run.lean: Run the LEAN approach

Description

Apply the LEAN approach to a given network and a list of pvalues

Usage

run.lean(ranking, network, ranked = F,  add.scored.genes = F, keep.nodes.without.scores = F,  verbose = F, n_reps = 10000, bootstrap = F, ncores = NULL)

Arguments

ranking

Either a file containing gene p-values or a named numerical vector of p-values with names matching node names used in the network

network

Either a file containing the network in sif format or an igraph graph object representing the network

ranked

whether to transform input p-values into a uniformly distributed list of p-values based on the genes' rank before p* calculation

add.scored.genes

whether to create one singleton node for each gene with a score but not occurring in the graph

keep.nodes.without.scores

whether to keep nodes of the graph that have no recorded score. For those nodes it is still possible to compute enrichment scores if at least one of their network neighbors has a recorded score.

verbose

whether to print additional status messages

n_reps

the number of samples each background distribution should consist of. Largely influences the run-time, but higher values needed for meaningful empirical pvalues !

bootstrap

whether to draw the pvalues of the background distributions with or without replacement

ncores

number of cores to be used in parallel computation. Default (NULL) leads to automatic guessing of max number of cores to be used (depending on operating system).

Value

A list object containing the results of the LEAN run. The list encompasses the following elements: A list object containing the results of the LEAN run. The list encompasses the following elements:

References

Gwinner et al., Network-based analysis of omics data: The LEAN method, MS submitted to Bioinformatics

Examples

Run this code

## Simple use case starting from a test network and p-value list
## Not run:  
# # compute LEAN p-values starting from a p-value file and a network file
# rank_file<-system.file('extdata/pvals_red.txt.gz', package='LEANR')
# net_file<-system.file('extdata/g_red.sif.gz', package='LEANR')
# system.time(res<-run.lean(ranking=rank_file, network=net_file, 
#   add.scored.genes=T, verbose=T, n_reps=1000, ncores=3))
# 
# # compute LEAN p-values starting from a list of gene scores and a graph
# data(pvals_red)
# data(g_red)
# system.time(res2<-run.lean(ranking=pvals_red, network=g_red, verbose=T, 
#    n_reps=1000, ncores=3))
# 
# ## End(Not run)