Spatial cellular network graph — spatialNearestNeighbors • SpatialMap

Create a cellular nearest neighbor graph from spatial Representation with a variety of methods. A wrapper around computeNearestNeighbors().

Usage

spatialNearestNeighbors(
  object,
  representation = "spatial",
  method = c("knn", "frnn", "snn", "del"),
  k = 5,
  eps = 50,
  analyze = "regions",
  name = NULL,
  ...
)

Arguments

object: A SpatialMap or Region object
representation: What Representation to use. Defaults to "spatial"
method: 'knn', 'frnn', 'snn', or 'del'. See vignette("Tutorial_Advanced_spatial") for more details.
k: How many nearest neighbors (knn and snn only)
eps: How large a radius (frnn only)
analyze: Passed to .smapply. Default "regions" ensures that separate samples are analyzed separately (ignores activeAnalysis). Any other setting may have unpredictable results.
name: A custom name for the graph. By default, will be named according to the method and parameter specified (e.g. spatial_knn_5, spatial_frnn_50, spatial_del)
...: Additional parameters passed to the NN methods.

Value

A SpatialMap or Region object with a new NN computed and added to each Region, with a name determined by name

Details

See vignette("AnalysisGuide5_Cell_interactions") for an example workflow that makes use of this function, and vignette("Tutorial_Advanced_spatial") for a detailed description of the different methods available for use with this function, as well as some options for advanced work with the resulting NN objects.

method = "knn": createKNNGraph.Region()
method = "snn": createSharedNNGraph.Region()
method = "frnn": createFixedRadiusNNGraph.Region()
method = "del": createDelaunayTriangulation.Region()