AlphaFold AI used to redesign gene-editing proteins, reducing off-target effects

Addressing Gene-Editing Safety

Gene-editing therapies are emerging, but a significant challenge has been ensuring their safety. Gene-editing systems can sometimes make unintended edits to the wrong DNA sequences, known as off-target effects, because the human genome contains many similar sequences.

Even if the probability of an off-target edit is low, the large number of cells typically edited in therapies makes such errors inevitable. Minimizing these errors is crucial for the widespread adoption and safety of gene-editing treatments.

AlphaFold's Role in Redesign

In a recent study published in Nature, researchers described using the AI protein-folding software AlphaFold to address this issue. They modified AlphaFold to pinpoint specific areas within gene-editing proteins that contribute to off-target effects.

Once these problematic regions were identified, the researchers altered them to reduce the occurrence of unintended edits, thereby enhancing the precision of gene-editing systems.

Components of Gene Editing

Gene-editing systems typically consist of three main components. The first is guide RNA, which pairs with the target genomic sequence. Selecting guide RNA sequences that have minimal similarity to other genomic locations is a standard practice to improve specificity.

The second component is a Cas protein, such as Cas9 from the CRISPR system. Cas proteins interact with both the guide RNA and genomic DNA, enforcing the specificity of the interaction. Improved Cas family members have been developed to reduce the tendency for off-target edits.