getOptimalFeatureSet: Determine the optimal feature set using Density Index (DI)

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

Description

Determine the optimal feature set using Density Index (DI)

Usage

getOptimalFeatureSet(
  filt.data,
  ordered.genes,
  elbow.pt = 25,
  k = 10,
  num.pcs = 20,
  error = 0
)

Arguments

filt.data	log-transformed filtered gene-expression matrix
ordered.genes	genes ordered after stepwise regression
elbow.pt	Elbow point to start determining optimal feature set
k	number of nearest neighbours for CI computation
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

optimal set of feature genes

bbbranjan/DUBStepR documentation built on May 27, 2021, 4:13 p.m.