Fig. 4.

STMVGAE can accurately identify spatial domains and SVGs on the BRCA dataset. (A) Manual annotating of BRCA dataset based on pathological features. (B) Bar plots of STMVGAE and five baseline methods on the BRCA dataset with three different evaluation matrices. (C) Domain identification on the BRCA dataset by STMVGAE, SEDR, STAGATE, DeepST, SpaGCN, and stLearn. (D) UMAP visualization and PAGA trajectory on the BRCA dataset by STMVGAE, SEDR, STAGATE, DeepST, SpaGCN, and stLearn. (E) STMVGAE Spatial expression patterns of SVG detected on different spatial domains in the BRCA dataset.