7.5 Visualize the MICA output

scMINER provides a function, MICAplot() to easily visualize the clustering results on a 2D plot, UMAP or tSNE. And it can be colored by multiple variables, including cluster label, sample source, nUMI, nFeature, pctMito and more.

7.5.1 Color-coded by cluster labels

library(ggplot2)
MICAplot(input_eset = pbmc14k_log2cpm.eset, color_by = "clusterID", X = "UMAP_1", Y = "UMAP_2", point.size = 0.1, fontsize.cluster_label = 6)

7.5.2 Color-coded by true label of cell types

MICAplot(input_eset = pbmc14k_log2cpm.eset, color_by = "trueLabel", X = "UMAP_1", Y = "UMAP_2", point.size = 0.1, fontsize.cluster_label = 4)

7.5.3 Color-coded by nUMI, for QC purpose

MICAplot(input_eset = pbmc14k_log2cpm.eset, color_by = "nUMI", do.logTransform = TRUE, point.size = 0.1)

## The values in "nUMI" have been transformed by log2(value + 1). To turn transformation off, set do.logTransform = FALSE.

7.5.4 Color-coded by nFeature, for QC purpose

MICAplot(input_eset = pbmc14k_log2cpm.eset, color_by = "nFeature", do.logTransform = TRUE, point.size = 0.1)

## The values in "nFeature" have been transformed by log2(value + 1). To turn transformation off, set do.logTransform = FALSE.

### Color-coded by pctMito, for QC purpose

MICAplot(input_eset = pbmc14k_log2cpm.eset, color_by = "pctMito", do.logTransform = FALSE, point.size = 0.1)