Artificial intelligence (AI) has been used by a team of researchers in the United States to engineer a complete genome for a new bacteriophage, marking the first reported effort of its kind and a step toward potential AI-designed phage therapy, according to a study published in the journal Science.
Led by Brian Hie, the study's principal investigator and corresponding author, an assistant professor at Stanford University and an innovation researcher at the Arc Institute, the researchers used two genome language models, Evo 1 and Evo 2, to generate complete phage genomes with realistic genetic architectures and specificity to the bacterial host, Escherichia coli C. (E. coli).
This genome marks the first generative design of a complete bacteriophage genome, the researchers say.
Bacteriophages, also known as phages, are viruses that infect bacteria, rather than humans, animals, or plants. The researchers generated and screened thousands of candidate sequences before selecting nearly 300 designs, the study said.
Of these, 285 were successfully synthesized and assembled inside E. coli cells, with 16 of these producing viable bacteriophages capable of reproducing and killing bacteria in laboratory tests, the study said.
Some of the engineered phages outperformed the naturally occurring E. coli-infecting phage, phiX174, in killing bacteria in head-to-head laboratory competition tests. The engineered phage cocktail was also effective against E. coli strains that had developed resistance to phiX174, suggesting the potential for AI-engineered phage therapy, the researchers said.
In a Perspective article accompanying the study, Thomas Inglesby and Moritz Hanke of Johns Hopkins University noted that the advances raise biosafety concerns. They called for mandatory legal screening of synthetic genetic material orders for potentially harmful sequences, rather than relying solely on voluntary safeguards.
