Researchers at MIT have made a significant breakthrough in the battle against antibiotic resistance by using artificial intelligence to develop new antibiotics. This innovative approach aims to combat two harmful drug-resistant bacteria: Neisseria gonorrhoeae, which causes gonorrhea, and methicillin-resistant Staphylococcus aureus (MRSA), often responsible for severe infections.
The announcement comes at a time when the pharmaceutical landscape remains bleak. The FDA has approved only a few new antibiotics in the last 45 years, and existing methods of drug discovery are struggling to keep up with the alarming rise in bacterial resistance. According to estimates, nearly 5 million people die each year due to drug-resistant infections.
MIT’s Antibiotics-AI Project is changing the way researchers approach antibiotic development. Instead of relying on libraries of existing chemicals, the team has turned to generative AI, exploring vast and untapped chemical spaces to find potential new compounds. Led by James Collins, a professor at MIT, the project generated over 36 million potential compounds through computation and then utilized machine-learning models to select those with antibacterial properties, safety for human use, and originality.
“We’re excited about the new possibilities that this project opens up for antibiotics development,” Collins stated, highlighting the fresh avenues this research creates.
The team employed two different strategies for each bacterium. For N. gonorrhoeae, they used a fragment-based approach, starting with a library of 45 million chemical fragments. They focused on a particular fragment known as F1, which was expanded into complete molecules using advanced algorithms. These efforts produced two candidates, with NG1 showcasing remarkable effectiveness in laboratory and animal models.
In the case of MRSA, the researchers allowed the AI algorithms to generate new compounds without a predefined starting point. This free exploration led to the creation of 29 million compounds, from which six demonstrated strong activity. One standout, DN1, proved effective in eliminating MRSA skin infections in mice.
The work at MIT has the potential to redefine antibiotic treatment, addressing resistance issues in a fundamentally different manner. Nonprofit Phare Bio, part of this project, is further refining the candidates with hopes of moving to preclinical trials. As Collins noted, this research paves the way for a new era of antibiotics, offering hope against the dire threat of superbugs.
"*" indicates required fields
