Extended Data Fig. 10: Single cell RNA sequencing cell cycle, lineage and pseudotime analysis.
a, Violin plots showing the expression of E2F1 and MYC genes in n = 448 untreated and n = 163 treated cells. Both genes are on average 10 log2 folds downregulated in treated cells. b, Stacked bar plot comparing the inferred lineage of each n = 448 untreated and n = 163 treated single cell with its cell cycle status as determined by the expression of G1/S, G2/M, MYC targets, and E2F targets signatures. c, TSCAN analysis displaying a PCA analysis of the data (n = 611 cells) and inferred pseudotime. The TSCAN package uses a cluster-based minimum spanning tree (MST) approach to order cells based on the gradual transition of their transcriptomes39. A line on the PCA highlights the constructed pseudo-temporal path. The table shows the number of treated and untreated cells within each cluster. The numbers of treated and untreated cells analysed are shown in the histograms at the top of the figure. d, Plot showing the distribution of deep, shallow quiescent and cycling cells (n = 611 cells), and lineage scores, across the TSCAN pseudotime clusters. e, Plots displaying the dynamic expression along the inferred pseudo-temporal path of representative genes with a known role in B-cell differentiation. Each cell (n = 611 cells) is shown as a circle (treated) or a triangle (untreated), the color of which reflects the cluster it belongs to (see panel c). f, Heatmap showing the top variable genes that contribute to clustering of n = 448 untreated and n = 163 chemotherapy-treated cells. PRS cells (n = 10 cells) as defined by this approach are evidenced by black arrows. g, Violin plots showing the expression of mitochondrial genes in treated, PRS and untreated cells as defined by either pseudotime (n = 163, 30 and 418 cells respectively; left plot) or t-SNE analysis (n = 163, 10 and 428 cells respectively; right plot).
