Skip to main content
. 2021 Nov 25;22(23):12755. doi: 10.3390/ijms222312755

Figure 2.

Figure 2

Assignment of cell line type to clusters generated with Seurat [18], implemented in rCASC [19]. (A) Five clusters were generated with Seurat (resolution = 0.1), using 2500 genes selected as the most variant within the 5000 most expressed in the RNA-5c experiment. (B) Four clusters were generated with Seurat (resolution = 0.1) using RNA-3c experiment, selecting the same 2500 genes used in (A). (C) RNA-5c hierarchical clustering (Euclidean distance, average linkage) of log2CPM clusters’ pseudo-bulk expression, row-mean centered, combined with the log2TPM, row-mean centered, for CCLE lung cell lines A449, NCIH838, NCIH2228, NCIH1975 and HCC827. (D) RNA-3c hierarchical clustering same parameters used in (C). In the RNA-5c experiment, hierarchical clustering of the clusters’ pseudo-bulks, combined with the expression data retrieved for the bulk RNAseq for the five cell lines (C) well correlate to each other. Only in the RNA-3c experiment, cluster X1 (D) shares similarity with more than one cell line (NCIH1975 and HCC827).