Assign fine cell type labels from enrichment scores

Selects the best-matching fine cell type label for each cluster based on minimum Fisher p-value, with ties broken by descending overlap similarity.

Usage

assign_fine_labels(cluster_col, scores)

Arguments

cluster_col

Factor or character vector. Cluster identifiers per cell, typically a colData column from a SingleCellExperiment. Values must match cluster_name entries in scores.

scores

A data.frame as returned by score_markers_against_panel, with columns cluster_name, marker_set_name, fisher_p, and similarity.

Value

A named list containing:

labels

Factor of fine cell type labels, one per cell, in the same order as cluster_col.

scores

A data.frame of the best label per cluster with columns cluster_name, marker_set_name, fisher_p, and similarity.