Skip to contents

Plot stored CellOracle results

Source: R/CellOraclePlot.R
CellOraclePlot.Rd

Display official CellOracle displacements, grid flow, predicted expression changes or a cluster-specific regulatory network without rerunning Python.

Usage

CellOraclePlot(
  object,
  tool_name = "CellOracle",
  plot_type = c("quiver", "flow", "effect", "network"),
  randomized = FALSE,
  arrow_scale = 1,
  pt.size = 0.5,
  pt.alpha = 0.4,
  top_n = 20L,
  cluster = NULL,
  theme_use = "theme_scop",
  theme_args = list(),
  palette = "Chinese",
  palcolor = NULL
)

Arguments

object

A Seurat object returned by RunCellOracle().

tool_name

Result name in object@tools.

plot_type

One of "quiver", "flow", "effect", or "network".

randomized

Show the official randomized displacement reference in quiver/flow plots. This is not an independent biological replicate.

arrow_scale

Positive multiplier of stored displacements in embedding units. This is a display parameter, not matplotlib's inverse scale argument.

pt.size

Background cell point size.

pt.alpha

Background cell opacity.

top_n

Number of genes in effect plots or edges in network plots.

cluster

Cluster name required for network plots when multiple networks are stored. Networks are never silently pooled across clusters.

theme_use

Theme function or function name.

theme_args

Named list of arguments passed to the theme function.

palette, palcolor

Palette and optional colors passed to thisplot::StatPlot() for effect bars.

Value

A ggplot object. Flow plots apply the stored upstream mass_filter (TRUE means excluded), without smoothing or resampling. Effect bars display mean absolute delta_X across cells, not fold changes or p-values. Network arrow widths reflect absolute upstream coefficients, not activation/repression.

Details

The effect example uses constructed stored predictions to demonstrate plotting without running a perturbation model. Effect bars use thisplot::StatPlot() with the stored values.

See also

RunCellOracle(), thisplot::StatPlot()

Examples

demo <- Seurat::CreateSeuratObject(
  counts = matrix(
    c(3, 1, 0, 2, 1, 4), nrow = 3,
    dimnames = list(c("GeneA", "GeneB", "GeneC"), c("Cell1", "Cell2"))
  )
)
#> Warning: Data is of class matrix. Coercing to dgCMatrix.
demo@tools$CellOracle <- list(gene_effect = data.frame(
  gene = c("GeneA", "GeneB", "GeneC"),
  mean_delta = c(-0.4, 0.8, 0.1),
  mean_abs_delta = c(0.6, 0.8, 0.2)
))
CellOraclePlot(demo, plot_type = "effect")