Skip to contents

Make dot plot of gene expression in cell groups

Source: R/dotplot.R
plotClusterGeneDot.Rd

This function produces dot plots. Each column represent a group of cells specified by groupBy, each row is a gene specified by features. The color of dots reflects mean of normalized expression of specified genes in each cell group and sizes reflects the percentage of cells expressing each gene in a group. We utilize ComplexHeatmap for simplified management of adding annotation and slicing subplots. This was inspired by the implementation in scCustomize.

Usage

plotClusterGeneDot(
  object,
  features,
  groupBy = NULL,
  splitBy = NULL,
  featureScaleFunc = function(x) log2(10000 * x + 1),
  cellIdx = NULL,
  legendColorTitle = "Mean\nExpression",
  legendSizeTitle = "Percent\nExpressed",
  viridisOption = "magma",
  verbose = FALSE,
  ...
)

Arguments

object

A liger object

features

Use a character vector of gene names to make plain dot plot like a heatmap. Use a data.frame where the first column is gene names and second column is a grouping variable (e.g. subset runMarkerDEG output)

groupBy

The names of the columns in cellMeta slot storing categorical variables. Expression data would be aggregated basing on these, together with splitBy. Default uses default clusters.

splitBy

The names of the columns in cellMeta slot storing categorical variables. Dotplot panel splitting would be based on these. Default NULL.

featureScaleFunc

A function object applied to normalized data for scaling the value for better visualization. Default function(x) log2(10000*x + 1)

cellIdx

Valid cell subscription. See subsetLiger. Default NULL for using all cells.

legendColorTitle

Title for colorbar legend. Default "Mean\nExpression".

legendSizeTitle

Title for size legend. Default "Percent\nExpressed"

viridisOption

Name of available viridis palette. See viridis. Default "magma".

verbose

Logical. Whether to show progress information. Mainly when subsetting data. Default FALSE.

...

Additional theme setting arguments passed to .complexHeatmapDotPlot and heatmap setting arguments passed to Heatmap. See Details.

Value

HeatmapList object.

Details

For ..., please notice that arguments colorMat, sizeMat, featureAnnDF, cellSplitVar, cellLabels and viridisOption from .complexHeatmapDotPlot are already occupied by this function internally. A lot of arguments from Heatmap have also been occupied: matrix, name, heatmap_legend_param, rect_gp, col, layer_fun, km, border, border_gp, column_gap, row_gap, cluster_row_slices, cluster_rows, row_title_gp, row_names_gp, row_split, row_labels, cluster_column_slices, cluster_columns, column_split, column_title_gp, column_title, column_labels, column_names_gp, top_annotation.

Examples

# Use character vector of genes
features <- varFeatures(pbmcPlot)[1:10]
plotClusterGeneDot(pbmcPlot, features = features)


# Use data.frame with grouping information, with more tweak on plot
features <- data.frame(features, rep(letters[1:5], 2))
plotClusterGeneDot(pbmcPlot, features = features,
                   clusterFeature = TRUE, clusterCell = TRUE, maxDotSize = 6)