Fig. 5: scNiche deciphers the cellular compositional and phenotypic heterogeneity between immune niches from human TBNC dataset.

a The proportion of immune, stromal, and tumor cells in each niche (top) and the enrichment scores of each cell type in each niche (bottom) across 19 mixed subtype samples and 15 compartmentalized subtype samples. The enrichment scores are calculated among all 13 niches, and the p-value is calculated with the one-sided Mann-Whitney U test (*p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001; ****p ≤ 0.0001). b The two spatially adjacent cell niches are enriched with immune cells from different lineages respectively, corresponding to different immune microenvironments. Each cell is colored by the niche labels and the cell type labels, and the proportions of CD4 T cells and other immune cells within the neighbors of each cell are displayed. c The proportions of Niche 6 and Niche 12 in each sample. Only samples with a proportion of Niche 6 (or Niche 12) exceeding 5% are shown. The average expression levels of monocyte markers, immunoregulatory proteins, and antigen presentation proteins of macrophages in Niche 6 and Niche 12 in each sample (d) and the proportions of macrophages, immune cells, and tumor cells in Niche 6 and Niche 12 in each sample (e). For Niche 6 (or Niche 12), only samples with a proportion of that niche exceeding 5% are considered. The p-value is calculated with the two-sided Mann-Whitney U test. Data are presented as boxplots (minima, 25th percentile, median, 75th percentile, and maxima). n = 6 (10) samples for Niche 6 (Niche 12). f Niche 6 in Patient 9 (top) and Patient 10 (bottom) is located at the tumor-immune border. g The spatial expression level of CD11b, CD11c, IDO, and PD-L1 in Patient 9 (top) and Patient 10 (bottom). Source data are provided as a Source Data file.