DUBStepR: DUBStepR - Obtain a list of feature genes to characterise...

In DUBStepR: Correlation-Based Feature Selection for Single-Cell RNA Sequencing Data

Description

DUBStepR - Obtain a list of feature genes to characterise cell types

Usage

DUBStepR(
  input.data,
  min.cells = 0.05 * ncol(input.data),
  species = "human",
  optimise.features = TRUE,
  k = 10,
  num.pcs = 20,
  error = 0
)

Arguments

input.data	input gene expression matrix (genes x cells)
min.cells	minimum number of cells to filter genes out
species	species to use for gene filtering: "human" (default), "mouse" and "rat"
optimise.features	Determine optimal feature set using density index
k	number of nearest neighbours. Default is 10
num.pcs	number of principal components to represent sc data. Default is 20
error	Acceptable error margin for kNN computation. Default is 0, but is set to 1 for large datasets

Value

Returns optimal feature set

Author(s)

ranjanb

Examples

res <- DUBStepR(input.data = pbmc_norm_small_data)

DUBStepR documentation built on Oct. 5, 2021, 5:08 p.m.