Skip to main content
Fig. 2 | Genome Biology

Fig. 2

From: scDART: integrating unmatched scRNA-seq and scATAC-seq data and learning cross-modality relationship simultaneously

Fig. 2

The results of scDART and baseline methods on the SNARE-seq mouse neonatal brain cortex dataset. a Latent embedding of scDART, Liger, Seurat, and UnionCom, visualized using PCA. Cells are colored with cell type annotation in the original paper. Red lines show the inferred trajectory backbone. All plots share the same legend as in the Liger plot. b Latent embedding of scDART, Seurat, Liger, and UnionCom, where cells are colored with data batches. All plots share the same legend as in the UnionCom plot. c Neighborhood overlap score. d Pseudotime consistency score, where Pearson correlation is used. e Pearson correlation between real-scRNA-seq and (y-axis) pseudo-scRNA-seq (predicted scRNA-seq from scATAC-seq data by scDART), (x-axis) linear transformation. 11 key DE genes are shown

Back to article page

Genome Biology

ISSN: 1474-760X

Contact us