Skip to contents

Plot Traditional Statistics

Source: R/plot.R
plotTradStats.Rd

Visualizes results from traditional cluster/cell type-level differential abundance testing. Shows effect sizes as heatmap with significance markers, providing a complementary view to landmark-based analysis for easier interpretation at the population level.

Usage

plotTradStats(x, ...)

# S3 method for class 'TDRObj'
plotTradStats(
  x,
  .model.name = "default",
  .split.by = "clustering",
  .coefs = NULL,
  .q = 0.1,
  .row.space.scaler = 0.2,
  .col.space.scaler = 0.07,
  .label.substr.rm = "",
  .order.ids = FALSE,
  ...
)

Arguments

x

A TDRObj, Seurat, SingleCellExperiment, or HDF5AnnData (anndataR) object processed through get.map() and get.lm(). Statistical results should be stored in .tdr.obj$map$lm.

...

Additional arguments passed to methods.

.model.name

Character string naming which model fit to use from .tdr.obj$map$lm (default "default"). Must match a name used in get.lm(.model.name = ...).

.split.by

Character: "clustering" (default) or "celltyping" - which population grouping to use.

.coefs

Character vector of coefficient names to plot. Defaults to all coefficients from traditional model.

.q

Numeric q-value threshold for significance stars (default 0.1).

.row.space.scaler

Numeric scaling for row height (default 0.2 inches per population).

.col.space.scaler

Numeric scaling for column width (default 0.07).

.label.substr.rm

Character substring to remove from labels (default "").

.order.ids

Logical whether to order IDs based on dendrogram order (default FALSE).

Value

A patchwork object combining two plots: (1) a dot plot showing mean cell percentages per population, and (2) a heatmap showing log fold changes colored by magnitude with significance indicated by asterisks.

Details

Traditional analysis tests for DA at the cluster/celltype level by aggregating cell counts per sample, then using standard models (typically edgeR). This provides:

  • Easier biological interpretation (known populations vs abstract landmarks)

  • Comparison to landmark-based results for validation

  • Detection of broad shifts affecting entire populations

Note: Traditional analysis is less sensitive to subtle within-cluster variation that landmark-based methods can detect.

See also

get.lm, plotTradPerc, plotBeeswarm

Examples

if (FALSE) { # \dontrun{
# After get.lm with traditional analysis
lm.obj <- get.lm(lm.obj, .design = design)

# Heatmap of cluster-level changes
plotTradStats(lm.obj, .split.by = "clustering")

# Cell type-level changes (using a different model)
plotTradStats(lm.obj, .model.name = "full", .split.by = "celltyping")
} # }