Deep Learning Models To Predict The Editing Efficiencies And Outcomes Of Diverse Base Editors

This article describes a study that compared the base editing and nuclease activities of Cas9 and BE variants. To do this, cell lines were generated that expressed the variants, and a high-throughput approach involving pairwise libraries of sgRNA-encoding and target sequences was used to measure the activities of BEs and Cas9 nucleases. The results showed that the levels of most Cas9 and BE variant proteins were comparable, except for one variant which showed statistically significant higher protein levels.