Fig. 1

The 225 off-target modification frequencies for 26 guide RNAs separated by number of mismatches. To indicate the optimal depth for an off-target search, validated off-target modification frequencies are shown on the x-axis, separated by number of mismatches from their guide sequence (rows along the y-axis). The studies are indicated by symbols, explained in the legend of the graph [3, 7, 26, 28, 29, 33, 38]. The row label specifies the number of mismatches, followed by a line showing the total number of off-targets predicted by CRISPOR for the 22 guide sequences (“genome hits”). The third line indicates the number of validated off-targets and the percentage of total validated off-targets that they represent. For example, at six mismatches, about 1.9 million potential off-targets were found in the genome for the 26 guide sequences, three of which were shown to be bona fide off-targets that were experimentally validated. The four off-targets with six mismatches make up only 1.8 % of all off-targets, so 98.2 % of the 225 off-targets differ by up to five mismatches. The off-targets with five and six mismatches make up 11.7 % of all off-targets