Supplementary Figure 6: Azimuth score and PAM identity are correlated with guide efficacy
From: Multiplexed precision genome editing with trackable genomic barcodes in yeast
(a) The Azimuth score for each guide in our dataset (x-axis) is plotted against log2-fold change after editing (y-axis) (Spearman rho -0.18, Pearson R = -0.19, both p < 2.2E-16). The trend line derives from a linear regression of logFC to Azimuth score (R2 = 0.037, p<2.2E-16). The color intensity depicts the count of barcodes per box (N = 23,866 barcodes). (b) Violin plots of the effect of PAM on log2-fold change after editing (black dots depict medians and black lines indicate the 25th and 75th percentiles). The number of barcodes (N) for each PAM type is indicated in the above plots. A one-sided Wilcoxon test was used for between-group comparisons; for each group, location shift and 99% confidence intervals (in square brackets) are as follows: TGG_AGG: 0.12 [0.05, Inf], p = 6.478E-05; TGG_CGG: 0.19 [0.1, Inf], p = 7.72E-07; TGG_GGG: 0.31 [0.22, Inf], p = 6.744E-16. Dead guides are shown for comparison.
