effective drug combinations.
• Further research is needed to address challenges and enhance ML model performance in drug combination prediction.
• Integration of diverse data sources and advanced ML techniques will be key for future advancements.