Fig. 5: MIA encompasses β-cells heterogeneity across and within biological conditions.
From: Delineating mouse β-cell identity during lifetime and in diabetes with a single cell atlas
a, Coarse β-cell states labeled based on sample metadata (excluding low-quality clusters) shown as a UMAP. b, Expression of known markers (marker groups are specified on the top of the plot), quality control metrics and sex ratios across coarse β-cell states displayed in separate dot-plot panels. In the marker expression panel, the dot size indicates the fraction of cells expressing a gene, whereas in other panels it is set to a fixed size. c, Expression of MIA-based markers of coarse β-cell states. d, Overview of the method used for extraction of GPs and subsequent cell clustering resolution selection or definition of consistently variable GPs across samples. e, Fine β-cell states defined based on the presence of a unique combination of GPs (excluding low-quality clusters) shown as a UMAP. f, Expression of known β-cell heterogeneity markers across fine β-cell states. Phenotypes associated with individual genes (top). The dotted boxes represent two distinct sets of maturity (orange) and dedifferentiation or diabetes markers (red); the solid cyan box shows overlap and expression similarity between maturity, immune-attack susceptibility and extreme insulin producer markers. g, Correlation between gene groups variable in all healthy samples and known β-cell heterogeneity markers on the healthy β-cell subset. Markers present within a specific gene group are annotated with an X. imm., immature; M, male, F, female; NOD-D, NOD diabetic; D.-inter, diabetic intermediate; insL/H, insulin low/high; str., stressed. In b, c and f, relative expression is computed as the average of cell groups normalized to [0,1] for each gene feature.
