Extended Data Fig. 5: Comparison of CaRPool-seq to alternative combinatorial perturbation approaches.
From: Efficient combinatorial targeting of RNA transcripts in single cells with Cas13 RNA Perturb-seq
a) Density plots showing the CD46-APC, CD55-FITC, and CD71-PE flow cytometry signal upon Cas9-nuclease mediated knockout (KO) and CRISPRi-mediated (KRAB-dCas9, KRAB-dCas9-MeCP2) knockdown with three alternative sgRNAs from established genome-wide KO20 and CRISPRi21 libraries. Vertical lines mark the threshold (2nd percentile of combined NT conditions) for CD-protein negative cells, indicating the percent negative cells for one replicate experiment. Single guide RNAs with the highest percentage of negative cells (bold) were selected for direct capture Perturb-seq experiments (NA = sgRNA not assayed). b) Cloning strategy for triple sgRNA plasmid vectors. Dual sgRNA constructs were cloned as described before 6. The third sgRNA was cloned behind a bovine U6 promoter using an alternative sgRNA scaffold tested before 6. c) Cell surface protein expression (log2-normalized UMI counts) of CD46, CD55 and CD71 in cells assigned with indicated (s)gRNAs. CaRPool-CITE-seq (n = 4,979 cells) and Perturb-seq experiments using Cas9-nuclease (n = 2,270), KRAB-dCas9 (n = 2,104) or KRAB-dCas9-MeCP2 (n = 2,326). d) Contour plots of data shown in (c). e) Protein level ADT-based clustering of single-cell expression profiles of merged CaRPool-CITE-seq (n = 6,986 cells) and Perturb-seq experiments using Cas9-nuclease (n = 2,836), KRAB-dCas9 (n = 2,911) or KRAB-dCas9-MeCP2 (n = 3,038) effector proteins as in Fig. 3e. Cells are labelled by the assigned target gene combination based on detected bcgRNA or sgRNAs and split by Perturb-seq.
