Fig. 3

Data distributions before and after rscreenorm. Lethality scores’ and rscreenorm scores’ empirical distributions of data corresponding to multiple independent genetic screens. Upper panels show data of genome-wide siRNA screens, lower panels data of CRISPR-Cas screens data by Hart et al. [17]. Lines in all panels represent density values for library features, per replicate. a Log2-transformed viabilities corrected for plate effects, multiplied by -1 to facilitate visual comparison with lethality and rscreenorm scores. b Lethality scores prior to rscreenorm. c Rscreenorm scores. d Hyperbolic-arc sine transformed cell counts per guide RNA, multiplied by -1 to facilitate visual comparison with lethality and rscreenorm scores. e Lethality scores prior to rscreenorm. f Rscreenorm scores