linksRanking: Ranking of links inferred from data

Description Usage Arguments Details Value Author(s) References See Also Examples

View source: R/linksRanking.R

Description

This function uses data (CNOlist) to rank links based on measurement error model as used by FEED method to reverse-engineer the network.

Usage

1
linksRanking(CNOlist, measErr=c(0.1, 0), savefile=FALSE)

Arguments

CNOlist

a CNOlist structure, as produced by makeCNOlist

measErr

a 2 value vector (err1, err2) defining the error model of the data as sd^2 = err1^2 + (err2*data)^2, default to c(0.1, 0)

savefile

TRUE to save the file in txt format, FALSE not. Default is FALSE.

Details

This function is similar to the fist step of FEED to reverse engineer the network strictly from data, i.e. the inference of Boolean tables, as described in (Eduati et al., PLoS ONE, 2010) and implemented in makeBTables. Links are ranked according to the upper limit value of parameterk allowing the presence of the link, where k is the parameter which is multiplied by the measurement error in order to assess the relevance of a link. The function returs link in decreasing order of importance and associate to each link a value (maximum value of k allowing the presence of the link) quantifying its relevance.

Value

this function returns a list with fields:

Lrank

a matrix in which each link is associated with a numerical value, links are ordered in decreasing order of reliability)

Author(s)

F.Eduati

References

F. Eduati, A. Corradin, B. Di Camillo, G. Toffolo. A Boolean approach to linear prediction for signaling network modeling. PLoS ONE; 5(9): e12789.

See Also

makeCNOlist, makeBTables

Examples

1
2
3
data(CNOlistDREAM,package="CellNOptR")
Lrank <- linksRanking(CNOlist=CNOlistDREAM, measErr=c(0.1, 0))
    

saezlab/CNORfeeder documentation built on April 2, 2020, 8:40 p.m.