clusterCells: Cluster cells in a SingleCellExperiment

In MarioniLab/scran: Methods for Single-Cell RNA-Seq Data Analysis

Cluster cells in a SingleCellExperiment

Description

A SingleCellExperiment-compatible wrapper around clusterRows from the bluster package.

Usage

clusterCells(
  x,
  assay.type = NULL,
  use.dimred = NULL,
  BLUSPARAM = NNGraphParam(),
  ...
)

Arguments

x	A SummarizedExperiment or SingleCellExperiment object containing cells in the columns.
assay.type	Integer or string specifying the assay values to use for clustering, typically log-normalized expression.
use.dimred	Integer or string specifying the reduced dimensions to use for clustering, typically PC scores. Only used when assay.type=NULL, and only applicable if x is a SingleCellExperiment.
BLUSPARAM	A BlusterParam object specifying the clustering algorithm to use, defaults to a graph-based method.
...	Further arguments to pass to clusterRows.

Details

This is largely a convenience wrapper to avoid the need to manually extract the relevant assays or reduced dimensions from x. Altering BLUSPARAM can easily change the parameters or algorithm used for clustering - see ?"BlusterParam-class" for more details.

Value

A factor of cluster identities for each cell in x, or a list containing such a factor - see the return value of ?clusterRows.

Author(s)

Aaron Lun

Examples

library(scuttle)
sce <- mockSCE()
sce <- logNormCounts(sce)

# From log-expression values:
clusters <- clusterCells(sce, assay.type="logcounts")

# From PCs:
sce <- scater::runPCA(sce)
clusters2 <- clusterCells(sce, use.dimred="PCA")

# With different parameters:
library(bluster)
clusters3 <- clusterCells(sce, use.dimred="PCA", BLUSPARAM=NNGraphParam(k=5))

# With different algorithms:
clusters4 <- clusterCells(sce, use.dimred="PCA", BLUSPARAM=KmeansParam(centers=10))

MarioniLab/scran documentation built on March 7, 2024, 1:45 p.m.