netFeatureMatrix: Generating network feature matrix

In mlDNA: Machine Learning-based Differential Network Analysis of Transcriptome Data

Description

This functions generates a feature matrix containing 33 network characteristics by differential network analysis.

Usage

netFeatureMatrix( net1, net2, nodes = NULL, knodes = NULL, 
                  cpus = 1, verbose = TRUE, netResFileDic,
                  features = c( "expDistance", "ASC", "corDistance", 
                                "AllConnectivity", "PosConnectivity", 
                                 "NegConnectivity", "closeness", 
                                 "eccentricity", "eigenvector", 
                                 "page.rank", "dis2knodes", 
                                 "closeness2knodes", "eccenticity2knodes") )

Arguments

net1	exp2net output for condition 1.
net2	exp2net output for condition 2.
nodes	a character vector, a set of genes will be analyzed.
knodes	a character vector, a set of interested genes (e.g., known stress-related genes).
cpus	an integrator value, the number of cpus will be used for parallel computing.
verbose	logical value indicate whether the progress information will be output.
netResFileDic	file directory where the network-related results will be output.
features	a character vector specifying the network characteristics will be analyzed.

Value

a numeric matrix, feature matrix generated from network comparision analysis.

Note

1) More information about these network properties has been described in (Ma and Wang, 2013).

2) To run netFeatureMatrix, the parameters "v" and "to" in "exp2net" function should be the whole genes in expmat.

Author(s)

Chuang Ma, Xiangfeng Wang.

References

[1] Chuang Ma, Xiangfeng Wang. Machine learning-based differential network analysis:a case study of stress-responsive transcriptomes in Arabidopsis thaliana. 2013(Submitted).

Examples

## Not run: 
   
   ##suppose the network-related results are stored at:
   netResFileDic = "/home/wanglab/mlDNA/network/"
   
   ##only consider a subset of genes here
   genes <- unique ( c( rownames(ControlExpMat)[1:1000], positiveSamples[1:100] ) )
   res_ControlSub <- exp2net( expmat = ControlExpMat[genes,], method = "GCC", 
                              pvalue = 0.01, cpus = 6, expDescribe = "Control_Sub", 
                              connListFlag = TRUE, distmatFlag = TRUE, 
                              saveType = "bigmatrix", netResFileDic = netResFileDic, 
                              v = genes, to = genes )
                   
    res_StressSub <- exp2net( expmat = SaltExpMat[genes,], method = "GCC", 
                              pvalue = 0.01, cpus = 6, expDescribe = "Stress_Sub", 
                              connListFlag = TRUE, distmatFlag = TRUE, 
                              saveType = "bigmatrix", netResFileDic = netResFileDic, 
                              v = genes, to = genes )

    ##generate network feature matrix
    nodes <- genes
    knodes <- intersect( genes, positiveSamples )
    netFeatureMat <- netFeatureMatrix( net1 = res_ControlSub, net2 = res_StressSub, 
                                       nodes = nodes, knodes = knodes, 
                                       cpus = 2, verbose = TRUE, 
                                       netResFileDic = netResFileDic,
                                       features = c( "expDistance", "ASC", 
                                             "corDistance", "AllConnectivity", 
                                             "PosConnectivity", "NegConnectivity", 
                                             "closeness","eccentricity", 
                                             "eigenvector", "page.rank", 
                                             "dis2knodes", "closeness2knodes", 
                                             "eccenticity2knodes") )
                                                     

## End(Not run)

mlDNA documentation built on May 2, 2019, 2:15 p.m.