Skip to main content
Fig. 4 | BMC Bioinformatics

Fig. 4

From: Joint analysis of scATAC-seq datasets using epiConv

Fig. 4

Aligning co-assay (SNARE-seq) data of mouse cerebral cortex to scATAC-seq (sci-ATAC-seq and dscATAC-seq) references increases the resolution of chromatin profiles. a Low dimensional embedding and clustering of co-assay data based on scRNA-seq profiles. b Joint embedding of scATAC-seq from co-assay data and scATAC-seq references. c The Adjusted Rank Index (ARI) and LISI of different methods. d Result of LSI + Seurat in details. LSI + Seurat groups cells with mixed identities together. e Joint embedding of scATAC-seq from co-assay data and scATAC-seq references, cells are colored by manually annotated joint clusters. Ex, excitatory neuron; In, inhibitory neuron; Oligo, oligodendrocyte; OPC, oligodendrocyte progenitor cell; Ast, astrocyte; Mic, microglia. f Heatmap of chromatin markers and their nearest genes. Left: chromatin profiles of sci-ATAC-seq. Middle: chromatin profiles of SNARE-seq. Right: gene expression profiles of SNARE-seq. Selected marker genes with highest fold changes are shown in the right panel. g Aggregated ATAC-seq signals near the TSS of OPC maker gene Vcan. h Left (scATAC-seq) and middle (co-assay): The fold changes of enrichment between ATAC-seq defined marker regions and RNA-seq defined associated regions. Right: the number of detected markers from scATAC-seq references and co-assay data for each cluster.

Back to article page