dms: Dense module search function

In aaronwolen/dmGWAS2: A fork of EW_dmGWAS

Description

dms constructs a node- and edge-weighted PPI network, performs dense module searching, generates simulation data from random networks, normalizes module scores using simulation data, removes un-qualified modules, and orders resultant modules according to their significance.

Usage

dms(network, geneweight, expr1, expr2 = NULL, d = 1, r = 0.1,
  lambda = "default")

Arguments

network	A data frame containing a symbolic edge list of the PPI network
geneweight	A data frame containing two columns: the first is unique gene identifier (should be coordinate with the node symbol used in PPI); the second is gene-based p-value derived from GWAS
expr1	A data frame containing gene expression data from case samples. The first column is gene identifier (should be coordinate with the node symbol used in PPI
expr2	A data frame containing gene expression data from control samples. The first column should be the same as expr1
d	An integer used to define the order of neighbour genes to be searched. This parameter is always set up as 1 in dmGWAS_3.0, but could be 1 or 2 in dmGWAS_1.0 and dmGWAS_2.X
r	A float indicating the cut-off for increment during module expanding process. Greater r will generate smaller module. Default is 0.1.
lambda	A float between 0 and 1 to balance node and edge weights. dmGWAS_NEW will estimate it by default

Value

dms returns a list containing relevant data and results, including:

GWPI	the edge-weighted network used for searching
lambda	lambda coefficient
genesets	list of genes comprising each dense module, named for the seed gene
genesets.length.null.dis	randomization data for normalization
module.score.matrix	contains Sm and Sn

Examples

## Not run: 
 res.list <- dms(network, geneweight, expr1, expr2, r=0.1)

## End(Not run)

aaronwolen/dmGWAS2 documentation built on May 10, 2019, 4:04 a.m.