Skip to contents

Rank markers for each cluster using scran::findMarkers()

Source: R/method-L2_cluster_enrichment.R
rank_cluster_markers.Rd

Runs findMarkers and stores the resulting ranked marker tables in metadata(sce) for downstream annotation, or optionally returns the ranked marker results directly as a list.

Usage

rank_cluster_markers(
  sce,
  cluster_col,
  assay_type = "logcounts",
  test_type = c("wilcox", "t"),
  direction = c("up", "down", "any"),
  pval_type = c("any", "some", "all"),
  min_prop = 0.25,
  metadata_key = "broad_cluster_markers",
  return_list = FALSE,
  BPPARAM = BiocParallel::SerialParam()
)

Arguments

sce

A SingleCellExperiment.

cluster_col

Character scalar. Column in colData(sce) containing the cluster labels to test.

assay_type

Character scalar. Assay to use. Defaults to "logcounts".

test_type

Character scalar. Differential expression test passed to findMarkers. One of "wilcox" or "t". Defaults to "wilcox".

direction

Character scalar. Direction of testing. One of "up", "down", or "any". Defaults to "up".

pval_type

Character scalar. A string specifying how p-values are to be combined across pairwise comparisons for a given group/cluster. For more information see combineMarkers. This can be one of "any", "some", or "all". Defaults to "any".

min_prop

Numeric scalar in (0, 1]. This sets the minimum proportion of pairwise comparisons in which a gene must be DE (in the expected direction) for it to contribute to that gene's summary statistic. This is only relevant when pval_type is set to "any", and will ensure that a gene will only be high-ranked if it is among the top-ranked genes in at least min.prop of the pairwise comparisons. If the pval_type is set to "some" or "all", then this will be ignored. For more information see combineMarkers. Defaults to 0.25, in order to be lenient and identify markers amongst similar populations.

metadata_key

Character scalar. Name of the metadata entry used to store the ranked marker results when return_list = FALSE. Defaults to "broad_cluster_markers".

return_list

Logical scalar. If FALSE (default), store the ranked marker tables in metadata(sce)[[metadata_key]] and return the modified sce. If TRUE, return the output list directly instead of modifying sce.

BPPARAM

A BiocParallelParam object. Defaults to SerialParam().

Value

If return_list = FALSE, the input SingleCellExperiment with ranked marker tables stored in metadata(sce)[[metadata_key]].

If return_list = TRUE, a named list with components:

marker_tables

The output of findMarkers.

params

A list of parameters used to generate the ranked marker tables.

Details

This function is intended as a reusable preprocessing step for:

  • broad label assignment from ranked broad markers, and

  • fine label assignment via Fisher's Exact Test on top-ranked genes.