filter_expr: Filter a gene expression data.frame

In csbl-usp/CEMiTool: Co-expression Modules identification Tool

Filter a gene expression data.frame

Description

Filter a gene expression data.frame

Usage

filter_genes(expr, pct = 0.75, apply_vst = FALSE)

Arguments

expr	A data.frame containing expression data
pct	Percentage of most expressed genes to keep.
apply_vst	Logical. If TRUE, will apply variance stabilizing transform before filtering data.

Details

This function takes a gene expression data.frame and applies a percentage filter (keeps the pct) most expressed genes). If apply_vst is TRUE, a variance stabilizing transformation is applied on gene expression values as long as mean and variance values have a Spearman's rho of over 0.5. This transformation is intended to remove this dependence between the parameters. One should then apply the select_genes function to get significant genes.

Value

A data.frame containing filtered expression data

Examples

# Get example expression data
data(expr0)
# Filter genes
expr_f <- filter_genes(expr0)
# Check selected genes
expr_f[1:5, 1:5]
# Filter genes and apply variance stabilizing transformation
expr_f2 <- filter_genes(expr0, apply_vst=TRUE)
# Check results
expr_f2[1:5, 1:5]
# Selected genes
selected <- select_genes(expr_f2)
# Get data.frame with only selected genes
expr_s <- expr_f2[selected, ]
# Check results
expr_s[1:5, 1:5]

csbl-usp/CEMiTool documentation built on March 27, 2024, 5:14 a.m.