Extended Data Fig. 3: Metrics for rigor and reproducibility of large-scale base editing screens. | Nature Biotechnology

Extended Data Fig. 3: Metrics for rigor and reproducibility of large-scale base editing screens.

From: Mapping variant effects on anti-tumor hallmarks of primary human T cells with base-editing screens

Extended Data Fig. 3

a, Density plots showing LFC values of different categories of guides from the ClinVar library at Day 35 post-electroporation of the long-term expansion screen arm. Dashed line represents the bottom 5% of the distribution of combined empty window and silent mutation controls. Indicated are the percentages of guides in each category falling below this threshold. sgRNAs generating variants in CD3D, CD3E, CD3G, or CD3Z were binned into the ‘CD3 complex’ category. The second donor from the screen is shown (in companion to Fig. 2a). b, Scatter plot showing LFC values of negative control sgRNAs (including both empty window and silent mutations) in both donors from the ClinVar Library at Day 28 post-electroporation in the long-term expansion screen arm. c-d, Distribution of robust rank aggregation (RRA) scores for gene-wise dropout analysis in the c, CD25 hi vs lo (activation) sort and d, CFSE lo vs hi (short-term proliferation) sort arms of the ClinVar library across both donors. The top 5 negatively selected genes in CD25 hi vs lo and in CFSE lo vs hi are listed. e, Shared positive control sgRNA (n = 600) were identified between the ClinVar and 12-gene tiling screens and sgRNA LFCs from matched long-term proliferation arm timepoints (Day 28 of ClinVar Screen, Day 26 of 12-gene Tiling Screen) are plotted. For each screen, the average LFC of each sgRNA across both donors is plotted. Simple linear regression with two-sided Pearson test (panel e).

Back to article page