Skip to contents

This function removes the outlier cells from a group based on the distance patterns, as observed on an UMAP embedding. The function considers a point to be outlier if the distance to the geometric median of the group is above a given threshold determined by a quantile.

Usage

remove_outlier_cells(
  cell_names,
  umap_emb,
  percentile_threshold = 0.75,
  gmedian_point = NULL
)

Arguments

cell_names

A vector with the names of the cells defining the group.

umap_emb

A matrix with the UMAP embedding of the cells. The matrix should have defined the rownames as the cell names.

percentile_threshold

The percentile to be used as threshold for the distance to the geometric median. This value will be used to calculate the quantile. Defaults to 0.75.

gmedian_point

The geometric median point of the group. If NULL, the function will calculate this point. Defaults to NULL.

Value

A vector with the names of the cells that are not considered outliers.