The rise of antimicrobial-resistant pathogens has outpaced the traditional methods of drug discovery and development, emphasizing a need for new and innovative approaches to identifying novel antibiotics. Artificial intelligence (AI) poses new opportunities to overcome the challenges in traditional drug discovery by accelerating the identification, design, and optimization of bioactive small molecules and antimicrobial peptides. AI-driven genome mining allows for the identification and prioritization of biosynthetic gene clusters, while advanced AI models facilitate molecular property prediction, predicted binding interactions, and novel structure design. This review explores the advancements that AI has enabled in antimicrobial discovery and design, as well as its current limitations. • Implementation of AI can decrease experimental time from hit to lead compound. • Models can parse through millions of hits at a fast rate. • Increased hit rates can accelerate AI-driven antibiotic discovery. • Property predictors can aid in specific antimicrobial targeting. • Small molecules and antimicrobial peptides are potent inhibitors.
Clements et al. (Wed,) studied this question.