calc_var_genes: Subset Genes

In kgellatl/SignalCell: Network Analysis for single cell RNA sequencing Data (sc-RNASeq)

Description

This will select genes based on minimum expression, coefficient of variation, or by a preliminary PCA.

Usage

calc_var_genes(
  input,
  method,
  assay = NULL,
  threshold = 1,
  minCells = 10,
  nComp = 10,
  log = F,
  fudge = F,
  fudge_val = 0.01
)

Arguments

input	the input sce
method	can either be "CV", "Malhanobis", or "Gini"
assay	if NULL will default to def_assay. Can provide a character assay argument.
threshold	UMI threshold for gene detection
minCells	number of cells expressed above threshold for a given gene
nComp	if method = PCA, the number of components to keep
log	whether or not to log scale the data
fudge	whether or not to add a fudge factor to the entire matrix
fudge_val	the value to add to matrix before transformation

Details

Genes will be first filtered by minimum expression selecting by subsetting to genes that are expressed above the threshold in more than minCells. If the method is CV, it will first subset the genes based on the expression cutoffs, then find the coefficient of variation across all genes. Next it will select the percentile of genes (cutoff) based on their coefficient of variation. The last method will perform PCA on the cells, and then look at the loadings of each gene. By finding genes that are off center (via malhanoobis distance) we can filter to include only genes that contribute significant variance to the data.

Examples

gene_subset <- subset_genes(input = sce, method = "PCA", assay = "counts")

kgellatl/SignalCell documentation built on Sept. 3, 2020, 8:45 a.m.