Extended Data Fig. 3: Results from Tiled edits self-targeting sensor screen demonstrate precise editing with low errors across 1 bp substitution edits.

a, Percentages of sequencing reads from sensor targets containing only the precise edit from two replicates of Tiled edits screen performed in PEmaxKO cells and collected at day 5 (left), day 8 (middle), and day 10 (right) post transduction of sensor library. Each data point represents an individual epegRNA-target pair, with best epegRNA design per edit indicated in green, and additional controls indicated in red (negative) and cyan (positive). Correlation between replicates (Pearson’s r) indicated. Inset plots show percentages of sequencing reads from sensor targets containing errors for replicate 1 (x-axis) and replicate 2 (y-axis) from respective timepoint. b, Replicate-averaged percentages of sequencing reads with only the precise edit (green), any error (red), or unedited sequence (blue) for each epegRNA design targeting the +10 T > C edit at target site 1 in Tiled edits screen. Results from day 10 post transduction of sensor library. Bars are stacked per epegRNA to show cumulative total of sequencing reads. Individual data points for replicates provided in Supplementary Table 2. c, Replicate-averaged percentages of sequencing reads containing only the precise edit or errors for epegRNA-target pairs from the Tiled edits screen for the best epegRNA per edit (maximum precise editing) on day 10 post transduction of sensor library. Boxplots indicate the median and interquartile range (IQR) for each group with whiskers extending 1.5 x IQR past the upper and lower quartiles. Left, all edits (n = 195 edits). Right, all edits except A > G substitutions, which may occur during ADAR-based editing of the epegRNA-target constructs during lentiviral packaging⁵³ (n = 167 edits). d, Maximum precise editing frequencies (replicate-averaged) across epegRNA designs for each edit in the Tiled edits screen, separated by target site. x-axis indicates the edit position relative to the nick. y-axis specifies the targeted edit. na indicates an unedited nucleotide at that position, which was not measured in the screen.