pipeline

input a seurat object or single cell matrix, get randomly selected n cells as super-cell seed. For each seed, merge k_merge nearest neighbor cells to the seed cell and calculate average expression.

1
2
3

Super_cell_creation(sc_object, k_filter = NULL, k_merge = 100,
  n = 5000, tag_cell = NULL, verbose = 1, sampling_ref = NULL,
  seed = 123)

`sc_object`	sc matrix with row in cells, col in genes. Or Seurat object, if so Seurat@scale.data will be used for calculation.
`k_filter`	remove cells that have no mutual nearest neighbor with k=k_filter, default not removing any cells.
`k_merge`	merge k=k_merge nearest neighbor to create each super-cell.
`n`	number of super-cell to pick, will be ignored if tag_cell!=NULL.
`tag_cell`	a vector of cell position for super-cell centers.
`verbose`	print time consumption to screen or not (0 not print, 1 print important ones, 2 print all).
`sampling_ref`	a vector group annotation. The random selection of n super-cell centers will try to keep the ratio between group the same.
`seed`	a integer for set.seed

an super-cell matrix, each row is a cell, eacho col is a gene. row names is the position of the super-cell center from input (row number in sc matrix, col number in Seurat@scale.data).

Arthurhe/Lightbulb documentation built on April 13, 2020, 5:12 p.m.