Regina Barzilay and Jim Collins win Pritzker Prize for AI and Science Research Excellence for work in AI-enabled antibiotic discovery
Regina Barzilay and Jim Collins received the 2026 Pritzker Prize for AI in Science Research Excellence for pioneering AI‑enabled antibiotic discovery, notably the deep‑learning‑driven identification of the broad‑spectrum antibiotic halicin (trained on 2,500 molecules) and the narrow‑spectrum antibiotic abaucin in 2023. Their work compressed a traditionally years‑long discovery cycle into weeks, demonstrating that generative AI can move from a discovery aid to a design tool for new drugs. This breakthrough signals to engineers that AI models can rapidly screen vast chemical spaces to pinpoint viable candidates against drug‑resistant pathogens, accelerating pipelines in biomedical AI applications.
⚡ Key Takeaways
- Deep learning model trained on 2,500 molecules identified halicin, a broad‑spectrum antibiotic effective against Clostridium difficile, Acinetobacter baumannii, and Mycobacterium tuberculosis.
- The same AI approach later identified abaucin in 2023, a narrow‑spectrum antibiotic targeting A. baumannii.
- AI reduced the antibiotic discovery timeline from years to a few weeks, highlighting a major performance‑time trade‑off.
- Generative AI is shifting from a discovery aid to a design tool, enabling targeted synthesis of novel compounds.
- Realistic AI models of human biology are still needed; current systems only automate parts of the pipeline.
Engineers can leverage AI‑driven chemical screening to dramatically shorten drug‑discovery cycles, delivering effective antibiotics faster and at lower cost, which is critical for combating antimicrobial resistance.
✅ Practical Steps
- Apply deep‑learning models trained on curated molecular datasets (e.g., 2,500‑molecule training set) to screen for antibacterial activity.
- Validate top AI‑predicted candidates experimentally against target pathogens such as A. baumannii or C. difficile.
- Incorporate generative AI techniques to design narrow‑spectrum compounds that minimize off‑target effects.
Want the full story? Read the original article.
Read on MIT News AI ↗