scCustomize (version 3.0.0)

Variable_Features_ALL_LIGER: Perform variable gene selection over whole dataset

Description

Performs variable gene selection for LIGER object across the entire object instead of by dataset and then taking union.

Usage

Variable_Features_ALL_LIGER(
  liger_object,
  num_genes = NULL,
  var.thresh = 0.3,
  alpha.thresh = 0.99,
  tol = 1e-04,
  do.plot = FALSE,
  pt.size = 1.5,
  chunk = 1000
)

Value

A LIGER Object with variable genes in correct slot.

Arguments

liger_object

LIGER object name.

num_genes

Number of genes to find. Optimizes the value of var.thresh to get this number of genes, (Default is NULL).

var.thresh

Variance threshold. Main threshold used to identify variable genes. Genes with expression variance greater than threshold (relative to mean) are selected. (higher threshold -> fewer selected genes).

alpha.thresh

Alpha threshold. Controls upper bound for expected mean gene expression (lower threshold -> higher upper bound). (default 0.99)

tol

Tolerance to use for optimization if num.genes values passed in (default 0.0001). Only applicable for rliger < 2.0.0.

do.plot

Display log plot of gene variance vs. gene expression. Selected genes are plotted in green. (Default FALSE)

pt.size

Point size for plot.

chunk

size of chunks in hdf5 file. (Default 1000)

References

Matching function parameter text descriptions are taken from rliger::selectGenes which is called by this function after creating new temporary object/dataset. https://github.com/welch-lab/liger. (License: GPL-3).

Examples

if (FALSE) {
liger_obj <- Variable_Features_ALL_LIGER(liger_object = liger_obj, num_genes = 2000)
}