According to an IT Home report citing the BBC, Stanford researchers fine-tuned the Evo1 and Evo2 genome language models to generate complete bacteriophage genomes of roughly 5,400 base pairs. They selected and synthesized 302 AI-generated candidates; 16 reportedly infected and killed E. coli, a yield of about 5.3%. The researchers describe this as the first successful AI design of complete, functional viral genomes. The result could expand the design space for phage therapy and antimicrobial-resistance research, although therapeutic efficacy, host specificity, safety, and the primary study’s review status remain unclear from the supplied report.
No heat snapshots are available in the last 24 hours.