Sungkyunkwan University Research Team Successfully Advances AI-Designed Base Editors, Paving the Way for Next-Generation Gene Therapy

A joint research team led by Sungkyunkwan University has developed OpenABE, an advanced gene-editing tool that addresses the accuracy issues found in earlier AI-designed models. By using structure-guided protein engineering based on AlphaFold predictions, researchers boosted gene-editing efficiency by up to 36 times. This brings the performance of AI-generated editors in line with current laboratory gold standards.

The new OpenABE 1.1 and 1.2 models resolve major hurdles in gene therapy, specifically the frequent off-target and bystander effects that previously made AI-designed tools risky for clinical use. These editors demonstrate high precision by targeting intended DNA sequences without altering nearby genetic material. This breakthrough allows for consistent results in both nuclear DNA and the more complex mitochondrial DNA.

Beyond basic editing capability, the team integrated these tools into engineered virus-like particles for cellular delivery. This delivery system ensures that researchers can selectively target specific genes for treatment. The combination of high-efficiency editing and safer delivery mechanisms creates a viable platform for addressing previously intractable genetic diseases.

This work expands upon previous advancements from the same group, including the development of OpenCRISPR-1, which significantly lowered off-target mutations compared to standard Cas9. By combining structure-guided engineering with AI predictions, the researchers have created a repeatable process for designing high-performance gene therapy tools. The results were published in a peer-reviewed international genetics journal, signaling a shift toward more reliable, AI-supported therapeutic development.