pwOmics (version 1.4.0)

consDynamicNet: Dynamic analysis.

Description

Generates continous data for dynamic analysis of protein, TF and gene data via smoothing splines. 50 time points are generated this way. The following nodes are considered: Nodes which are part of the static consensus graphs from corresponding time points of the two different measurement types. In case a node is not significantly changed at a certain point in time its FC is assumed to remain constant at this time point. Calculation of the consensus-based dynamic net parameters are based on the ebdbNet R package [1]. The number of time points generated via smoothing splines (50) is based on their results for median AUCs of ROC curves. The number of forward time units a node is assumed to influence other nodes can be specified via the laghankel parameter. The cutoff determining the percent of total variance explained by the singular values generated by singular value decomposition (SVD) of the block-Hankel matrix H in order to specify the hidden state dimension K (for further details see [1]).

Usage

consDynamicNet(data_omics, consensusGraphs, laghankel = 3, cutoffhankel = 0.9, conv.1 = 0.15, conv.2 = 0.05, conv.3 = 0.05, verbose = TRUE, max.iter = 100, max.subiter = 200)

Arguments

data_omics

OmicsData object.

consensusGraphs

result from static analysis: consensus graph generated by staticConsensusNet function.

laghankel

integer specifying the maximum relevant time lag to be used in constructing the block-Hankel matrix.

cutoffhankel

cutoff to determine desired percent of total variance explained; default = 0.9 as in [1].

conv.1

value of convergence criterion 1; default value is 0.15 (for further details see [1]).

conv.2

value of convergence criterion 2; default value is 0.05 (for further details see [1]).

conv.3

value of convergence criterion 3; default value is 0.05 (for further details see [1]).

verbose

boolean value, verbose output TRUE or FALSE

max.iter

maximum overall iterations; default value is 100 (for further details see [1]).

max.subiter

maximum iterations for hyperparameter updates; default value is 200 (for further details see [1]).

Value

list of 2 elements: 1) output parameters of dynamic network inference with ebdbNet package 2) splines data generated.

References

1. A. Rau, F. Jaffrezic, J.-L. Foulley, R. W. Doerge (2010). An empirical Bayesian method for estimating biological networks from temporal microarray data. Statistical Applications in Genetics and Molecular Biology, vol. 9, iss. 1, article 9.

Examples

#please run with whole database files (prepared according to vignette)
data(OmicsExampleData)
data_omics = readOmics(tp_prots = c(0.25, 1, 4, 8, 13, 18, 24),
tp_genes = c(1, 4, 8, 13, 18, 24), OmicsExampleData,
PWdatabase = c("biocarta", "kegg", "nci", "reactome"),
TFtargetdatabase = c("chea", "pazar"))
## Not run: 
# data_omics = readTFdata(data_omics)
# data_omics_plus = readPWdata(data_omics,
# loadgenelists = FALSE)
# data_omics = identifyPWs(data_omics_plus)
# data_omics = identifyTFs(data_omics)
# data_omics = enrichPWs(data_omics)
# data_omics = identifyRsofTFs(data_omics, only_enriched = FALSE,
# noTFs_inPW = 1, order_neighbors = 10)
# data_omics = identifyPWTFTGs(data_omics, only_enriched = FALSE)
# statConsNet = staticConsensusNet(data_omics)
# consDynamicNet(data_omics, statConsNet)
# ## End(Not run)