Run spatial variable feature detection — RunSpatialVariableFeatures • scop

Score genes by spot-level spatial autocorrelation. The package "moran" and "geary" methods use a lightweight coordinate KNN graph. "SPARKX" and "nnSVG" use optional external backends when their packages are installed.

Usage

RunSpatialVariableFeatures(
  object = NULL,
  assay = NULL,
  layer = "data",
  features = NULL,
  method = c("moran", "geary", "SPARKX", "nnSVG"),
  image = NULL,
  coord.cols = c("x", "y"),
  k = 6,
  nfeatures = 2000,
  min_spots = 5,
  nperm = 0,
  set_variable_features = TRUE,
  store_results = TRUE,
  verbose = TRUE,
  seed = 11,
  coordinate_space = c("raw", "legacy_display"),
  backend = c("cpp", "r"),
  cores = NULL,
  max_dense_gb = 2,
  ...,
  srt = NULL
)

Arguments

object: A Seurat object.
assay: Assay to use. NULL uses the default assay.
layer: Assay layer used for expression values.
features: Features to score. If NULL, current variable features are used; if no variable features are present, all assay features are used.
method: Spatial variable feature detection method.
image: Spatial image name. Required when multiple images are present; a single image is selected automatically when NULL.
coord.cols: Metadata coordinate columns used when no image is available.
k: Number of nearest spatial neighbors per spot.
nfeatures: Number of top spatial features stored in srt@tools[["SpatialVariableFeatures"]].
min_spots: Minimum number of spots with non-zero expression required for a feature to be tested.
nperm: Number of label permutations used for empirical p values. The default 0 skips p-value calculation.
set_variable_features: Whether to set the top spatial features as variable features for assay.
store_results: Whether to store the full result in srt@tools.
verbose: Whether to print the message. Default is TRUE.
seed: Random seed used for permutation tests.
coordinate_space: Coordinate system used for distance-sensitive analysis. The default is raw, unscaled acquisition coordinates. Use "legacy_display" explicitly to reproduce the display-scaled coordinates used before scop 0.9.0. Distance thresholds and weights use the selected coordinate units; k is a unitless neighbor count.
backend: Backend used by the package "moran" and "geary" methods. "cpp" is the default; use "r" for the reference implementation.
cores: Number of OpenMP threads for C++ Moran/Geary scoring. NULL uses the process OpenMP default.
max_dense_gb: Maximum estimated GB for an expression matrix conversion in the R-reference or nnSVG path. The C++ and SPARK-X paths retain sparse input. This guard bounds one conversion, not total backend memory.
...: Additional arguments passed to external backends.
srt: Deprecated alias for object; supply exactly one of the two. It will be removed in scop 1.0.0.

Value

A Seurat object with spatial variable feature results stored in srt@tools[["SpatialVariableFeatures"]]. Top feature names are available at srt@tools[["SpatialVariableFeatures"]]$summary$top_features.

Examples

data(visium_human_pancreas_sub)
spatial <- Seurat::NormalizeData(
  visium_human_pancreas_sub,
  assay = "Spatial",
  verbose = FALSE
)
spatial <- Seurat::FindVariableFeatures(
  spatial,
  assay = "Spatial",
  nfeatures = 100,
  verbose = FALSE
)

SpatialSpotPlot(
  object = spatial,
  features = Seurat::VariableFeatures(spatial, assay = "Spatial")[1:2]
)


spatial <- RunSpatialVariableFeatures(
  object = spatial,
  assay = "Spatial",
  nfeatures = 50
)
#> ◌ [2026-10-11 23:51:16] Running spatial variable feature detection
#> ✔ [2026-10-11 23:51:16] Spatial variable features completed: 100 tested, 50 ranked in the top set
#> ℹ   Scope method "moran" ("cpp" backend); assay "Spatial", layer "data"; 1986 spots; coordinates "raw"
#> ℹ   Saved 50 top features as assay `VariableFeatures`; full result in returned object tool bundle `SpatialVariableFeatures`
#> ℹ   Plot returned object `SpatialVariableFeaturePlot(<returned_object>, plot_type = "combined", assay = "Spatial", image = "slice1", coord.cols = c("x", "y"))`
SpatialVariableFeaturePlot(object = spatial, plot_type = "combined", nfeatures = 2)