AI-guided mutations make viruses infect bacteria up to 1 million times better

Biochemists at the University of Wisconsin–Madison built an AI model that selects mutations in bacteriophages to fight antibiotic-resistant bacteria. Published in Cell Systems, the work could speed up development of alternatives to traditional antibiotics.

• AI model trained on lab data on phage-bacteria interactions
• Selected mutations boosted infection efficiency up to 1 million times
• Findings published in the journal Cell Systems
• Goal is alternatives to antibiotics against resistant infections