Calculate silhouette widths from a Seurat embedding — sn_calculate

Silhouette widths summarize how well cells are separated by a categorical metadata label in the selected embedding.

Usage

sn_calculate_silhouette(
  x,
  label,
  reduction = "pca",
  dims = NULL,
  cells = NULL,
  max_cells = 3000,
  stratify_by = label,
  seed = 717
)

Arguments

x: A Seurat object.
label: Metadata column used as the grouping label.
reduction: Reduction name used to extract embeddings. Defaults to "pca".
dims: Optional integer vector of embedding dimensions to retain.
cells: Optional character vector of cell names to score.
max_cells: Optional integer cap used to subsample cells before running the metric. Defaults to 3000 because silhouette needs a full distance matrix.
stratify_by: Optional metadata column used to preserve representation during subsampling. Defaults to label.
seed: Random seed used when max_cells triggers subsampling.

Value

A data frame with per-cell silhouette widths.

Examples

if (FALSE) { # \dontrun{
data("pbmc_small", package = "Shennong")
pbmc <- sn_run_cluster(
  pbmc_small,
  batch = "sample",
  species = "human",
  verbose = FALSE
)
sil_tbl <- sn_calculate_silhouette(
  pbmc,
  label = "seurat_clusters",
  reduction = "harmony"
)
head(sil_tbl)
} # }