nem: Nested Effects Models - main function

Description

The main function to perform model learning from data

Usage

nem(D,inference="nem.greedy",models=NULL,control=set.default.parameters(setdiff(unique(colnames(D)),"time")), verbose=FALSE)
"print"(x, ...)

Arguments

data matrix with experiments in the columns (binary or continious)

inference

search to use exhaustive enumeration, triples for triple-based inference, pairwise for the pairwise heuristic, ModuleNetwork for the module based inference proposed in Fr\"ohlich et al. 2008, ModuleNetwork.orig for the module based inference proposed in Fr\"ohlich et al. 2007, nem.greedy for greedy hillclimbing, nem.greedyMAP for alternating MAP optimization using log odds or log p-value densities, mc.eminem for EM based inference using log odds or log p-value densities, BN.greedy, BN.exhaustive for a conventional Bayesian Network treatment using binomial or normal distribution assumptions, dynoNEM for MCMC based inference from time series data, mc.eminem for EM based inference

models

a list of adjacency matrices for model search. If NULL, an exhaustive enumeration of all possible models is performed.

control

list of parameters: see set.default.parameters

verbose

do you want to see progression statements? Default: TRUE

nem object

...

other arguments to pass

Value

graph: inferred directed S-gene graph (graphNEL object)
mLL: log posterior marginal likelihood of model(s)
pos: posterior over effect positions
mappos: MAP estimate of effect positions
selected: selected E-gene subset
LLperGene: likelihood per selected E-gene
avg: in case of MCMC: posterior mean S-gene graph (edge weighted adjacency matrix)
control: hyperparameter as in function call
local.maxima: local maxima of the EM procedure
graphs.sampled: sampled graphs
EB: samples of the empirical Bayes prior
acc_list: list that indicates whether the corresponding sampled S-gene graph has been accepted (new local maximum (1), same local maximum (0)) or rejected(-1) in the MCMC sampling process - length(acc_list)=mcmc.nsamples + mcmc.nburnin

Details

If parameter Pm != NULL and parameter lambda == 0, a Bayesian approach to include prior knowledge is used. Alternatively, the regularization parameter lambda can be tuned in a model selection step via the function nemModelSelection using the BIC criterion. If automated subset selection of effect reporters is used (default), the regularization parameter delta can be tuned via the BIC model selection criterion. Per default it is fixed to 1 / (no. S-genes + 1).

The function plot.nem plots the inferred phenotypic hierarchy as a directed graph, the likelihood distribution of the models (only for exhaustive search) or the posterior position of the effected genes.

References

Markowetz, F.; Bloch, J. & Spang, R., Non-transcriptional Pathway Features Reconstructed from Secondary Effects of RNA interference. Bioinformatics, 2005, 21, 4026 - 4032\

Markowetz, F.; Kostka, D.; Troyanskaya, O. & Spang, R., Nested Effects Models for High-dimensional Phenotyping Screens. Bioinformatics, 2007, 23, i305 - i312\

Fr\"ohlich, H.; Fellmann, M.; S\"ultmann, H.; Poustka, A. & Beissbarth, T. Large Scale Statistical Inference of Signaling Pathways from RNAi and Microarray Data. BMC Bioinformatics, 2007, 8, 386\

Fr\"ohlich, H.; Fellmann, M.; S\"ultmann, H.; Poustka, A. & Beissbarth, T. Estimating Large Scale Signaling Networks through Nested Effect Models with Intervention Effects from Microarray Data. Bioinformatics, 2008, 24, 2650-2656\

Tresch, A. & Markowetz, F., Structure Learning in Nested Effects Models Statistical Applications in Genetics and Molecular Biology, 2008, 7\

Zeller, C.; Fr\"ohlich, H. & Tresch, A., A Bayesian Network View on Nested Effects Models EURASIP Journal on Bioinformatics and Systems Biology, 2009, 195272\

Fr\"ohlich, H.; Tresch, A. & Beissbarth, T., Nested Effects Models for Learning Signaling Networks from Perturbation Data. Biometrical Journal, 2009, 2, 304 - 323\

Fr\"ohlich, H.; Sahin, \"O.; Arlt, D.; Bender, C. & Beissbarth, T. Deterministic Effects Propagation Networks for Reconstructing Protein Signaling Networks from Multiple Interventions. BMC Bioinformatics, 2009, 10, 322\

Fr\"ohlich, H.; Praveen, P. & Tresch, A., Fast and Efficient Dynamic Nested Effects Models. Bioinformatics, 2011, 27, 238-244\

Niederberger, T.; Etzold, S.; Lidschreiber, M; Maier, K.; Martin, D.; Fr\"ohlich, H.; Cramer, P.; Tresch, A., MC Eminem Maps the Interaction Landscape of the Mediator, PLoS Comp. Biol., 8(6): e1002568, 2012.

Examples

Run this code

   data("BoutrosRNAi2002")
   D <- BoutrosRNAiDiscrete[,9:16]
   control = set.default.parameters(unique(colnames(D)), para=c(0.13, 0.05))   
   res1 <- nem(D,inference="search", control=control)
   res2 <- nem(D,inference="pairwise", control=control)
   res3 <- nem(D,inference="triples", control=control)
   res4 <- nem(D,inference="ModuleNetwork", control=control)
   res5 <- nem(D,inference="nem.greedy", control=control)        
   res6 = nem(BoutrosRNAiLods, inference="nem.greedyMAP", control=control)
   

   par(mfrow=c(2,3))
   plot.nem(res1,main="exhaustive search")
   plot.nem(res2,main="pairs")
   plot.nem(res3,main="triples")
   plot.nem(res4,main="module network")
   plot.nem(res5,main="greedy hillclimber")      
   plot.nem(res6,main="alternating MAP optimization")

Run the code above in your browser using DataLab