AlphaFold3-based contact modelling enables precise DNA base editing

Researchers have developed a new method for precise DNA base editing by integrating AlphaFold3-based contact modelling, according to a study published in Nature. The approach enables more accurate targeting of single-nucleotide changes in the genome, which could improve gene therapy and genetic research.

The technique affects scientists and clinicians working on genome editing, particularly those using base editors to correct disease-causing mutations. By predicting protein-DNA interactions more reliably, the method reduces off-target edits, making treatments safer and more effective.

This matters because current base editing tools can inadvertently modify unintended DNA sites, limiting their therapeutic potential. The new modelling approach enhances precision, potentially accelerating the development of gene therapies for genetic disorders.

The study, published on 12 February 2025, involved researchers from multiple institutions. They used AlphaFold3 to model the structure of base editor proteins bound to DNA, predicting contact points that determine editing accuracy. The method achieved a 90% success rate in targeting specific bases in human cells, compared to 70% with conventional design.

Earlier work on base editing, first reported in 2016, has faced challenges with off-target effects. This new modelling step builds on AlphaFold's protein structure predictions, which won the 2024 Nobel Prize in Chemistry.

The next steps include testing the method in animal models and eventually in human clinical trials. The researchers also plan to adapt the approach for other genome editing tools, such as prime editors, to expand its applications.