Fig. 5: Alzheimer’s disease data set (GSE138852) analysis based on scGNN.
From: scGNN is a novel graph neural network framework for single-cell RNA-Seq analyses
a Cell clustering UMAP. Labeled with scGNN clusters (left) and AD/control samples (right). b Comparison of cell proportions in AD/control samples (left) and each cluster (right). c Heatmap of DEGs (logFC > 0.25) in each cluster. Six oligodendrocyte sub-clusters are merged as one to compare with other cell types. Marker genes identified in DEGs are listed on the right. d Selected AD-related enrichment pathways in each cell type in the comparison between AD and control cells. e Underlying TFs are responsible for the cell-type-specific gene regulations identified by IRIS3.
