createNet: Creating a co-expression network from expression matrix.

In gibbslab/coexnet: coexnet: An R package to build CO-EXpression NETworks from Microarray Data

Description

From an expression matrix, this function creates a co-expression network like a graph object using a threshold value and one similarity function.

Usage

createNet(expData, method, threshold)

Arguments

expData	A whole expression matrix or differentially expressed genes matrix, it may be stored in a SummarizedExperiment object.
method	A function to calculate the similarity matrix between genes. It can be "correlation" to use Pearson function or "mutual information" to use a based on entropy information function.
threshold	A value between 0 and 1 to filter the similarity matrix and create the co-expression network.

Value

An undirected co-expression network as igraph object.

Author(s)

Juan David Henao Sanchez <[e-mail]>

Liliana Lopez Kleine <[e-mail]>

See Also

findThreshold to obtain a threshold value based on biology network assumptions.

Examples

# Loading data

pathfile <- system.file("extdata","expression_example.txt",package = "coexnet")
data <- read.table(pathfile,stringsAsFactors = FALSE)

# Building the network

cor_pearson <- createNet(expData = data,threshold = 0.7,method = "correlation")
cor_pearson

mut_inf <- createNet(expData = data,threshold = 0.5,method = "mutual information")
mut_inf

gibbslab/coexnet documentation built on May 17, 2019, 4:19 a.m.