Fig. 1: NE features and tumor spatial locations were associated with immune cell abundance and interactions.

a Schematic diagram of the ZJCC scRNA-seq study design. b Projection of the data onto the Chan et al. reference dataset for precise cell type classification on Uniform Manifold Approximation and Projection (UMAP). Different colors represent sample sources. c Annotations of SCLC cells of different molecular subtypes (SCLC-A, SCLC-N, SCLC-P) and non-malignant cells. d Standard PAGA graph representing the topological information of the main immune cell types and spatial locations. e The network of main immune subsets from different tissues. Colors of dots in the network represent different sources. The size of the dots represents the abundance of each cell type. The weight of the lines represents the closeness of the interaction between each pair of immune subsets. f Crosstalk between cancer cells and immune cells via receptors and ligands identified by CellPhoneDB. Line weight represents the probability of cellular crosstalk. g Cell types projected with PCoA into Euclidean distance based on their potential CCI interaction. h Schematic of the 10x Flex FFPE sample scRNA-seq study design. i UMAP showing projection of the 10x Flex data with cell type classification.