Plot heatmap of genes with expression change along TSCAN pseudotime

Source: R/runTSCAN.R

A wrapper function which visualizes outputs from the runTSCANDEG function. Plots the top genes that change in expression with increasing pseudotime along the path in the MST. runTSCANDEG has to be run in advance with using the same pathIndex of interest.

plotTSCANPseudotimeHeatmap(
  inSCE,
  pathIndex,
  direction = c("both", "increasing", "decreasing"),
  topN = 50,
  log2fcThreshold = NULL,
  useAssay = NULL,
  featureDisplay = metadata(inSCE)$featureDisplay
)

Arguments

inSCE

Input SingleCellExperiment object.

pathIndex

Path index for which the pseudotime values should be used. Should have being used in runTSCANDEG.

direction

Should we show features with expression increasing or decreeasing along the increase in TSCAN pseudotime? Choices are "both", "increasing" or "decreasing".

topN

An integer. Only to plot this number of top genes along the path in the MST, in terms of FDR value. Use NULL to cancel the top N subscription. Default 30.

log2fcThreshold

Only output DEGs with the absolute values of log2FC larger than this value. Default NULL.

useAssay

A single character to specify a feature expression matrix in assays slot. The expression of top features from here will be visualized. Default NULL use the one used for runTSCANDEG.

featureDisplay

Whether to display feature ID and what ID type to display. Users can set default ID type by setSCTKDisplayRow. NULL will display when number of features to display is less than 60. FALSE for no display. Variable name in rowData to indicate ID type. "rownames" or TRUE for using rownames(inSCE).

Value

A ComplexHeatmap in .ggplot class

Author

Nida Pervaiz

Examples

data("mouseBrainSubsetSCE", package = "singleCellTK")
mouseBrainSubsetSCE <- runTSCAN(inSCE = mouseBrainSubsetSCE,
                                useReducedDim = "PCA_logcounts")
#> Sat Mar 18 10:30:15 2023 ... Running 'scran SNN clustering' with 'louvain' algorithm
#> Sat Mar 18 10:30:15 2023 ...   Identified 2 clusters
#> Sat Mar 18 10:30:15 2023 ... Running TSCAN to estimate pseudotime
#> Sat Mar 18 10:30:16 2023 ...   Clusters involved in path index 2 are: 1, 2
#> Sat Mar 18 10:30:16 2023 ...   Number of estimated paths is 1
terminalNodes <- listTSCANTerminalNodes(mouseBrainSubsetSCE)
mouseBrainSubsetSCE <- runTSCANDEG(inSCE = mouseBrainSubsetSCE,
                                   pathIndex = terminalNodes[1])
plotTSCANPseudotimeHeatmap(mouseBrainSubsetSCE,
                           pathIndex = terminalNodes[1])