Fig. 6: Differentiation trajectory reconstruction of HBECs and MTECs single-cell transcriptomic datasets.
From: Heterogeneity-preserving discriminative feature selection for disease-specific subtype discovery
a, b, c, d UMAP visualizations (N = 2970 cells) using pre-annotated markers and PHet’s selected features and their corresponding SPRING plots, respectively. Coloring represents the previously annotated cell types. The trajectories are visually represented by red-colored arrows. e A SPRING plot using PHet’s features displaying two distinct trajectories: Donor 1 (black) and Donors 2 and 3 (red). f A bar plot representing the relative abundance of cell types grouped by donors. g SPRING plots of the selected top features predicted by PHet. The color gradient from black to green indicated cells enriched with the corresponding feature. h, i, j, k UMAP visualizations (N = 14163 cells) using pre-annotated markers and PHet’s selected features and their corresponding SPRING plots, respectively. Coloring represents the previously annotated cell types. The trajectories are visually represented by red-colored arrows. l Two distinct clusters are displayed in the SPRING plot, with one cluster (injured) located at the top and another (uninjured) at the bottom. m The relative abundance of cells between these clusters is shown as a bar plot. Pre-annotated markers and PHet’s selected features of HBECs and MTECs are provided as a Source Data file.
