Much of the conversations around artificial intelligence (AI) as of late has been in the realm of what are the promise and perils it could bring. Can it increase the risk of bioterrorism or help us combat disease?1 In the midst of this, sits the growing threat of antimicrobial resistance (AMR) and the hurdles to develop effective drugs. According to the World Health Organization (WHO), AMR is one of the top global public health threats, directly responsible for 1.27 million deaths and contributing to 4.95 million deaths in 2019, and expected to cost $1 trillion USD in healthcare costs by 2050.2
At this very intersection is a possibility for growth—harnessing the power of AI to address AMR and accelerate drug discovery. A new article in Cureus discusses the utility of AI resources to help close the gap between drug discovery and deployment in the age of AMR.3 In their paper, the researchers noted the timeline of antimicrobial drug discovery and emergence of resistance, which serves as a stark reminder for the fast pace of resistance and the slowness of discovery/development.
What You Need to Know
Artificial intelligence is being explored as a tool to accelerate drug discovery and combat antimicrobial resistance (AMR), which is a major global health threat responsible for millions of deaths and projected to cost $1 trillion in healthcare by 2050.
The review highlights how AI can optimize the drug development process, citing the case of halicin, an antibiotic discovered using machine learning algorithms. AI has also been applied in precision prescribing to reduce resistance risks.
Despite its promise, AI in antibiotic development faces challenges such as the need for high-quality datasets and validation through experimental trials.
The authors performed a narrative review to assess how AI might impact time and cost of drug discovery/development, optimize processes, and overall capacity to predict resistant patterns. The review included 1228 records screened, 490 assessed for eligibility, and ultimately 53 were included in the report. Citing the case study of halicin as a prime example for how AI can be utilized to aid in drug discovery, the authors note that while initially suggested as a treatment for diabetes, “in 2020, through the application of AI and the use of machine learning algorithms, the ZINC15 database, which is a collection of almost 1500 million chemical compounds, was used to discover how halicin exhibited unique antibacterial activity against several strains of harmful bacteria, including MDR bacteria,” the authors wrote.
In this same vein, the increasing role of AI in clinical medicine has been noted, with the authors emphasizing how machine learning has aided in precision prescribing as a means to reduce potential resistance. Ultimately, the authors underscored the significant potential AI has in not only drug discovery/development, but also enhancing antimicrobial stewardship.
“Despite its immense promise, the implementation of AI in antibiotic development faces challenges, such as the need for high-quality, comprehensive datasets, and the validation of AI-generated findings through rigorous experimental trials,” the authors wrote.
Ultimately, though, it will come down to larger efforts in not only AI utilization, but larger training, integration, guardrail development.
References
1.Bloomfield, D., Pannu, J., Zhu, A. W., Ng, M. Y., Lewis, A., Bendavid, E, et al (2024). AI and biosecurity: The need for governance. Science, 385(6711), 831-833.
2.Antimicrobial resistance. WHO. November 21, 2023. https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance
3.Zavaleta-Monestel E, Rojas-Chinchilla C, Campos-Hernández J, et al. (January 31, 2025) Utility of Artificial Intelligence in Antibiotic Development: Accelerating Discovery in the Age of Resistance. Cureus 17(1): e78296. doi:10.7759/cureus.78296
