Fig. 1: Schematic overview.
From: SOTIP is a versatile method for microenvironment modeling with spatial omics data
a–h Overview of SOTIP. SOTIP takes 2D/3D spatial omics data (spatially resolved gene expression, protein or metabolic profiles) as input (a). Next, the cell (spot) type clustering is obtained solely by the expression matrix. Then the histogram of cell clusters is obtained combing the spatial information and clustering result (b). The manifold distance is approximated by combing clustering result and the expression matrix (c). The ME graph is built using optimal transport between histograms characterizing microenvironments (MEs) (d). Finally, multiple tasks can be performed based on the MECN graph, including spatial heterogeneity quantification (e), spatial domain identification (f), differential microenvironment analysis (g), and other downstream tasks (h). i The supported spatial data.
