plot.COMPASSResult: Plot a COMPASSResult

In COMPASS: Combinatorial Polyfunctionality Analysis of Single Cells

Description

This function can be used to visualize the mean probability of response; that is, the probability that there is a difference in response between samples subjected to the 'treatment' condition, and samples subjected to the 'control' condition.

Usage

## S3 method for class 'COMPASSResult'
plot(
  x,
  y,
  subset = NULL,
  threshold = 0.01,
  minimum_dof = 1,
  maximum_dof = Inf,
  must_express = NULL,
  row_annotation,
  palette = colorRampPalette(brewer.pal(10, "Purples"))(20),
  show_rownames = FALSE,
  show_colnames = FALSE,
  measure = NULL,
  order_by = FunctionalityScore,
  order_by_max_functionality = TRUE,
  markers = NULL,
  ...
)

Arguments

x	An object of class COMPASSResult.
y	This argument gets passed to row_annotation, if row_annotation is missing. It can be used to group rows (individuals) by different conditions as defined in the metadata.
subset	An R expression, evaluated within the metadata, used to determine which individuals should be kept.
threshold	A numeric threshold for filtering under-expressed categories. Any categories with mean score < threshold are removed.
minimum_dof	The minimum degree of functionality for the categories to be plotted.
maximum_dof	The maximum degree of functionality for the categories to be plotted.
must_express	A character vector of markers that should be included in each subset plotted. For example, must_express=c("TNFa & IFNg") says we include only subsets that are positive for both TNFa or IFNg, while must_express=c("TNFa", "IFNg") says we should keep subsets which are positive for either TNFa or IFNg.
row_annotation	A vector of names, pulled from the metadata, to be used for row annotation.
palette	The colour palette to be used.
show_rownames	Boolean; if TRUE we display row names (ie, the individual ids).
show_colnames	Boolean; if TRUE we display column names (ie, the column name associated with a cytokine; typically not needed)
measure	Optional. By default, we produce a heatmap of the mean gammas produced in a model fit. We can override this by supplying a matrix of suitable dimension as well; these can be generated with the Posterior* functions – see Posterior for examples.
order_by	Order rows within a group. This should be a function; e.g. FunctionalityScore, mean, median, and so on. Set this to NULL to preserve the original ordering of the data.
order_by_max_functionality	Order columns by functionality within each degree subset. to TRUE.
markers	specifies a subset of markers to plot. default is NULL, which means all markers.
...	Optional arguments passed to pheatmap.

Value

The plot as a grid object (grob). It can be redrawn with e.g. grid::grid.draw().

Examples

## visualize the mean probability of reponse
plot(CR)

## visualize the proportion of cells belonging to a category
plot(CR, measure=PosteriorPs(CR))

COMPASS documentation built on Nov. 8, 2020, 8:05 p.m.