BioTechniques News
Beatrice Bowlby
In a world-first, scientists have used genome language models to generate new viruses that selectively target bacteria, raising hopes for new medicines.
For the first time ever, Stanford University (CA, USA) researchers have designed hundreds of complete bacteriophage genomes using generative AI models. After testing the effectiveness of these phages against Escherichia coli, the team reports 16 novel viruses with potent bacteria-killing capabilities, potentially ushering in a new era of AI-guided therapeutic design.
Using AI to create complex biological systems has the potential to transform biotechnology. For example, genome language models, trained not on text but instead on millions of genomes from all domains of life, can learn evolutionary patterns that shape DNA in nature and use this to generate entirely new sequences. While progress has been made at the scale of individual genes and gene circuits, whole-genome design has thus far eluded scientists.
Bacteriophages are considered a promising jumping off point thanks to their small size, tractability and broad applications in molecular biology, microbial engineering and medicine. The team behind the latest breakthrough used the phage ΦX174 as a template for the genome language models Evo 1 and Evo 2, which then wrote novel, complete phage genomes with specificity against E. coli.
These were computationally evaluated, and almost 300 of the most promising designs were chemically synthesized and tested in laboratory conditions, ultimately yielding 16 viable viruses with diverse sequences, structures and fitness profiles. The phages are genetically distinct from any known natural phages and capable of infecting and destroying E. coli.
Fungal gene editing world-first points to promising cancer medicines
Unraveling some of the mysteries of fungi with a novel gene-editing tool, researchers have identified a number of molecules with promising anti-cancer properties.
When the researchers tested a cocktail of their new bacteriophages on E. coli that had evolved resistance to ΦX174, they quickly overcame the pathogens, whereas a mixture of naturally occurring ΦX174-like phages failed to do the same.
Meanwhile, cryo-electron microscopy revealed that one of the phages uses an evolutionarily distant DNA packaging protein in its capsid, which may help to explain its success.
Overall, the work is an intriguing proof-of-concept for AI-guided biological design on the whole-genome level and demonstrates a potential path toward new therapies against rapidly evolving pathogens.
“This is an important milestone for synthetic genomics. For the first time, we are seeing AI move beyond predicting biological sequences to generating entire functional genomes that work in the laboratory. While these are relatively small bacteriophage genomes, the significance extends far beyond phages. It suggests that genome language models are beginning to learn the design principles encoded by evolution, opening the door to AI-assisted genome writing,” Patrick Cai from the Manchester Institute of Biotechnology (UK), who was not involved in the study, told the Science Media Centre (London, UK). “The next challenge is to scale these approaches to much larger and more complex genomes while ensuring that every computational prediction is matched by rigorous experimental validation.”
Although important questions around biosafety and regulation still need to be addressed, the team is excited about what the future holds for their groundbreaking approach. “New doors in science are now open because of what we can do with these models,” study author Samuel King concluded.
The post World-first AI-designed genomes generate 16 powerful new bacteria-killing viruses appeared first on BioTechniques.
Powered by WPeMatico