Skip to contents

Plot High-Dimensional Dataset Embeddings for Reference and Simulated Single-Cell Data

Source: R/10_visualizations.R
plot_dataset_embeddings.Rd

Generates publication-ready comparative visualizations of empirical reference and simulated single-cell datasets across UMAP, t-SNE, or PCA coordinate spaces. Supports side-by-side comparison, comprehensive multi-method facet grids, and co-embedded overlays, with customizable color mappings and themes.

Usage

plot_dataset_embeddings(
  embedding_data,
  reduction = c("umap", "tsne", "pca"),
  layout = c("facet", "side_by_side", "overlay"),
  color_by = c("cell_type", "cluster", "library_size", "dataset", "batch"),
  selected_methods = NULL,
  pt_size = 0.8,
  alpha = 0.75,
  palette = "Set1",
  title = NULL,
  base_size = 12
)

Arguments

embedding_data

A data.frame produced by compute_dataset_embeddings.

reduction

Character string specifying reduction used: "umap", "tsne", or "pca".

layout

Character string specifying the comparison layout:

"facet"

Faceted grid showing Reference alongside all selected simulated datasets simultaneously (default).

"side_by_side"

Direct side-by-side comparison of Reference versus a single chosen simulator.

"overlay"

Overlaid single coordinate space showing Reference and Simulated cells together.

color_by

Feature used for coloring cells: "cell_type" (default), "cluster", "library_size", "dataset", or "batch".

selected_methods

Optional character vector of simulator names to display. If NULL, displays all available simulators.

pt_size

Numeric point size (default: 0.8).

alpha

Numeric transparency in [0, 1] (default: 0.75).

palette

Character string specifying discrete color palette from RColorBrewer (default: "Set1").

title

Optional title character string. If NULL, an informative default title is generated.

base_size

Base font size for ggplot2 rendering (default: 12).

Value

A ggplot2 object.

Examples

data(example_scrna, package = "scSimEval")
emb <- compute_dataset_embeddings(
  reference = example_scrna$ref,
  simulated = list("Splatter" = example_scrna$sim),
  reduction = "umap",
  cell_types = example_scrna$cell_types
)
#> Warning: You're computing too large a percentage of total singular values, use a standard svd instead.
#> Warning: You're computing too large a percentage of total singular values, use a standard svd instead.
plot_dataset_embeddings(emb, reduction = "umap", layout = "facet")