Run semla region neighbor detection — RunSemlaRegionNeighbors • scop

Use semla::RegionNeighbors() to identify neighboring spots for selected metadata labels and write the returned columns to Seurat metadata. SCOP provides no dedicated plot for this result; retrieve its schema record with GetSpatialResult() and inspect the recorded metadata columns.

Usage

RunSemlaRegionNeighbors(
  srt,
  column_name,
  column_labels = NULL,
  mode = "outer",
  column_key = NULL,
  image_type = "tissue_lowres",
  verbose = TRUE,
  ...
)

Arguments

srt: A Seurat object with spatial image data.
column_name: Metadata column containing labels.
column_labels: Labels to find neighbors for. If NULL, semla uses all labels in column_name.
mode: Neighbor selection mode passed to semla.
column_key: Prefix for metadata columns returned by semla.
image_type: Image scale used by semla::UpdateSeuratForSemla() when the object does not already contain a Staffli object.
verbose: Whether to print the message. Default is TRUE.
...: Additional arguments passed to semla.

Value

A Seurat object.

Examples

data(visium_human_pancreas_sub)
spatial <- visium_human_pancreas_sub
spatial$region <- ifelse(
  spatial$x > stats::median(spatial$x),
  "right",
  "left"
)
spatial$right_border <- spatial$region == "right" &
  abs(spatial$x - stats::median(spatial$x)) < stats::sd(spatial$x) * 0.25

SpatialSpotPlot(
  spatial,
  group.by = c("region", "right_border"),
  overlay_image = FALSE,
  coord.cols = c("x", "y")
)


spatial <- RunSemlaRegionNeighbors(
  spatial,
  column_name = "region",
  column_labels = "right",
  mode = "outer",
  column_key = "right_border",
  verbose = FALSE
)
#> ℹ Found VisiumV2 object(s).
#> 
#> ── Collecting data from @images slot 
#> ! Array coordinates are not available for non-VisiumHD datasets. Please consider using semla's own functions to load the data. See '?ReadVisiumData()'
#> Warning: 2 spots had 0 neighbors.