With the assistance of artificial intelligence, researchers at MIT developed novel antibiotics aimed at fighting difficult-to-treat bacteria, including multi-drug-resistant Neisseria gonorrhoeae and Staphylococcus aureus (MRSA). Two distinct strategies were employed: one directed generative AI to craft molecules based on a predicted active chemical fragment, and another allowed algorithms to freely generate molecules without constraints. This process revealed over 36 million potential compounds, which were then screened for their antimicrobial properties.
The most promising candidates feature unique structures not found in existing antibiotics and employ novel mechanisms that disrupt bacterial cell membranes. These innovations could lead to antibiotics that are more resilient to resistance, as drug-resistant bacterial infections are responsible for nearly five million deaths annually. The team is optimistic about applying their approach to discover drugs targeting other bacterial species, showcasing the transformative potential of AI in drug development and broadening access to previously unexplored chemical spaces.
👉 Pročitaj original: MIT Technology Review
