plotDendrogram: Plot dendrogram of ClusterExperiment object

In clusterExperiment: Compare Clusterings for Single-Cell Sequencing

Description

Plots the dendrogram saved in a ClusterExperiment object

Usage

## S4 method for signature 'ClusterExperiment'
plotDendrogram(
  x,
  whichClusters = "dendro",
  leafType = c("samples", "clusters"),
  plotType = c("colorblock", "name", "ids"),
  mergeInfo = "none",
  main,
  sub,
  clusterLabelAngle = 45,
  removeOutbranch = TRUE,
  legend = c("side", "below", "none"),
  nodeColors = NULL,
  colData = NULL,
  clusterLegend = NULL,
  ...
)

Arguments

x	a ClusterExperiment object.
whichClusters	argument that can be either numeric or character vector indicating the clusterings to be used. See details of getClusterIndex.
leafType	if "samples" the dendrogram has one leaf per sample, otherwise it has one per cluster.
plotType	one of 'name', 'colorblock' or 'id'. If 'Name' then dendrogram will be plotted, and name of cluster or sample (depending on type of value for leafType) will be plotted next to the leaf of the dendrogram. If 'colorblock', rectangular blocks, corresponding to the color of the cluster will be plotted, along with cluster name legend. If 'id' the internal clusterIds value will be plotted (only appropriate if leafType="clusters").
mergeInfo	What kind of information about merge to plot on dendrogram. If not equal to "none", will replicate the kind of plot that mergeClusters creates, and the input to mergeInfo corresponds to that of plotInfo in mergeClusters.
main	passed to the plot.phylo function to set main title.
sub	passed to the plot.phylo function to set subtitle.
clusterLabelAngle	angle at which label of cluster will be drawn. Only applicable if plotType="colorblock".
removeOutbranch	logical, only applicable if there are missing samples (i.e. equal to -1 or -2), leafType="samples" and the dendrogram for the samples was made by putting missing samples in an outbranch. In which case, if this parameter is TRUE, the outbranch will not be plotted, and if FALSE it will be plotted.
legend	character, only applicable if plotType="colorblock". Passed to phydataplot in ape package that is used to draw the color values of the clusters/samples next to the dendrogram. Options are 'none', 'below', or 'side'. (Note 'none' is only available for 'ape' package >= 4.1-0.6).
nodeColors	named vector of colors to be plotted on a node in the dendrogram (calls nodelabels). Names should match the internal name of the node (the "NodeId" value, see clusterDendrogram).
colData	index (by integer or name) the sample data stored as a DataFrame in colData slot of the object. Only discrete valued ("character" or "factor" variables) will be plotted; indexing of continous variables will be ignored. Whether that data is continuous or not will be determined by the properties of colData (no user input is needed). This argument is only relevant if plotType=="colorblock" and leafType=="samples"
clusterLegend	Assignment of colors to the clusters or sample data (as designated by colData argument) plotted with the dendrogram . If NULL or a particular variable/cluster are not assigned a color, colors will be assigned internally for sample data and pull from the clusterLegend slot of the x for the clusters.
...	arguments passed to the plot.phylo function of ape that plots the dendrogram.

Details

If leafType="clusters", the plotting function will work best if the clusters in the dendrogram correspond to the primary cluster. This is because the function colors the cluster labels based on the colors of the clusterIds of the primaryCluster

Value

A dendrogram is plotted. Returns (invisibly) a list with elements

• plottedObject the phylo object that is plotted.

• originalObject the phylo object before adjusting the node/tip labels.

See Also

mergeClusters,plot.phylo, nodelabels,tiplabels

Examples

data(simData)

#create a clustering, for 8 clusters (truth was 3) 
cl <-clusterSingle(simData, subsample=FALSE, 
sequential=FALSE, 
mainClusterArgs=list(clusterFunction="pam", clusterArgs=list(k=8)))

#create dendrogram of clusters and then 
# merge clusters based ondendrogram: 
cl <- makeDendrogram(cl) 
cl <- mergeClusters(cl,mergeMethod="adjP",DEMethod="limma",
   cutoff=0.1,plot=FALSE) 
plotDendrogram(cl) 
plotDendrogram(cl,leafType="samples",whichClusters="all",plotType="colorblock")

clusterExperiment documentation built on Feb. 11, 2021, 2 a.m.