Artificial intelligence in small molecule drug discovery from 2018 to 2023: Does it really work?

Area of Science:

• Computational chemistry and pharmacology
• Bioinformatics and cheminformatics
• Drug discovery and development

Background:

• Artificial intelligence (AI) is revolutionizing drug design, offering advanced methods for target identification and new drug development.
• Integrating AI techniques streamlines the drug development process, significantly enhancing early-stage discovery efficiency.
• AI applications in small molecule drug design are rapidly expanding, impacting various stages of research.

Purpose of the Study:

• To provide a comprehensive review of AI applications in small molecule drug design.
• To focus on key AI areas: protein structure prediction, virtual screening, molecular design, and ADMET prediction.
• To explore the role, limitations, and impact of AI on decision-making in drug development.

Main Methods:

• Literature review of AI methodologies applied to drug design.
• Analysis of AI's contribution to protein structure prediction.
• Examination of AI in molecular virtual screening and de novo molecular design.
• Assessment of AI for predicting absorption, distribution, metabolism, excretion, and toxicity (ADMET).

Main Results:

• AI significantly enhances efficiency and reduces workload in early-stage drug discovery.
• AI methods show promise in protein structure prediction, virtual screening, and molecular design.
• AI-driven ADMET prediction aids in identifying potential drug candidates with favorable pharmacokinetic profiles.
• AI integration impacts decision-making processes, highlighting its role as a supportive tool.

Conclusions:

• AI offers substantial benefits for small molecule drug design, particularly in early stages.
• AI should be viewed as a powerful tool to augment, not replace, human expertise in drug development.
• Careful consideration of AI's limitations and the importance of human-guided decision-making are essential for successful drug discovery.