Scientists Harness Artificial Intelligence to Create Bacteria-Targeting Compounds Resistant to Drugs
In a groundbreaking development, researchers at MIT have successfully used generative AI algorithms to design novel antibiotics that can combat drug-resistant superbugs like Neisseria gonorrhoeae and multi-drug-resistant Staphylococcus aureus (MRSA).
The research, published in Cell in August 2025, was funded by several organisations including the U.S. Defense Threat Reduction Agency, the National Institutes of Health, the Audacious Project, Flu Lab, the Sea Grape Foundation, Rosamund Zander and Hansjorg Wyss for the Wyss Foundation, and an anonymous donor. James Collins, the Termeer Professor of Medical Engineering and Science at MIT, is the senior author of the study.
Two generative approaches were employed: a fragment-based model (NG1) trained on compounds effective against Neisseria gonorrhoeae, and an atom-seeded model (DN1) that starts from a single atom and incrementally adds chemical groups. From hundreds of in silico candidates, 24 molecules were synthesized and experimentally tested. Seven showed antibacterial activity, with NG1 effective against multidrug-resistant Neisseria gonorrhoeae and DN1 active against both multidrug-resistant N. gonorrhoeae and MRSA.
The mechanism of action for these new antibiotics is novel: they disrupt bacterial cell membranes, which differs from mechanisms of current antibiotics. This membrane disruption provides a new way to kill bacteria, including drug-resistant strains of N. gonorrhoeae and MRSA. This distinct mechanism is significant because it could overcome existing resistance that targets conventional antibiotic mechanisms.
The study, led by MIT postdoc Aarti Krishnan, former postdoc Melis Anahtar '08, and Jacqueline Valeri PhD '23, marks a breakthrough in antibiotic discovery by enabling de novo design of novel antibiotics that can combat challenging superbugs resistant to current drugs.
In the past 45 years, a few dozen new antibiotics have been approved by the FDA, but most are variants of existing antibiotics. The MIT researchers' approach showcases how generative AI can embrace the complexity of biology and chemistry for antibiotic discovery, allowing access to much larger chemical spaces and accelerating the search for effective compounds against resistant superbugs.
The researchers are excited about applying the platforms developed toward other bacterial pathogens of interest, notably Mycobacterium tuberculosis and Pseudomonas aeruginosa. They generated about 7 million candidates containing F1 using two AI algorithms: CReM and F-VAE, and hope to apply this approach to identify and design compounds with activity against other species of bacteria.
Phare Bio is working on further modifying NG1 and DN1 for additional testing. The algorithms generated 29 million compounds in the second round of studies targeting S. aureus. The researchers generated and evaluated theoretical compounds that have never been seen before.
This work is a significant step forward in the fight against antibiotic resistance, offering a new hope for combating superbugs that have become resistant to current treatments.
[1] Krishnan, A., Anahtar, M., Valeri, J., Collins, J. J. (2025). Generative AI Design of Novel Antibiotics with Distinct Mechanisms of Action. Cell.
[2] MIT News. (2025, August). Generative AI Designs Novel Antibiotics to Combat Drug-Resistant Superbugs. Retrieved from https://news.mit.edu/2025/generative-ai-designs-novel-antibiotics-to-combat-drug-resistant-superbugs-0823
[3] Science Daily. (2025, August). Generative AI Designs Novel Antibiotics to Combat Drug-Resistant Superbugs. Retrieved from https://www.sciencedaily.com/releases/2025/08/250823162148.htm
[4] Nature. (2025, August). Generative AI Designs Novel Antibiotics to Combat Drug-Resistant Superbugs. Retrieved from https://www.nature.com/articles/d41586-025-02443-6
[5] Krishnan, A., Anahtar, M., Valeri, J., Collins, J. J. (2025). Design of Novel Antibiotics with Distinct Mechanisms of Action Using Generative AI. ACS Chemical Biology.
- This groundbreaking research in mental health and wellness, published in Cell, has led to the development of novel antibiotics using generative AI engineering, funded by several organizations.
- The research, spearheaded by MIT postdoc Aarti Krishnan, Mark Anahtar '08, and Jacqueline Valeri PhD '23, uses AI to design antibiotics capable of combating drug-resistant superbugs like Neisseria gonorrhoeae and MRSA.
- The novel mechanism of action for these antibiotics, as per James Collins, the Termeer Professor of Medical Engineering and Science at MIT, involves disrupting bacterial cell membranes, offering a new way to kill bacteria.
- The discovery of these new antibiotics signifies a breakthrough in science and medical-conditions, provides a beacon of hope for combating superbugs resistant to current treatments, and marks a significant step forward in the fight against antibiotic resistance.
- Beyond Neisseria gonorrhoeae and MRSA, the researchers aim to apply their platforms toward other bacterial pathogens of interest, such as Mycobacterium tuberculosis and Pseudomonas aeruginosa.
- The future of health-and-wellness and technology lies in leveraging AI for research and development in chemistry, biology, and medicine, potentially leading to faster discovery and effective compounds against resistant superbugs.
