Washington, Aug. 10 (SANA)U.S. researchers have usedartificial intelligenceto design microscopic viruses capable of attacking and killingbacteria, a development that could contribute to new approaches for treating antibiotic-resistant infections.

Euronews reported on Saturday that a study by researchers fromStanford University and the Arc Instituteshowed that generative AI models can be used to design complete genomes of bacteriophages,virusesthat specifically infect bacteria.

The researchers produced 16 synthetic bacteriophages that successfully infected Escherichia coli (E. coli) and inhibited its growth under laboratory conditions.

The team used genome language models, including Evo 1 and Evo 2, to generate new genetic sequences inspired by the ΦX174 bacteriophage, which is known to target E. coli. Candidate designs were then synthesized and tested in laboratory experiments.

The study found that some AI-designed bacteriophages had characteristics different from naturally occurring phages. A combination of synthetic phages was also able to target E. coli strains that had shown resistance to natural bacteriophages, highlighting their potential role in future phage-based therapies.

The researchers said the current experiments focused only on bacteriophages that target bacteria and included safeguards aimed at reducing possible misuse of the models. They described the work as an early step toward using AI to design biological systems for therapeutic applications.

The development comes as antibiotic resistance continues to reduce the effectiveness of medicines used to treat bacterial infections, increasing interest in alternative treatment strategies.