Researchers from Stanford University and the Arc Institute used generative AI to design complete viral genomes from scratch, and 16 out of 302 AI-generated bacteriophage candidates were confirmed to be fully functional and capable of infecting E. coli bacteria. Several of the AI-designed variants showed replication advantages of up to 65 times compared to the natural template they were modeled on. The team, led by Brian Hie and involving collaborators from NVIDIA and UC Berkeley, used genome language models called Evo 1 and Evo 2 trained on trillions of DNA base pairs. These AI-generated phages were also tested in cocktail therapies against resistant bacterial strains and showed significant efficacy. The research raises biosecurity questions, as the capability to design functional viral genomes from scratch could warrant careful governance conversations before the technology advances further.
Source: Read the original article
