Using artificial intelligence, researchers at McMaster University and the Massachusetts Institute of Technology have discovered a new antibiotic that could be used to combat a drug-resistant pathogen that kills hospital patients who are at risk.
The method they used could also expedite the development of additional antibiotics to treat numerous other difficult bacteria.
The World Health Organization identified Acinetobacter baumannii as one of the most dangerous antibiotic-resistant bacteria in the world, prompting an urgent need for new treatments. Famously challenging to destroy, A. baumannii can cause pneumonia, meningitis, and contaminate wounds, all of which can prompt demise.
A. baumanni is usually found in hospitals because it can live for a long time on surfaces. The pathogen can acquire antibiotic-resistance gene DNA from other bacterial species in its environment.
“We can rapidly explore vast regions of chemical space using AI, significantly increasing the chances of discovering fundamentally new antibacterial molecules,”
Stokes, who belongs to McMaster’s Global Nexus School for Pandemic Prevention and Response.
Abaucin is the name given to a new antibacterial compound discovered by researchers using an artificial intelligence algorithm to predict new structural classes of antibacterial molecules. The study was published today in the journal Nature Chemical Biology.
It has been hard to find new antibiotics for A. baumannii through traditional screening. The scope of traditional methods is limited, expensive, and time-consuming.
There are hundreds of millions, if not billions, of molecules with antibacterial properties that can be accessed by modern algorithmic methods.
Lead creator Jonathan Stirsup is an associate teacher in the branch of natural chemistry and biomedical science at McMaster College. A new antibacterial compound was discovered by researchers to treat the pathogen Acinetobacter baumannii. Credit: McMaster College.
“This work approves the advantages of AI in the quest for new anti-microbials,” says Jonathan Stirsup, lead creator on the paper and an associate teacher in McMaster’s Branch of Biomedicine and Organic Chemistry, who led the work with James J. Collins, a teacher of clinical design and science at MIT, and McMaster graduate understudies Gary Liu and Denise Catacutan.
According to Stokes, who is a student at McMaster’s Global Nexus School for Pandemic Prevention and Response, “we can rapidly explore vast regions of chemical space using AI, significantly increasing the chances of discovering fundamentally new antibacterial molecules.”
According to Collins, the Life Sciences faculty lead at the MIT Abdul Latif Jameel Clinic for Machine Learning in Health, “AI approaches to drug discovery are here to stay and will continue to be refined.” We know that algorithmic models work, so the next step is to use them widely to find new antibiotics faster and for less money.
Abaucin is particularly encouraging, the specialists report, since it just focuses on A. baumannii, an essential finding that implies the microorganism is less inclined to quickly foster medication obstruction and could prompt more exact and successful medicines.
The majority of antibiotics are broad-spectrum in nature, which means that they kill all bacteria and disrupt the gut microbiome, making it possible for a wide range of serious infections, such as C. difficile, to develop.
According to Stokes, “pathogens have the ability to evolve and adjust to every trick we throw at them,” and “we know broad-spectrum antibiotics are suboptimal.” We can significantly accelerate the rate at which we discover new antibiotics using AI techniques at a lower cost. New antibiotics should investigate this extensively.
More information: James Collins, Deep learning-guided discovery of an antibiotic targeting Acinetobacter baumannii, Nature Chemical Biology (2023). DOI: 10.1038/s41589-023-01349-8. www.nature.com/articles/s41589-023-01349-8
