SpatialCompassV (SCOMV)
SCOMV is a computational tool designed to quantify spatial distribution patterns at both the gene and cell levels relative to tumor regions, for characterizing the spatial relationships between tumors and surrounding cells.
Using SCOMV, the spatial distribution patterns were categorized into four representative types:
Internal, primarily expressed within the tumor.
Peripheral, circumferentially distributed around the tumor.
Partially peripheral, restricted to specific segments of the tumor boundary.
Ubiquitous, broadly distributed across the tissue.
It represents the spatial distribution patterns of cells and genes relative to a user-defined reference region (e.g., tumor regions) using vector-based features derived from their spatial relationships. Reference regions are delineated using SpatialKnifeY (SKNY).
By integrating information across multiple fields of view, SCOMV further identifies spatially differential genes (spatially DEGs) that exhibit region-specific spatial reorganization beyond conventional differential expression analysis.
Input
cell_feature_matrix.h5
An HDF5 file containing the gene expression count matrix generated by the sequencing pipeline. The matrix stores the number of transcripts detected for each gene (features) in each cell or spot. It also includes metadata such as feature IDs, feature names, barcodes, and matrix indices in a compressed sparse format.
cells.csv.gz
A compressed CSV file containing metadata for each cell or spot. This file typically includes cell barcodes and associated attributes such as spatial coordinates, quality metrics, or annotation labels produced by the processing pipeline.
Overview
