Filter a dataframe with cluster markers.

Usage

filter_markers(
  markers,
  top_n,
  top_n_wt = c("top", "fdr", "lfc", "auc"),
  distinct_by = NULL
)

markers

A dataframe-like object.

top_n

A numeric scalar: number of top markers to keep.

top_n_wt

A character scalar: name of column in markers by which top markers will be determined:

"top": the minimum rank across all pairwise comparisons. Only possible if markers were computed with pval.type="any".
"fdr": False Discovery Rate
"lfc": log2 fold change
"auc": Area Under Curve. Only possible if marker test was "wilcox".

distinct_by

A character scalar: name of column in markers by which unique rows will be kept.

A filtered markers object.