In a landscape overshadowed by worries regarding the misuse of Artificial Intelligence for disinformation and deep fakes, a glimmer of positivity emerges. Scientists have achieved a remarkable breakthrough using AI technology, uncovering a new antibiotic with the ability to eradicate a highly perilous superbug.
Recently featured in the science journal Nature Chemical Biology, the published study reveals the successful identification of an antibiotic by researchers from the Massachusetts Institute of Technology and McMaster University. This groundbreaking achievement, detailed on Thursday, highlights the antibiotic’s potent antimicrobial properties against the deadly hospital superbug.
The World Health Organization (WHO) has highlighted the concerning nature of Acinetobacter baumannii bacteria, as it possesses the ability to develop new resistance mechanisms and transfer genetic material, thereby making other bacteria drug-resistant. These bacterial strains are regarded as the “greatest threat” to human health.
Healthcare facilities, including hospitals and nursing homes, are particularly vulnerable to this bacterium, as it poses a risk to patients who require ventilators, blood catheters, or have surgical wounds. Notably, Acinetobacter baumannii can persist on environmental surfaces and shared equipment for extended periods. Its transmission occurs through contaminated hands. Apart from causing bloodstream infections, it can also lead to infections in the urinary tract and lungs.
According to the Centers for Disease Control and Prevention (CDCP), Acinetobacter baumannii has the ability to colonize or reside in a patient’s body without causing any infections or noticeable symptoms.
What The Study Revealed
In this recent study, researchers leveraged an Artificial Intelligence algorithm to screen an extensive library of antibacterial molecules, predicting novel structural classes. Through AI screening, the team successfully identified a new antibacterial compound named “Abaucin.”
Utilizing the trained AI model, scientists analyzed 6,680 previously unencountered compounds. Within just an hour and a half, the analysis generated several hundred compounds, with 240 selected for further laboratory testing. Through these tests, nine potential antibiotics were identified, including the promising candidate, Abaucin.
To evaluate its efficacy, the researchers tested Abaucin in a wound infection model in mice and observed its ability to suppress the infection.
Jonathan Stokes, an assistant professor at McMaster University’s Department of Biomedicine and Biochemistry and one of the study’s leaders, emphasized the validation of machine learning in the quest for new antibiotics, as reported by The Guardian.
“Using AI, we can rapidly explore vast regions of chemical space, significantly increasing the chances of discovering fundamentally new antibacterial molecules,” he said.