Fig. 1. Fibroblast identification through single-cell RNA-sequencing analysis of whole-tissue homogenates derived from human NSCLC tumour samples.

a Schematic illustrating sample processing and analysis methodology used to generate the target lung drop-seq (TLDS) dataset, comprised of human control (n = 6) and NSCLC (n = 12) samples. The Figure was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license. b Barplots showing key demographics of the TLDS dataset, further details provided in Supplementary Data 1. c. 2D visualisation (UMAP dimensionality reduction) of pooled data from all samples’ whole-tissue scRNA-seq data, highlighting different cell types. Further analysis is shown in Supplementary Fig. 1a–d. AT1 alveolar type 1 cells, Mono/Mac monocytes and macrophages. d Feature plots showing the expression of canonical markers for mural cells (MCAM and RGS5), mural cells and myofibroblasts (ACTA2) and fibroblasts (DPT), in the stromal cell cluster (subset from panel c). e Volcano plot showing genes differentially expressed (Bonferroni adj.P < 0.01 and absolute logFC >0.5, shown in red) between mural cells and fibroblasts from a recently published human lung cell atlas¹⁸ (HLCA). Genes included in the consensus gene signatures identified for mural cells or fibroblasts are labelled. f Scatter plot showing the classification of fibroblasts and mural cells based on consensus gene signatures (defined in e), applied to stromal cells from the HLCA dataset¹⁸. g Barplots showing the results of the gene signature classification results compared to the original HLCA publication’s cell type annotation. h Feature plot showing the expression of consensus fibroblast marker gene signature in the TLDS dataset. i Feature plot showing the expression of consensus mural marker gene signature in the TLDS dataset. j UMAP plot showing the result of fibroblast and mural cell demarcation in the TLDS dataset using the consensus gene signature approach.