Predicting base editing outcomes with an attention-based deep learning algorithm trained on high-throughput target library screens
This work results from a collaboration with the research group of Prof. Dr. Gerald Schwank at the University of Zurich.
Base editors consist of a DNA-targeting CRISPR-Cas module and a single-stranded DNA deaminase. These chimeric ribonucleoprotein complexes enable transition of C•G into T•A base pairs and vice versa on genomic DNA. So far, base editors have not been widely used in research or therapy, due to broad variation in their editing efficiencies on different genomic loci. To overcome this hurdle, this study performed an extensive analysis of A- and C base editors on a library of 28,294 lentivirally integrated genetic sequences to establish BE-DICT, an attention-based deep learning algorithm capable of predicting base editing outcomes with high accuracy. The BE-DICT tool can in principle be trained on any novel base editor variant, facilitating the application of base editing for research and therapy.