MIT Researchers Leverage AI to Identify Antibiotic That Can Kill Drug-Resistant Bacteria
As bacteria continue to evolve to withstand the effects of antibiotics, it has rendered bacterial infections more challenging to treat. The issue of “antibiotic resistance” has already become a critical health issue.
The overuse and misuse of antibiotics have made the issue worse. Researchers have now turned to artificial intelligence to find innovative ways to fight against this issue and MIT researchers might have found a way to crack the code.
Using artificial intelligence, MIT researchers identified a new class of antibiotics that could kill a drug-resistant bacterium that causes more than 10,000 deaths in the U.S. every year. The research was based on using deep learning – a type of AI that teaches computers to process data in a way that is inspired by the human brain.
The researchers were successful in showing that these compounds could kill methicillin-resistant Staphylococcus aureus (MRSA), which is resistant to several types of antibiotics including methicillin, penicillin, and amoxicillin. MRSA can cause various infections including some that are life-threatening.
The compounds identified by the MIT research that can kill drug-resistant bacteria also show very low toxicity when tested with human cells, making them good candidates for human use.
James Collins, one of the lead researchers of the study and the Termeer Professor of Medical Engineering and Science at MIT says “The insight here was that we could see what was being learned by the models to make their predictions that certain molecules would make for good antibiotics. Our work provides a framework that is time-efficient, resource-efficient, and mechanistically insightful, from a chemical-structure standpoint, in ways that we haven’t had to date.”
AI has recently been at the forefront of research in the world of medicine and healthcare. Last month, researchers from the Oxford Martin School published a ground-breaking study where they used AI to detect antimicrobial resistance (AMR). This new advancement is set to pave the way for novel and rapid antimicrobial susceptibility tests.
Using deep learning for new drugs is not a new phenomenon but as AI models become more sophisticated, the capabilities of such systems get stronger. One of the key insights of this study was the researchers were able to pinpoint what kind of data is used by deep learning models to make antibiotic potency predictions.
This knowledge can empower other researchers to develop drugs that might work even better than the ones identified by this study. According to Felix Wong, the lead co-author of the MIT study, the study will help “open the black box” to help other researchers understand how DL models work.
The MIT researchers have shared their findings with Phare Bio, which is a social venture using novel AI and Deep Learning to tackle the world’s most urgent threats. Collins is one of the founders of Phare Bio. The startup plans on doing a more detailed analysis of the data and identifying potential clinical use cases of the compounds. Meanwhile, the authors of the study will focus on using their DL models to seek compounds that can kill other types of bacteria.
This article originally appeared in Datanami.