Automated high-throughput genome editing platform with an AI learning in situ prediction model

A great number of cell disease models with pathogenic single-nucleotide variants (SNVs) are needed for genome-editing therapeutics and basic research, but traditional generation of these models is manual, time-consuming, costly, and error-prone. The authors devise an automated high-throughput platform that can automatically edit thousands of samples within a week, yielding edited cells with high efficiency. Leveraging the large in situ genomic dataset generated by the platform, they introduce the Chromatin Accessibility Enabled Learning Model (CAELM), which incorporates both chromatin accessibility and sequence context to predict cytosine base editor (CBE) performance and accurately forecasts in situ base editing outcomes. This work is expected to accelerate development of base editor-based genetic therapies.