(a) Comparison of GC content versus indel frequency. Individual targets are denoted by open squares. The solid line is a linear trend line of the data (R2 = 0.0233). (b) Analysis of sgRNA indel efficiency based on binding the non-coding (–, N = 10) or coding (+, N = 32) strand. Box-and-whisker plot denotes median, quartiles, and extremes of the data. (c) Heat map for positions 2 to N of the sgRNA showing the relative increase in indel efficiency for each specific position. N is the first position of the PAM sequence. Color scale represents increased (orange) or decreased (cyan) indel efficiency relative to average. (d) Percentage of indels that caused out-of-frame changes to the target sequence. Individual targets are denoted by open squares. (e) Histogram of deletion positions relative to the NGG site bound by Cas9. Data are included from all targets aligned to the NGG. (f) Distribution of deletion sizes. Data are combined from all targets. The x-axes of e and f are log scales.