human_mouse_ct_marker_enrich: Consensus cell-type naming (Fisher's Exact)

In scMappR: Single Cell Mapper

Consensus cell-type naming (Fisher's Exact)

Description

This function completes the Fisher's exact test cell-type naming for all cell-types.

Usage

human_mouse_ct_marker_enrich(
  gene_lists,
  theSpecies = "human",
  cell_marker_path = "",
  naming_preference = -9
)

Arguments

gene_lists	A named list of vectors containing cell-type markers (mouse or human gene-symbols) which will be named as a cell-type via the Fisher's exact test method.
theSpecies	The species of the gene symbols: "human" or "mouse".
cell_marker_path	If local, path to cell-type marker rda files, otherwise, we will try to download data files.
naming_preference	Either -9 if there is no expected cell-type or one of the categories from get_naming_preference_options(). This is useful if you previously have an idea of which cell-type you were going to enrich.

Details

Fisher's exact test method of cell-type identification using the Panglao and CellMarker databases. It extracts significant pathways (pFDR < 0.05). Then, if naming_preference != -9, it will extract the enriched cell-types within the cell-types identified with the naming preferences option. Generally, this method seems to be biased to cell-types with a greater number of markers.

Value

List with the following elements:

cellTypes	most likely marker for each cell-type from each database.
marker_sets	all enriched cell-types for each cluster from each dataset.

Examples

data(POA_example)
POA_generes <- POA_example$POA_generes
POA_OR_signature <- POA_example$POA_OR_signature
POA_Rank_signature <- POA_example$POA_Rank_signature
Signature <- POA_Rank_signature
rowname <- get_gene_symbol(Signature)
rownames(Signature) <- rowname$rowname
genes <- rownames(Signature)[1:100]
enriched <- human_mouse_ct_marker_enrich(gene_lists = genes, theSpecies = "mouse", 
                                         cell_marker_path = "", naming_preference = "brain")
 

scMappR documentation built on July 9, 2023, 6:26 p.m.