Fig. 4: Integration of mouse brain scRNA-seq and scATAC-seq datasets. | Nature Communications

Fig. 4: Integration of mouse brain scRNA-seq and scATAC-seq datasets.

From: scMODAL: a general deep learning framework for comprehensive single-cell multi-omics data alignment with feature links

Fig. 4

a UMAP plots of unintegrated scRNA-seq and scATAC-seq datasets, colored by cell types. b UMAP plots of integrated embeddings produced by scMODAL, colored by modalities (left) and cell types (right). c Correlation heatmap of Louvain clusters computed based on the mean gene expression or gene activity profiles of the clusters. The dendrogram was computed with Scanpy80 to show the hierachical clustering result of the Louvain clusters. d UMAP plots of integrated embeddings produced by scMODAL, colored by Louvain cluster labels. Mouse brain cortical layer marker gene expression levels in the RNA modality (e) and gene activity scores in the ATAC modality (f) in Louvain clusters 1, 3 and 7. Fam107a expression levels in the RNA modality, activity scores in the ATAC modality and imputed expression levels in the ATAC modality shown in UMAP (g) and heatmap (h). i Gene-peak links inferred using scMODAL's imputation result. Source data are provided as a Source Data file.

Back to article page