RegionNeighbours: Autodetect region neighbours

In jbergenstrahle/STUtility: Analysis and visualization of Spatial Transcriptomics data

Autodetect region neighbours

Description

This function allows you to automatically identify neighbours of a selected region.

Usage

RegionNeighbours(
  object,
  id,
  column.key = "nbs_",
  keep.idents = FALSE,
  keep.within.id = FALSE,
  verbose = FALSE
)

Arguments

object	Seurat object.
id	Group label used to define region, e.g. a cluster id. The region will be selected using the active identity (see 'Idents').
column.key	Sets the name of the column in the meta.data slot where the output is stored. [default: 'nbs_']
keep.idents	If set to TRUE, the identities of the neighbours are kept in the output, otherwise all neighbours will be named 'nbs_id' where id is the group label defined by the aprameter 'ìd' [default: FALSE]
keep.within.id	If set to TRUE, all id spots are kept, otherwise only the spots with outside neighbours are kept
verbose	Print messages

Details

One way of using method this is to find spots surrounding a certain cluster. First, you need to make sure the identity of the Seurat object is set to the meta.data column that you want to use, so for example 'se <- SetIdent(se, value = "seurat_clusters")' if you want to use the default seurat clusters. Then you select the label that defined the region of interest using the 'id' parameter, so for example 'ìd = "1"' will use cluster 1 as the region. If you set the 'keep.idents' parameter to TRUE, the cluster ids of the neighbouring spots will be kept in the result, otherwise they will be returned as one single goup. You can also activate the 'keep.within.id' parameter to include all spots of the selected region in the output, otherwise only the spots along the region border will be kept.

Examples

## Not run: 
# Cluster data, find neighbours to cluster 10 and then plot the results
se <- FindNeighbours(se) %>% FindClusters()
se <- RegionNeighbours(se, id = 10)
ST.FeaturePlot(se, features = "nbs_10")

# Find neighbours to cluster 10 and keep all id spots
se <- RegionNeighbours(se, id = 10)
ST.FeaturePlot(se, features = "nbs_10", keep.within.id = TRUE)

## End(Not run)

jbergenstrahle/STUtility documentation built on March 14, 2023, 7:15 a.m.