RankMarkerGenes: Rank the top marker genes for each cluster from spectral...

In countland: Analysis of Biological Count Data, Especially from Single-Cell RNA-Seq

Rank the top marker genes for each cluster from spectral clustering.

Description

Rank the top marker genes for each cluster from spectral clustering.

Usage

RankMarkerGenes(C, method = "prop-zero", subsample = FALSE)

Arguments

C	countland object
method	prop-zero to rank by proportion of cells that are non-zero (default), or rank-sums to rank using Wilcoxon rank-sums test
subsample	if TRUE, use subsampled counts, otherwise use counts (default=FALSE)

Value

countland object with slots marker_genes and marker_full

Examples

gold_path <- system.file("testdata", package = "countland", mustWork = TRUE)
gold.data <- Seurat::Read10X(data.dir = gold_path)
C <- countland(gold.data)
C <- Dot(C)
C <- Embed(C,n_components=5)
C <- Cluster(C,n_clusters=3)
C <- RankMarkerGenes(C,method='prop-zero',subsample=FALSE)

countland documentation built on May 29, 2024, 7:13 a.m.