Fig. 2: Benchmarking and application of vertical integration using SpatialMETA.

a Schematic diagram illustrating the vertical integration process of SpatialMETA. b Summary of benchmarking metrics assessing the vertical integration performance of different tools on ST and SM data. The tools are arranged in descending order based on their overall scores. Source data are provided as a Source Data file. c–e Spatial plots comparing the original (upper left and middle panels) and denoised (lower left and middle panels) gene expression (left panels) and metabolite intensity (middle panels) data processed by SpatialMETA. Bar plots display the Pearson correlation coefficient (PCC) and cosine similarity between the original and denoised data for each method (right panels). Spatial plots visualizing Leiden clustering results derived from different tools for ccRCC (f), GBM (g), and mouse brain (h) datasets. Bar plots depicting spatial continuity scores for ccRCC (i), GBM (j), and mouse brain (k) datasets across tools, bars are color-coded by method. Stacked bar plots showing marker scores for ST (green) and SM (yellow) data across different tools for RCC (l), GBM (m), and mouse brain (n) samples. Imm immune clusters, Endo endothelial clusters, Stro stromal clusters, Mal malignant clusters. Note that the color legend is placed at the bottom of the figure, and the colors of clusters do not directly correspond to the same captured structures across different methods.