Fig. 1

The impact of multiple on-targets on sgRNA log-fold changes. a Guide numbers as a function of the number of perfect alignments (on-target alignments); non-targeting controls (NTCs) were excluded from this analysis. b Guide-level log-fold changes (LFCs) averaged across cell lines as a function of the number of on-targets. The number of on-targets was calculated as the number of perfect alignments between the reference genome and the 20-nt spacer; we excluded guides with single-mismatch alignments to prevent the confounding effect of single-mismatch off-targets. c Combined effect of multiple on-targets on the CERES score. For each gene, we calculated the maximum number of targeted loci (x-axis) as the maximum number of perfect alignments for a guide designed to target that gene in the Avana library. d, e Effects of multiple on-target alignments on LFCs in the breast cancer cell line HMC-1-8 and in the lymphoma cell line SR-786. Solid lines represent second-degree polynomial fits (see “Material and methods”). f Distribution of cell-specific fitted on-target activity (average log-fold changes) as a function of the number of on-target alignments (solid line: median across cell lines; shaded area: full range of on-target activity across cell lines). g Average on-target toxicity as a function of Cas9 activity score