cellTypeSpecificity: Get cell type specificity of gene expression

In GabrielHoffman/dreamlet: Scalable differential expression analysis of single cell transcriptomics datasets with complex study designs

Get cell type specificity of gene expression

Description

For each gene, compute fraction of overall expression attributable to each cell type

Usage

cellTypeSpecificity(pb, ...)

Arguments

pb	SingleCellExperiment of pseudobulk data where easy assay is a cell type.
...	other arguments passed to edgeR::calcNormFactors()

Details

Sum counts for each cell type, and compute the fraction of counts-per-million attributable to each cell type for each gene

Value

matrix of the fraction of expression attributable to each cell type for each gene.

Examples

library(muscat)
library(SingleCellExperiment)

data(example_sce)

# create pseudobulk for each sample and cell cluster
pb <- aggregateToPseudoBulk(example_sce,
  assay = "counts",
  cluster_id = "cluster_id",
  sample_id = "sample_id",
  verbose = FALSE
)

# Compute cell type specificity of each gene
df <- cellTypeSpecificity(pb)

# Violin plot of specificity scores for each cell type
# Dashed line indicates genes that are equally expressed
# across all cell types.  For K cell types, this is 1/K
plotViolin(df)

# Compute the maximum specificity score for each gene
scoreMax <- apply(df, 1, max)
head(scoreMax)

# For each cell type, get most specific gene
genes <- rownames(df)[apply(df, 2, which.max)]

# Barplot of 5 genes
plotPercentBars(df, genes = genes)

# heatmap of 5 genes that are most cell type specific
dreamlet::plotHeatmap(df, genes = genes)

GabrielHoffman/dreamlet documentation built on Oct. 29, 2024, 4:04 a.m.