Fig. 4: ResST corrects batch effects for vertical and horizontal data integration of multiple tissue sections. | Communications Biology

Fig. 4: ResST corrects batch effects for vertical and horizontal data integration of multiple tissue sections.

From: A graph self-supervised residual learning framework for domain identification and data integration of spatial transcriptomics

Fig. 4

a H&E image of posterior mouse brain from the original study. b The spatial clustering results from Method 2 (with MDD), Method 1 (without MDD), Harmony, and Scanorama (from left to right). c UMAP plots before batch effects correction and spatial location of sections 151673 and 151675 in DLPFC. d Vertical integration results with two DLPFC sections. UMAP plots after batch effects correction and spatial clustering from ResST, Harmony, and Scanorama (left). Spots are colored according to the spatial domains identified by the respective clustering methods (right). e Horizontal integration results from ResST (top) and SpaGCN (bottom) with two mouse brain samples which consist of sagittal anterior and posterior regions.

Back to article page