Attach Cell Segmentation Polygons to a NICHESObject
Source:R/add_segmentation.R
add_segmentation.RdStores a pre-built sf object of cell segmentation polygons in the
$segmentation slot of a NICHESObject.
NICHESv2 does not extract segmentation data from Seurat objects – that step
is platform-specific and belongs in the user's own preprocessing script.
Once you have built an sf object (using sf), call this function
to attach it so that downstream plotting utilities can overlay cell
boundaries.
The sf object must have a cell column whose values are cell
barcodes, and every cell in niches.obj$cell.list must have a
corresponding polygon – since both the count matrix and the segmentation
are extracted from the same source object, missing polygons indicate a
barcode mismatch that must be resolved upstream. An error is raised if any
cell is absent. Extra polygons for barcodes not in the object are permitted
with a warning.
See also
NICHESv2_example for an example object whose
$segmentation slot is NULL and ready to receive an
attached sf object; merge_NICHESObjects which
uses sf::rbind() to combine $segmentation slots across
merged objects.
Examples
# NICHESv2 expects 'seg.sf' to be an sf data.frame with:
# * a 'cell' column matching niches.obj$cell.list
# * an sf geometry column of polygons (one per cell)
# The exact construction is platform-specific (Visium HD, Xenium, MERFISH,
# CosMx, etc.). The skeleton below shows the required shape, using a unit
# square for every cell:
if (requireNamespace("sf", quietly = TRUE)) {
data(NICHESv2_example)
# Replace 'polygon.list' with polygons from your segmentation pipeline:
polygon.list <- lapply(NICHESv2_example$cell.list, function(cell) {
sf::st_polygon(list(matrix(c(0, 0, 1, 0, 1, 1, 0, 1, 0, 0),
ncol = 2, byrow = TRUE)))
})
seg.sf <- sf::st_sf(
cell = NICHESv2_example$cell.list,
geometry = sf::st_sfc(polygon.list)
)
niches.obj <- add_segmentation(NICHESv2_example, seg.sf)
}
#> [add_segmentation] Segmentation attached: 80 polygon(s) for 80 cell(s).