STopover captures spatial colocalization and interaction in the tumor microenvironment using topological analysis in spatial transcriptomics data

Abstract

Unraveling the spatial configuration of the tumor microenvironment (TME) is key to understanding tumor-immune interactions to translate them into immuno-oncology. With the advent of spatially resolved transcriptomics (SRT), the TME could be dissected for whole cell types across numerous RNAs. We suggest a novel approach, STopover, which performs topological analysis to compute the colocalization patterns between cell types and map the location of cell-cell interactions. While gradually lowering the threshold for the feature, the connected components (CCs) were extracted based on the spatial distance between the unit tissue region and the persistence of the CCs. Local and global Jaccard indices were calculated between the CCs of a feature pair to measure the extent of spatial overlap. The STopover was applied to various lung cancer data obtained from SRT platforms, both barcode and image-based SRT, and could explain the infiltration patterns of immune and stromal cells in the TME. Moreover, the method predicted the top cell-cell communication based on the ligand-receptor database and highlighted the main region of the interaction. STopover is a tool to decipher spatial interaction in the tissue and shed light on the pathophysiology underlying the microenvironment.

Spatial Transcriptomics Reveals Spatially Diverse Cancer-Associated Fibroblast in Lung Squamous Cell Carcinoma Linked to Tumor Progression

“…Using CellDART 35 with the reference scRNA-seq data, we mapped the spatial distribution of cell types within the TME, with a focus on epithelial cells and fibroblasts ( Figure 6a ; complete cell type mapping results are available in Supplementary Figure 6). To uncover molecular patterns in regions where fibroblasts and epithelial cells spatially coincide, we analyzed topologically overlapping cell type patterns in the barcode-based ST data 36 . This analysis highlighted regions with a high density of specific cell types and their intersections with other cell populations ( Figure 6b ).…”

“…To uncover molecular patterns in regions where fibroblasts and epithelial cells spatially coincide, we analyzed topologically overlapping cell type patterns in the barcode-based ST data 36 . This analysis highlighted regions with a high density of specific cell types and their intersections with other cell populations (Figure 6b).…”

“…This analysis involved 25 configuring CellDART with specific parameters: 80,000 pseudospots, 3,000 iterations, and 10 mixed cells per spot 35 . Subsequently, we utilized all pairs of cell type scores from these spots for topological colocalization analysis via STopover, applying default settings to assess the topological overlap patterns for each cell type within the barcode-based ST data 36 .…”