Predicting off-target effects for end-to-end CRISPR guide design

Abstract

To enable more effective guide design we have developed the first machine learning-based approach to assess CRISPR/Cas9 off-target effects. Our approach consistently and substantially outperformed the state-of the-art over multiple, independent data sets, yielding up to a 6-fold improvement in accuracy. Because of the large computational demands of the task, we also developed a cloud-based service for end-to-end guide design which incorporates our previously reported on-target model, Azimuth, as well as our new off-target model, Elevation (https://www.microsoft.com/en-us/research/project/crispr).