Fig. 2: The effect of mutating TF binding sites and protein phosphorylation sites on cellular fitness determined by lineage tracing of editing events.

a, The effect of mutating protein phosphorylation sites of CDK1 on fitness of HAP1 cells. log₂(day 8/day 2) is shown for each sequence tag pair with read count >5 on day 2 after calculating the ratio of read counts for mutated/original sequences at both timepoints. The CGE method was used for measuring the effect of Y15F mutation on the fitness also after introducing this mutation to HAP1 cells using prime editing³³. In a–d, dots represent individual cell lineages harboring a unique barcode—that is, internal replicates for which median (red line) and P value are calculated (two-sided Wilcoxon signed-rank test separately for each experiment, no multiple comparison adjustments; see Supplementary Table 3 for statistical details and Supplementary Table 4 for sequencing depth and editing efficiency). b, The effect of mutating MYC binding motifs (E-box) at promoters of MYC target genes on fitness of HAP1 cells (see also Extended Data Fig. 4)—synonymous mutation in the MYC coding region as a negative control. log₂(day 8/day 2) is shown for each sequence tag pair with two flanking mutations and read count >50 on day 2 after calculating the ratio of read counts for mutated/original sequences at both timepoints (see also Supplementary Table 5). c, The effect of E-box mutation on MYC occupancy and H3K27ac at promoters of MYC target genes. log₂(IP sample/input) is shown for each sequence tag pair with two flanking mutations and read count >100 in the input after calculating the ratio of read counts for mutated/original sequences. Genome browser snapshots with ChIP-seq and ATAC-seq tracks demonstrate robust MYC binding to the targeted sites in wild-type HAP1 cells. d, Reproducibility of the CGE method shown for the E-box at the PPAT promoter from two independent experiments (Exp 1 and Exp 2) and from two internal replicate groups (IR1 and IR2) (Methods). e, The key advantages of the CGE method are high statistical power due to internal replicates and mitigation of the confounding effects characteristic of CRISPR–Cas9-based methods by excluding the unedited cells.