AI Creates 16 Novel Viruses Based on Synthetic Genetic Information

Scientists have reportedly created 16 types of novel viruses that never existed in the world, based on genetic information designed by AI for the first time.

These new viruses were found to actually reproduce.

On the 6th (local time), scientists affiliated with Stanford University and the Arc Institute in California published a paper titled "Generative Design of Bacteriophages Using Genomic Language Models" in the journal Science.

The research team selected a bacteriophage called PhiX174 as the baseline framework for the genome to be designed by the AI model.

A bacteriophage refers to a virus that infects bacteria, uses them as hosts, and multiplies, commonly abbreviated as a phage.

The genomic language models Evo 1 and 2 were used as the AI models.

After learning millions of natural genomes and additionally studying genomic data from PhiX174 and other closely related phages, the models generated candidates that had the potential to become reproducible new phages.

The research team chemically synthesized about 300 candidates deemed to have the highest probability of success and tested them in the laboratory.

As a result, they confirmed that 16 of these phages could actually reproduce and that their genetic information differed from existing phages known to exist in nature.

In an interview with the BBC, corresponding author Professor Brian Hie addressed the possibility of creating full-fledged living organisms through AI-driven genome generation, stating, "It will probably take a lot of effort, but it's not impossible," and adding that the research team is "certainly interested in pushing in that direction."

Professor Patrick Cai, director of the Manchester Institute of Biotechnology at the University of Manchester in the UK, evaluated the study as an "important milestone," noting that "genomic language models are starting to learn the design principles etched into evolution."

He pointed out that this has opened the door to designing genomes using AI.

However, concerns are also being raised that if such technology develops further without restrictions and is abused, it could potentially be used to create biological weapons with lethal toxicity or mass destruction capabilities.

Dr. Tom Inglesby and Dr. Moritz Haenle, researchers at the Johns Hopkins Center for Health Security, assessed in a commentary on the paper that this study "raises urgent biosafety and biosecurity questions."

Dr. Haenle expressed concern that in the future, someone might be able to instruct a genomic language model, "Create a flu genome engineered to be more transmissible or more lethal."

Reported by Jung Da-eun | Video by Kim Min-ji | Graphics by Yang Hye-min | Produced by SBS Digital News
※ Please note: This article was translated by AI and may contain errors.