Fig. 2

Cell hashing and genomic barcoding both facilitate effective sample multiplexing. (a,b) tSNE plots showing distribution of cells based on HTO (Cell Hashing Oligonucleotide; time course dataset; left) and genomic barcode reads (signalling perturbation dataset; right). (c,d) UMAPs showing distribution of cells in each timepoint, treatment, and barcoding line after quality control filtering (see Table 3) in the time course (c) and signalling perturbation (d) datasets. Further details about assignment of treatment groups to each sequencing library, timepoint, and barcode are shown in Table 1. (e) Pearson correlation between average gene expression of cells in each individual sample in the signalling perturbation dataset.