AnnotationScatterPlot: Visualiation of a custom annotation over nonlinear...

In ILoReg: ILoReg: a tool for high-resolution cell population identification from scRNA-Seq data

Description

The AnnotationScatterPlot enables visualizing arbitrary class labels over the nonliner dimensionality reduction, e.g. t-SNE or UMAP.

Usage

AnnotationScatterPlot.SingleCellExperiment(
  object,
  annotation,
  return.plot,
  dim.reduction.type,
  point.size,
  show.legend
)

## S4 method for signature 'SingleCellExperiment'
AnnotationScatterPlot(
  object,
  annotation = NULL,
  return.plot = FALSE,
  dim.reduction.type = "",
  point.size = 0.7,
  show.legend = FALSE
)

Arguments

object	of SingleCellExperiment class
annotation	a character vector, factor or numeric for the class labels.
return.plot	return.plot whether to return the ggplot2 object or just draw it. Default is FALSE.
dim.reduction.type	"tsne" or "umap". Default is tsne.
point.size	point size. Default is 0.7.
show.legend	a logical denoting whether to show the legend on the right side of the plot. Default is TRUE.

Value

ggplot2 object if return.plot=TRUE

Examples

library(SingleCellExperiment)
sce <- SingleCellExperiment(assays = list(logcounts = pbmc3k_500))
sce <- PrepareILoReg(sce)
## These settings are just to accelerate the example, use the defaults.
sce <- RunParallelICP(sce,L=2,threads=1,C=0.1,k=5,r=1)
sce <- RunPCA(sce,p=5)
sce <- RunTSNE(sce)
sce <- HierarchicalClustering(sce)
sce <- SelectKClusters(sce,K=5)
## Change the names to the first five alphabets and Visualize the annotation.
custom_annotation <- plyr::mapvalues(metadata(sce)$iloreg$clustering.manual,
                                     c(1,2,3,4,5),
                                     LETTERS[1:5])
AnnotationScatterPlot(sce,
                      annotation = custom_annotation,
                      return.plot = FALSE,
                      dim.reduction.type = "tsne",
                      show.legend = FALSE)

ILoReg documentation built on Nov. 8, 2020, 8:20 p.m.