Deep sequencing analysis

Deep sequencing was performed using the MiSeq system (Illumina, San Diego CA). Specific primer pairs (Supplementary Table ³) were designed, each amplifying 100–200 bp fragments spanning the target site or one of the predicted top 10 putative off-target sites (Supplementary Table ³), and each incorporating adaptor sequences for the MiSeq processing. Briefly, genomic DNA samples were initially amplified using primers listed in Supplementary Table ³, then a nested adaptor PCR was run using the primers listed in the footnote to Supplementary Table ³. Following amplifications, PCR products were barcoded by running a 14-cycle barcode PCR. The barcoded PCR products were purified using a PCR purification kit (Qiagen), pooled and sequenced using the Illumina MiSeq platform. A custom-written computer script was used to extract “high-quality sequence reads”, based on their alignment with the wild-type template sequence. Sequences that contained a deletion or insertion resulting in shorter or longer amplicons which involve any base in the ZFN cleavage site were classified as an NHEJ-mediated deletion or insertion, respectively; together, such mutations are classified as “indels”. Sequences with the expected changes (3 substitutions, 1 to wild-type and 2 for novel BsaWI site (Fig. 1a) were classified as “corrected”.