Stanford's Evo 2 AI Model Generates Phages to Combat Drug-Resistant E. coli
Researchers at Stanford have synthesized nearly 300 phages using DNA sequences generated by the Evo 2 AI model, narrowing them down to 16 strains capable of killing E. coli.
Stock photo for illustration only, not from the actual event
- Stanford researchers used the Evo 2 AI model to synthesize nearly 300 viral genomes.
- Laboratory tests identified 16 phages with exceptionally strong E. coli-killing activity.
- A computational framework helped cut synthesis costs by prioritizing viable candidates.
- The team released Evo 2 as open-source software and plans to target more complex DNA.
Researchers at Stanford University have successfully synthesized nearly 300 phages derived from DNA sequences produced by the Evo 2 generative artificial intelligence model. Following rigorous laboratory testing, the team narrowed the group down to 16 phages that demonstrated particularly powerful E. coli-killing activity.
The research centers on bacteriophage ΦX174, pronounced "FYE-ex-1-7-4". Brian Hie, an assistant professor of chemical engineering and a Dieter Schwarz Foundation Stanford Data Science Faculty Fellow, created Evo 2, with bioengineering graduate student Samuel King leading the experimental work detailed in the paper.
Evo 2 generates entirely new DNA sequences starting from a small snippet of a phage genome. The researchers instructed the model to produce a complete ΦX174 genome in a single left-to-right pass.
Stock photo for illustration only, not from the actual event
Brian Hie described the generation process, stating, "In this case, we wanted the model to generate the entire genome end-to-end in a single left-to-right pass. We didn’t add anything." This procedure produced thousands of candidate genomes before the team selected specific sequences for chemical synthesis and laboratory evaluation.
ΦX174 served as a relatively compact test system because its genome contains fewer than 6,000 base pairs, compared to the roughly 3 billion base pairs found in the human genome. Even with a compact 5,400-character sequence, Hie noted that interpreting DNA gene-by-gene remains a formidable task.
"If the bacteria gains resistance to a single phage, it’s game over for the medication. But if you have multiple genetically distinct phages in a mixture, it would be harder for the bacteria to develop resistance to the entire cocktail."
The project investigates whether an AI model can create entire, viable viral genomes rather than simply proposing localized DNA edits. Samuel King developed a computational framework to assess traits from ΦX174 and related phages, filtering down the candidate pool before chemical synthesis.
The application of generative AI to bacteriophage design represents a crucial advancement in combating antimicrobial resistance (AMR). Utilizing multi-phage cocktails mimics natural predator-prey dynamics and complicates the ability of bacteria to evolve resistance compared to standard single-target antibiotics.
Stanford reported that a cocktail combining the 16 selected phages rapidly overcame resistance in an E. coli strain that was immune to native ΦX174. Furthermore, Hie suggested that similar methodology could target methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa, a leading cause of hospital-acquired infections.
Hie has made Evo 2 available as open-source software, allowing researchers globally to download and utilize the model for genome design. While the release has prompted security discussions, Hie argues that existing natural pathogens pose greater risks and that AI tools can aid in defending against biological threats.
Stock photo for illustration only, not from the actual event
Source: AI News
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment