netinf.cv: Function performing network inference by combining priors and...

In predictionet: Inference for predictive networks designed for (but not limited to) genomic data

Description

The function netinf.cv perform a cross-validation loop and infers a gene network by combining priors and genomic data in each fold. This allows to estimate the predictive ability of the inferred network as well as edge stability.

Usage

netinf.cv(data, categories, perturbations, priors, predn, priors.count = TRUE, priors.weight = 0.5, maxparents = 3, subset, method = c("regrnet", "bayesnet"), ensemble = FALSE, ensemble.maxnsol=3, predmodel = c("linear", "linear.penalized", "cpt"), nfold = 10, causal = TRUE, seed, bayesnet.maxcomplexity = 0, bayesnet.maxiter = 100, verbose = FALSE) 

Arguments

data	matrix of continuous or categorical values (gene expressions for example); observations in rows, features in columns.
categories	if this parameter missing, 'data' should be already discretize; otherwise either a single integer or a vector of integers specifying the number of categories used to discretize each variable (data are then discretized using equal-frequency bins) or a list of cutoffs to use to discretize each of the variables in 'data' matrix. If method='bayesnet', this parameter should be specified by the user.
perturbations	matrix of 0, 1 specifying whether a gene has been perturbed (e.g., knockdown, overexpression) in some experiments. Dimensions should be the same than data.
priors	matrix of prior information available for gene-gene interaction (parents in rows, children in columns). Values may be probabilities or any other weights (citations count for instance). if priors counts are used the parameter priors.count should be TRUE so the priors are scaled accordingly.
predn	indices or names of variables to fit during network inference. If missing, all the variables will be used for network inference.
priors.count	TRUE if priors specified by the user are number of citations (count) for each interaction, FALSE if probabilities or any other weight in [0,1] are reported instead.
priors.weight	real value in [0,1] specifying the weight to put on the priors (0=only the data are used, 1=only the priors are used to infer the topology of the network).
maxparents	maximum number of parents allowed for each gene.
subset	vector of indices to select only subset of the observations.
method	regrnet for regression-based network inference, bayesnet for bayesian network inference with the catnet package.
ensemble	TRUE if the ensemble approach should be used, FALSE otherwise.
ensemble.maxnsol	Number of equivalent solutions chosen at each step.
predmodel	type of predictive model to fit; linear for linear regression model, linear.penalized for regularized linear regression model, cpt for conditional probability tables estimated after discretization of the data.
nfold	number of folds for the cross-validation.
causal	'TRUE' if the causality should be inferred from the data, 'FALSE' otherwise
seed	set the seed to make the cross-validation and network inference deterministic.
bayesnet.maxcomplexity	maximum complexity for bayesian network inference, see Details.
bayesnet.maxiter	maximum number of iterations for bayesian network inference, see Details.
verbose	TRUE if messages should be printed, FALSE otherwise.

Details

bayesnet.maxcomplexity and bayesnet.maxiter are parameters to be passed to the network inference method (see cnSearchOrder and cnSearchSA from the catnet package for more details).

Value

method	name of the method used for network inference.
topology	topology of the model inferred using the entire dataset.
topology.coeff	if method='regrnet' topology.coeff contains an adjacency matrix with the coefficients used in the local regression model; parents in rows, children in columns. Additionally the beta_0 values for each model in the first row of the matrix
topology.cv	topology of the networks inferred at each fold of the cross-validation.
topology.coeff.cv	if method='regrnet' topology.coeff contains an adjacency matrix with the coefficients used in the local regression model; parents in rows, children in columns. Additionally the beta_0 values for each model in the first row of the matrix. Inferred at each fold of the cross-validation
prediction.score.cv	list of prediction scores (R2, NRMSE, MCC) computed at each fold of the cross-validation.
edge.stability	stability of the edges inferred during cross-validation; only the stability of the edges present in the network inferred using the entire dataset is reported.
edge.stability.cv	stability of the edges inferred during cross-validation.
edge.relevance	mean relevance score for each across folds in cross-validation.
edge.relevance.cv	relevance score for each across computed during cross-validation.

Author(s)

Benjamin Haibe-Kains, Catharina Olsen

Examples

## load gene expression data for colon cancer data, list of genes related to RAS signaling pathway and the corresponding priors
data(expO.colon.ras)
## create matrix of perturbations (no perturbations in this dataset)
pert <- matrix(0, nrow=nrow(data.ras), ncol=ncol(data.ras), dimnames=dimnames(data.ras))

## number of genes to select for the analysis
genen <- 10
## select only the top genes
goi <- dimnames(annot.ras)[[1]][order(abs(log2(annot.ras[ ,"fold.change"])), decreasing=TRUE)[1:genen]]
mydata <- data.ras[ , goi, drop=FALSE]
myannot <- annot.ras[goi, , drop=FALSE]
mypriors <- priors.ras[goi, goi, drop=FALSE]
mydemo <- demo.ras
mypert <- pert[ , goi, drop=FALSE]

########################
## regression-based network inference
########################
## number of fold for cross-validation
res <- netinf.cv(data=mydata, categories=3, perturbations=mypert, priors=mypriors, priors.weight=0.5, method="regrnet", nfold=3, seed=54321)

## MCC for predictions in cross-validation
print(res$prediction.score.cv)

## export network as a 'gml' file that you can import into Cytoscape
## Not run: rr <- netinf2gml(object=res, file="predictionet_regrnet")

########################
## bayesian network inference
########################
## infer a bayesian network network from data and priors
## number of fold for cross-validation
## Not run: res <- netinf.cv(data=mydata, categories=3, perturbations=mypert, priors=mypriors, priors.count=TRUE, priors.weight=0.5, method="bayesnet", nfold=3, seed=54321)

## MCC for predictions in cross-validation
## Not run: print(res$prediction.score.cv)

## export network as a 'gml' file that you can import into Cytoscape
## Not run: rr <- netinf2gml(object=res, file="predictionet_bayesnet")

predictionet documentation built on Nov. 8, 2020, 7:48 p.m.