Fig. 5: Distinct Gene Signatures Define Unique Cell Types in Black American (BA) and White American (WA) Tumor Niches.

a Overview of the approach used to identify niche-associated gene signatures. For BA- and WA-associated communities (BA-community 1 and WA-community 1), extended signature genes (ESGs) were derived based on co-expression within the query (i.e., BA-community 1 or WA-community 1 genes), co-localized spots. ESGs were computed per sample and compared between BA and WA, with genes significantly elevated in BA (i.e., BA > WA for BA-niche) or in WA (i.e., WA > BA for WA-niche) selected for further analysis. Each query generated a distinct BA-niche and WA-niche, composing of ST spots uniquely expressing BA-community 1 or WA-community 1 genes. b Representative heat map showing cell-type-specific expression of ESGs associated with the BA niche defined by expansion of BA-community 1. These ESGs correspond to various cell types, including Cancer-Associated Fibroblasts (CAF), endothelial cells, and myeloid cells, while the BA-niche is poor in T-cells and neutrophils. c Same as in (b), but for the heat map showing cell-type-specific expression of ESGs associated with the WA niche defined by expansion of WA-community 1. ESGs in this niche predominantly correspond to exhausted T-cells and neutrophils, while the WA-niche is poor in CAFs and endothelial cells. d Average quantification of the number of co-localized Visium spots per sample across 10 BA (blue) and 10 WA (red) TNBC samples corresponding to BA- and WA-associated niches. Statistical significance was computed using a T-test. e, f Gene set enrichment analysis (GSEA) of ESGs associated with the BA-niche (e) and WA-niche (f), with -log10 P-values shown. P-values were derived from a 1-sided hypergeometric test, adjusted using g:SCS method (g:Profiler). Source data are provided as a Source Data file.