Run BayesSpace spatial clustering

Usage

RunBayesSpace(
  object,
  q,
  assay = NULL,
  platform = c("Visium", "VisiumHD", "ST"),
  image = NULL,
  use_reduction = NULL,
  dims = 1:15,
  preprocess = TRUE,
  n.PCs = 15,
  n.HVGs = 2000,
  skip.PCA = !is.null(use_reduction),
  spatial_preprocess_params = list(),
  spatial_cluster_params = list(),
  cluster_colname = "BayesSpace_cluster",
  init_colname = "BayesSpace_init",
  store_sce = TRUE,
  verbose = TRUE,
  coord.cols = c("col", "row"),
  srt = NULL
)

Arguments

object: A Seurat object.
q: Number of BayesSpace clusters.
assay: Assay to use. NULL uses the default assay.
platform: Spatial sequencing platform.
image: Name of the Seurat spatial image used to recover spot coordinates when they are not already present in metadata. For regular Visium data with only pixel x/y coordinates, BayesSpace array coordinates are inferred from the spatial grid.
use_reduction: Optional Seurat reduction to pass to BayesSpace as PCA.
dims: Dimensions from use_reduction to use.
preprocess: Whether to run BayesSpace::spatialPreprocess().
n.PCs, n.HVGs: Parameters passed to spatialPreprocess().
skip.PCA: Whether to skip PCA inside spatialPreprocess().
spatial_preprocess_params: Additional parameters passed to BayesSpace::spatialPreprocess().
spatial_cluster_params: Additional parameters passed to BayesSpace::spatialCluster().
cluster_colname: Metadata column used for BayesSpace clusters.
init_colname: Metadata column used for BayesSpace initial clusters.
store_sce: Whether to store the BayesSpace SingleCellExperiment in srt@tools.
verbose: Whether to print the message. Default is TRUE.
coord.cols: Two metadata columns containing raw x/y coordinates when no spatial image is available. Non-default selections also determine the coordinates passed to BayesSpace, overriding other metadata coordinate columns. The default selection preserves existing array_col/array_row or col/row indices when available. Standard array names (col/row, array_col/array_row, or arraycol/arrayrow, case-insensitive) preserve native array indices; other column names are treated as pixel coordinates from which a regular array grid is inferred.
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 BayesSpace clusters in metadata and raw results in srt@tools[["BayesSpace"]].

Examples

data(visium_human_pancreas_sub)
keep_spots <- unique(round(seq(1, ncol(visium_human_pancreas_sub), length.out = 400)))
spatial <- visium_human_pancreas_sub[, keep_spots]
spatial <- RunBayesSpace(
  spatial,
  q = 3,
  n.PCs = 5,
  n.HVGs = 200,
  store_sce = FALSE,
  spatial_cluster_params = list(nrep = 100, burn.in = 10, thin = 5),
  coord.cols = c("x", "y"),
  verbose = FALSE
)
#> Neighbors were identified for 377 out of 400 spots.
#> Fitting model...
#> Calculating labels using iterations 11 through 100.
SpatialSpotPlot(
  spatial,
  group.by = "BayesSpace_cluster",
  overlay_image = FALSE,
  coord.cols = c("x", "y"),
  pt.size = 1.5
)