Artificial intelligence is being hailed as a potential game-changer in the fight against superbugs. A recent study published in the journal Cell suggests that generative AI algorithms can be trained to design new, potent antibiotics that can combat even the most resistant bacterial strains.
Led by MIT professor Jim Collins, the research team used a neural network framework to virtually screen over 45 million chemical fragments and generate millions of potential molecules with predicted antibacterial activity. The team then filtered through these candidates, identifying seven that proved to be effective in laboratory tests, including two that showed particularly strong efficacy against gonorrhea and staphylococcus skin infections.
The discovery is a significant breakthrough in the field of antibiotic development, where the lack of new treatments has led to increased mortality rates. The research was conducted by Phare Bio, a social-enterprise non-profit created by Collins, which aims to advance these and other AI-discovered antibiotics toward clinical development.
The team’s approach sets down a new path for antibiotic discovery, using generative AI as a design tool rather than just a discovery tool. This could potentially lead to the creation of novel structural classes of molecules that have not been seen before.
While the technology still has limitations, including synthetic tractability and production costs, the researchers believe that their approach holds great promise for developing new, effective antibiotics. The study’s findings were recently published in Cell and are being hailed as a significant breakthrough in the fight against superbugs.
Source: https://spectrum.ieee.org/ai-drug-design-mit-antibiotics