Remodelling structure-based drug design using machine learning.

Area of Science:

• Biomedicine
• Computational Chemistry
• Pharmacology

Background:

• Rational Drug Development and Structure-Based Drug Design (SBDD) have historically driven drug discovery.
• SBDD identified novel molecules but faced limitations due to post-delivery drug failures from adverse interactions.
• Technological advancements and clinical research spurred the development of Artificial Intelligence (AI) and Machine Learning (ML).

Purpose of the Study:

• To review the evolution of drug discovery methodologies.
• To highlight the technological advancements and methodologies of Machine Learning (ML) in drug design.
• To emphasize ML's contributions to biomedicine and explore future potential.

Main Methods:

• Review of historical drug development approaches (Rational Drug Development, SBDD).
• Focus on Machine Learning (ML) methodologies and technological advancements.
• Analysis of ML's application in predicting drug side-effects and aiding drug design.

Main Results:

• ML, a subset of AI, excels at data analysis and predictive modeling with minimal human intervention.
• ML algorithms utilize large training datasets to iteratively improve predictions of new output values.
• ML tools have demonstrated significant success in predicting drug side-effects, enhancing drug design.

Conclusions:

• Machine Learning (ML) is transforming drug discovery and development.
• ML offers powerful computational capabilities for identifying novel therapeutics and predicting adverse drug interactions.
• The review underscores the significant impact and future potential of ML in biomedicine.