Artificial intelligence: Machine learning approach for screening large database and drug discovery.

Area of Science:

• Computational chemistry
• Drug discovery
• Bioinformatics

Background:

• Drug discovery faces challenges like disease outbreaks and drug resistance.
• Virtual screening is a key computer-aided method for identifying drug targets from large molecular databases.
• Deep learning and artificial neural networks have significantly advanced drug discovery processes.

Purpose of the Study:

• To review the application of artificial intelligence (AI) and machine learning (ML) in virtual screening of large molecular databases.
• To highlight the advantages, drawbacks, and potential hazards of AI/ML techniques in drug discovery, with a focus on antivirals and infectious diseases.

Main Methods:

• Review of existing literature on AI/ML applications in drug discovery and virtual screening.
• Analysis of AI/ML tools and their use in various drug discovery processes.
• Discussion of the translation of AI/ML tools from non-communicable to infectious disease research.

Main Results:

• AI/ML algorithms have modernized various drug discovery stages, including virtual screening, toxicity prediction, and drug repositioning.
• Numerous web servers and AI/ML tools are available for screening chemical databases.
• Challenges exist in applying AI/ML tools, primarily developed for non-communicable diseases, to infectious disease treatment discovery.

Conclusions:

• AI/ML offers significant potential for accelerating the identification of novel antiviral and other therapeutic agents through virtual screening.
• Addressing the limitations and potential hazards of AI/ML is crucial for successful translation to infectious disease drug discovery.
• Further research is needed to optimize AI/ML approaches for tackling infectious disease challenges.