Efficient C•G-to-G•C base editors developed using CRISPRi screens, target-library analysis, and machine learning
Key Finding
CRISPRi screen of 476 DNA repair genes identified factors affecting C-to-G editing. Generated suite of 10 engineered CGBEs with machine learning models predicting editing outcomes (R=0.90). Enables correction of 546 disease SNVs with >90% precision and up to 70% efficiency.