Researchers at Stanford University have successfully used artificial intelligence to design brand new viruses with fully functional genomes, a development described as a significant milestone in synthetic biology. The AI models, named Evo1 and Evo2, were trained on genetic codes from viruses, bacteria, plants and people. They operate on principles similar to large language models such as ChatGPT, but instead of predicting text sequences, they predict genetic sequences.
The Stanford team generated 302 candidate designs and synthesized them in the laboratory. Of these, 16 proved effective and were able to replicate and function as bacteriophages—viruses that infect specific bacterial species. Specifically, the phage successfully killed E. coli bacteria. The researchers implemented multiple safety measures during the work, including excluding viruses capable of infecting complex organisms from the training database and conducting all experiments in a secure laboratory setting using only bacteriophages rather than viruses affecting humans.
Scientists in the field have hailed the achievement as a transformative development for biotechnology. Experts note the technology could lead to new treatments for antibiotic-resistant infections and unlock possibilities for developing new drugs and therapies to address major health challenges. However, the breakthrough has simultaneously triggered concerns from the scientific community regarding biosafety and biosecurity. Commentators from Johns Hopkins University’s Center for Health Security emphasized that the capability to design viral genomes raises urgent questions about preventing misuse and ensuring the technology is not applied to create harmful pathogens.
The researchers acknowledge the distinction between designing viral genomes and creating living organisms. The bacteriophages used in this study contain approximately 5,400 base pairs of genetic code, compared to around 500,000 base pairs for the smallest living cell genome and 3 billion for the human genome. Researchers have expressed interest in exploring whether the technology could eventually be applied to simpler organisms, though they note such endeavors would require substantial additional effort.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI