AI designs 16 synthetic viruses to combat superbugs

In a significant leap for biotechnology, scientists have for the first time employed artificial intelligence to design and synthesize 16 entirely new viruses. Researchers from Stanford University and the Arc Institute utilized genomic language models, named Evo 1 and Evo 2, to generate complete, functional viral genomes capable of replication. This marks a paradigm shift, demonstrating AI's power to write novel genetic code.

The newly created organisms are bacteriophages, viruses that exclusively infect and destroy specific bacteria. In this study, they were programmed to target Escherichia coli (E. coli). Crucially, these synthetic phages lack the ability to harm human cells, making them safe for therapeutic applications. The AI models were trained on vast datasets of genetic material from viruses, bacteria, plants, and humans, learning the rules of genetic sequences to create viable, unprecedented viral architectures.

From thousands of theoretical designs, scientists selected the most promising 302 and synthesized them in the lab. When tested against E. coli, 16 of these AI-designed viruses proved exceptionally effective at infecting and eradicating bacterial colonies. This breakthrough holds direct implications for the global public health crisis of antibiotic resistance. In lab trials, a combination of these 16 phages successfully overcame the defenses of E. coli strains resistant to natural bacteriophages, showcasing AI's potential to generate adaptive biological therapies against evolving pathogens.