Fig. 5: Robustness tests and batch effect correction. | Nature Communications

Fig. 5: Robustness tests and batch effect correction.

From: scCASE: accurate and interpretable enhancement for single-cell chromatin accessibility sequencing data

Fig. 5

a–c Robustness of different methods in various scenarios. a Robustness of different methods to the noise level. b Robustness of different methods to the cell type imbalance degrees. c Robustness of different methods to the dataset size. d The values of different metrics of raw data and the data enhanced by various methods on the Mixed-protocols dataset. UMAP visualizations of (e) raw data and (f) the data enhanced by scCASE on the Mixed-protocols dataset. Cell type annotation labels, batch labels and clustering labels are projected onto the visualizations. Source data are provided as a Source Data file.

Back to article page