Fig. 5: ZR tumour cells harbouring single dominant LBs form independent lineage trajectories.
From: Dominant clones leverage developmental epigenomic states to drive ependymoma

a, Schematic of PiggyBac TrackerSeq lineage barcoding system and raw single-cell RNA data processing workflow. GFP + IUE cells were isolated and sequenced, and cells with a single LB were used for downstream analysis. b, LB abundance across six IUE biological replicates. Samples IUE-ZR-3 to IUE-ZR-5 included technical replicates. Sample IUE-ZR-1, derived from a P4 mouse brain, showed high diversity in the TrackerSeq library shortly after IUE. c, Integrated UMAP of all 11 IUE TrackerSeq datasets, with tumour cell clusters defined by ZR fusion target gene expression enclosed with a dotted line. d, UMAP visualization highlighting tumour cell clusters with ZR fusion target gene signal. e, Projection of cells harbouring dominant LBs from different samples on to tumour cell clusters in the UMAP, with assessment of barcode distribution across samples. f, UMAP reconstruction performed using tumour cells harbouring dominant LBs. Lineage trajectories were inferred by Slingshot on the basis of Seurat cluster assignments. Arrows denote two distinct differentiation trajectories identified through lineage analysis. g, Bar plot illustrating the relative proportions of tumour cell types derived from the dominant LB (LB-1). h,i, Pseudotime analysis of human ZR (h) and mouse IUE ZR (i) tumour cells showing progression from early-stage progenitor-like cells to later-stage differentiated cells. j, Schematic summarizing the main findings of our study, showing how shifts in cellular differentiation and progenitor programs correspond with changes in PLAG motif activity, providing insights into molecular mechanisms driving tumorigenesis from neurodevelopmental origins.