The most popular gene-editing tool, CRISPR-Cas9, generates breaks in the genome that are subsequently repaired by a mix of cellular pathways. Yet, the repair outcomes are not random. Using machine-learning algorithms to analyze large amounts of Cas9-mediated, genome-wide editing events in a range of cells, Shen et al., Allen et al., and Chakrabarti et al. uncovered sequence determinants of repair outcomes and devised rules to predict editing products. These findings provide insights into the repair process and instruct the design of guide RNAs to achieve more precise editing.
Nature 563, 646 (2018). Nat. Biotechnol. 37, 64 (2019). Mol. Cell 10.1016/j.molcel.2018.11.031 (2018).
(원문: 여기를 클릭하세요~)