filter_genes: Filter genes with quality control criteria

In ccfindR: Cancer Clone Finder

Description

Select genes with high relative variance in count data for further analysis

Usage

filter_genes(object, markers = NULL, vmr.min = 0,
  min.cells.expressed = 0, max.cells.expressed = Inf,
  rescue.genes = FALSE, progress.bar = TRUE, save.memory = FALSE,
  plot = TRUE, log = "xy", cex = 0.5)

Arguments

object	scNMFSet object.
markers	A vector containing marker genes to be selected. All rows in rowData that contain columns matching this set will be selected.
vmr.min	Minimum variance-to-mean ratio for gene filtering.
min.cells.expressed	Minimum no. of cells expressed for gene filtering.
max.cells.expressed	Maximum no. of cells expressed for gene filtering.
rescue.genes	Selected additional genes whose (non-zero) count distributions have at least one mode.
progress.bar	Display progress of mode-gene scan or VMR calculation with save.memory = TRUE.
save.memory	For a very large number of cells, calculate VMR row by row while avoiding calls to as.matrix(). Progress bar will be displayed unless progress.bar=FALSE.
plot	Plot the distribution of no. of cells expressed vs. VMR.
log	Axis in log-scale, c('x','y','xy').
cex	Symbol size for each gene in the plot.

Details

Takes as input scNMFSet object and scatterplot no. of cells expressed versus VMR (variance-to-mean ratio) for each gene. Optionally, genes are filtered using minimum VMR together with a range of no. of cells expressed.

Value

Object of class scNMFSet.

Examples

set.seed(1)
s <- scNMFSet(matrix(stats::rpois(n=1200,lambda=3),40,30))
s <- filter_genes(s,vmr.min=1.0,min.cells.expressed=28,
        rescue.genes=FALSE)

ccfindR documentation built on Nov. 8, 2020, 5:12 p.m.