Stanford scientists use AI to build a virus targeting E. coli

Scientists at Stanford University have achieved a significant breakthrough by using artificial intelligence to design a virus that targets harmful bacteria like E. coli. This marks the creation of the first virus specifically engineered by AI. Researchers employed a generative AI model named EVO 2 to construct a bacteriophage, a type of virus that infects and kills bacteria. Bacteriophages essentially act as the "enemy of my enemy," as Dr. Greg German, a physician and microbiologist at Unity Health Toronto, explained. These viruses have been in a natural battle with bacteria for billions of years, and scientists are now exploring how to harness them for human medical use. The goal is to develop bacteriophages as a new class of antibiotics to fight infections caused by bacteria, such as E. coli, which are increasingly resistant to existing treatments. German is leading an international research trial in Toronto focused on using viruses to treat drug-resistant infections, a $25 million initiative involving 212 women with E. coli-related urinary tract infections. The Stanford team designed 300 novel phages, finding 16 effective against E. coli. Published in the journal Science, the study notes these AI-generated genomes are distinct from natural ones and possess prespecified traits, allowing for targeted action. Notably, the AI-built phage has a genetic code of around 6,500 units, considerably smaller than the 50,000 to 200,000 units found in typical bacteriophages.

“

Bacterial phage eats bacteria, reprograms them to make more phage just like human viruses attack human cells to make more human viruses.