Majority voting to harmonise cell labels on given over clusters

Source: R/label_transfer.R

Majority voting to harmonise cell labels on given over clusters

MajorityVote(
  feature.mat = NULL,
  over.clusters = NULL,
  cell.types,
  k = 20,
  min.prop = 0
)

Arguments

feature.mat

A signature score matrix for over clustering, rows are cells, columns are features.

over.clusters

A named vector contains clusters for majority vote. Default: NULL

cell.types

A data.frame recording cell types at different granularity.

k

k of kmeans for over cluster. Default: 20

min.prop

The minimum proportion of cells required to support naming of the subcluster by a cell type. Default: 0

Value

A data.frame contains consensus predicted cell types.