minet (version 3.30.0)

clr: Context Likelihood or Relatedness Network

Description

clr takes the mutual information matrix as input in order to return the infered network - see details.

Usage

clr( mim, skipDiagonal=1 )

Arguments

mim

A square matrix whose i,j th element is the mutual information between variables $X_i$ and $X_j$ - see build.mim.

skipDiagonal

Skips the diagonal in the calculation of the mean and sd, default=1.

Value

clr returns a matrix which is the weighted adjacency matrix of the network. In order to display the network, load the package Rgraphviz and use the following comand plot( as( returned.matrix ,"graphNEL") )

Details

The CLR algorithm is an extension of relevance network. Instead of considering the mutual information $I(Xi;Xj)$ between features $Xi$ and $Xj$, it takes into account the score $sqrt(zi^2+zj^2)$, where $$ z_i = \max \bigg\lbrace 0, \frac{I(X_i;X_j)-\mu_i}{\sigma_i} \bigg\rbrace $$ and $mean(Xi)$ and $sd(Xi)$ are, respectively, the mean and the standard deviation of the empirical distribution of the mutual information values $I(Xi,Xk)$, k=1,...,n.

References

Jeremiah J. Faith, Boris Hayete, Joshua T. Thaden, Ilaria Mogno, Jamey Wierzbowski, Guillaume Cottarel, Simon Kasif, James J. Collins, and Timothy S. Gardner. Large-scale mapping and validation of escherichia coli transcriptional regulation from a compendium of expression profiles. PLoS Biology, 2007.

See Also

build.mim, aracne, mrnet, mrnetb

Examples

data(syn.data)
mim <- build.mim(syn.data,estimator="spearman")
net <- clr(mim)