tni.graph {RTN}R Documentation

Compute a graph from TNI objects.

Description

Extract results from a TNI object and compute a graph.

Usage

tni.graph(object, tnet = "dpi", gtype="rmap", minRegulonSize=15, tfs=NULL, amapFilter="quantile", 
amapCutoff=NULL, ntop=NULL, ...)

Arguments

object

an object of class 'TNI' TNI-class.

tnet

a single character value specifying which network information should be used to compute the graph. Options: "ref" and "dpi".

gtype

a single character value specifying the graph type. Options: "rmap", "amap", "mmap" and "mmapDetailed". The "rmap" option returns regulatory maps represented by TFs and targets (regulons); "amap" computes association maps among regulons (estimates the overlap using the Jaccard Coefficient); "mmap" and "mmapDetailed" return modulated maps derived from the tni.conditional function.

minRegulonSize

a single integer or numeric value specifying the minimum number of elements in a regulon. Regulons with fewer than this number are removed from the graph.

tfs

a vector with transcription factor identifiers.

amapFilter

a single character value specifying which method should be used to filter association maps (only when gtype="amap"). Options: "phyper","quantile" and "custom".

amapCutoff

a single numeric value (>=0 and <=1) specifying the cutoff for the association map filter. When amapFilter="phyper", amapCutoff corresponds to a pvalue cutoff; when amapFilter="quantile", amapCutoff corresponds to a quantile threshold; and when amapFilter="custom", amapCutoff is a JC threshold.

ntop

when gtype="mmapDetailed", ntop is an optional single integer value (>=1) specifying the number of TF's targets that should be used to compute the modulated map. The n targets is derived from the top ranked TF-target interations, as defined in the mutual information analysis used to construct the regulon set.

...

additional arguments passed to tni.graph function.

Value

a graph object.

Author(s)

Mauro Castro

Examples


data(dt4rtn)

# select 5 regulatoryElements for a quick demonstration!
tfs4test <- dt4rtn$tfs[c("PTTG1","E2F2","FOXM1","E2F3","RUNX2")]

## Not run: 

rtni <- tni.constructor(expData=dt4rtn$gexp, regulatoryElements=tfs4test, 
        rowAnnotation=dt4rtn$gexpIDs)
rtni <- tni.permutation(rtni)
rtni <- tni.bootstrap(rtni)
rtni <- tni.dpi.filter(rtni, eps=0)

# compute regulatory maps
g <- tni.graph(rtni, tnet="dpi", gtype="rmap", tfs=tfs4test)

# option: plot the igraph object using RedeR
library(RedeR)
rdp <- RedPort()
calld(rdp)
addGraph(rdp,g)
addLegend.shape(rdp,g)
addLegend.color(rdp,g,type="edge")
relax(rdp,p1=20,p3=50,p4=50,p5=10,p8=50,ps=TRUE)
#...it should take some time to relax!

# compute association maps
resetd(rdp)
g <- tni.graph(rtni, tnet="ref", gtype="amap", tfs=tfs4test)
addGraph(rdp,g)
addLegend.size(rdp,g)
addLegend.size(rdp,g,type="edge")


## End(Not run)

[Package RTN version 2.4.6 Index]