A new method that combines AI and gene editing improves the precision of gene therapies.

A scientific team has developed a new DNA editing method that combines cutting-edge genetic engineering with artificial intelligence (AI). This new technique paves the way for more precise modeling of human diseases and lays the groundwork for next-generation gene therapies.

Precise and selective DNA editing through small point mutations or gene integration using CRISPR/Cas has great potential for applications in biotechnology and gene therapy.

But while it's essential that these edits don't cause unwanted genetic changes and that they maintain genomic integrity to avoid side effects, DNA repair sometimes causes genetic errors. Now, a team of scientists from the University of Zurich (UZH), Ghent University in Belgium, and ETH Zurich has developed a new method that significantly improves the precision of genome editing.

They did this using an AI tool called Pythia , which predicts how cells repair their DNA after it's been cut by gene-editing tools like CRISPR/Cas9. "Our team developed tiny DNA repair templates , which act like molecular glue and guide the cell to make precise genetic changes," says Thomas Naert, a pioneer of this technology at UZH and a postdoctoral researcher at Ghent University.

These AI-designed templates were first tested in human cell cultures , where they enabled highly precise genetic edits and integrations, and validated in other organisms , such as Xenopus, a small tropical frog used in biomedical research, and in live mice, where the researchers successfully edited the DNA of brain cells.

Learning and predicting DNA repair patterns

"DNA repair follows patterns; it's not random. And Pythia uses these patterns to our advantage," Naert notes. Traditionally, when CRISPR cuts DNA, scientists rely on the cell's natural repair mechanisms to fix the break. But while these repairs follow predictable patterns, they can lead to unwanted outcomes, such as the destruction of surrounding genes.

"What we've modeled on a large scale is that this DNA repair process obeys consistent rules that AI can learn and predict ," he says. Furthermore, the method can also be used to fluorescently label specific proteins.

"This is incredibly powerful because it allows us to directly observe what individual proteins do in healthy and diseased tissues," he says. Another advantage of the new method is that it works well in all cells , even in organs without cell division, such as the brain, according to the study.

The Oracle of Delphi of Gene Editing

Pythia is named after the high priestess of the oracle at the Temple of Apollo at Delphi in ancient times, who was consulted to predict the future. Similarly, this new tool allows scientists to predict the results of gene editing with remarkable accuracy.

"Just as meteorologists use AI to predict the weather, we use it to forecast how cells will respond to genetic interventions. That kind of predictive capability is essential if we want gene editing to be safe , reliable, and clinically useful," says Soeren Lienkamp of UZH and ETH Zurich, lead author of the study.

"What excites us most is not just the technology itself, but also the possibilities it opens up. Pythia combines the large-scale prediction of AI with real-life biological systems . From cultured cells to whole animals, this close connection between modeling and experimentation is increasingly useful, for example, in precision gene therapies," adds Lienkamp.

This work opens up new possibilities for understanding genetic diseases and developing gene therapies, including those for neurological diseases, that are safer and more effective.