Particle.news

Download on the App Store

MIT Uses Generative AI to Design Two Novel Antibiotics That Beat Superbugs in Mice

Human trials remain years away despite AI's success in creating novel antibacterial scaffolds.

AI designs new antibiotics to combat drug-resistant superbugs
AI, drug discovery, pharmaceuticals, MRSA

Overview

  • Peer‑reviewed results in Cell show AI‑designed compounds, DN1 and NG1, killed MRSA and drug‑resistant Neisseria gonorrhoeae in lab assays and cleared infections in mouse models.
  • Researchers computationally explored tens of millions of molecules, narrowing to synthesizable candidates while filtering out designs resembling existing antibiotics or predicted to be toxic.
  • For MRSA, more than 29 million structures were screened and DN1 emerged as the lead; for gonorrhoea, only two of the top 80 AI designs could be synthesized and NG1 proved active.
  • The nonprofit Phare Bio is conducting medicinal chemistry to optimize both leads, with refinement and safety work expected before any human testing can begin.
  • Scientists caution that complex synthesis, manufacturing hurdles, stewardship needs, and weak commercial incentives will require coordinated investment despite the public‑health urgency of rising resistance.