Gene editing, particularly through CRISPR technology, has shown immense potential in solving challenges in various fields such as agriculture, biotechnology, and human health. However, there are limitations when using CRISPR-based gene editors in non-native environments like human cells. Artificial intelligence (AI) offers an alternative approach by using large language models (LLMs) trained on biological diversity to design optimal gene editors. In this study, researchers curated a massive dataset of CRISPR operons and used AI to design a programmable gene editor capable of precision editing in the human genome. The AI-generated gene editor, named OpenCRISPR-1, exhibited improved activity and specificity compared to the prototypical gene editing effector, SpCas9. It was also compatible with base editing. OpenCRISPR-1 is now publicly available for ethical and widespread use in research and commercial applications.
| Signal | Change | 10y horizon | Driving force |
|---|---|---|---|
| Gene editing with AI | From traditional gene editing to AI-assisted gene editing | Precision editing of the human genome with optimal properties | Bypassing evolutionary constraints and generating gene editors with AI |