ganet.hyperg: Hyper-geometric test for Overlapping Significance

In altayg/ganet: Global Assessment of NETwork Inference Algorithms Based on the Available Literature of Gene/Protein Interactions

Description

Determines a statistical p-value that shows the significance of the overlap between two networks.

Usage

  ganet.hyperg(nOverlap, nPredicted, nFocusedSet, nUniverse)

Arguments

nOverlap	The number of unique overlap.
nPredicted	The number of predcitions.
nFocusedSet	The number of elements (interactions) in the database of interest, regarding the genes of raw dataset.
nUniverse	The number of biologically possible (expected) interactions regarding the genes of raw dataset .

Details

This may be used for instance to search for whether the number of predicted gene/protein interactions (nPredicted) are overlapping in a biological interaction database of interest with a significant p-value.

Value

It results an environment 'res' with two output: res$stats and res$precision.

stats	All the statistical results from hyper-geometric test.
precision	The rate of accurate predictions over all the predicted links.

Note

This package is intended to be updated with respect to feedbacks and also with updates of the current databases.

Author(s)

Gokmen Altay

References

1. G. Altay, N. Altay, D.E. Neal, Global assessment of network inference algorithms based on available literature of gene/protein interactions, Turk J Biol, 37:547-555, 2013. 2. G. Altay, J Zapardel, B Peters, "Determining preprocessing combinations and sample size for Gene Network Inference over ...", bioArxiv, 2017.

See Also

ganet.ComLinks, ganet.combine, ganet.UniqNetSimp

Examples

 # data(ganet.ex.net)  # assume this is the predicted protein interactions
#  data(hprd)          # This is the HPRD database to search for in it.
  
#  Validated <- ganet.ComLinks(netlist=as.matrix(ganet.ex.net), netdata=as.matrix(hprd))
  
#  data(ganet.ex.genes)  # load the example unique genes of the example 
#row dataset to compute the number of universe
#  nUniverse <- length(ganet.ex.genes)*(length(ganet.ex.genes) - 1)/2 
##all possible interactions with these genes. 
#  #This is the upper-bound and over-optimistic.
#We recommend to take no more than half of it.

#  nUniverse <- nUniverse/2
  
#  #nFocusedSet <- length( union(which(hprd[,1] %in% ganet.ex.genes),
                    #which(hprd[,2]%in% ganet.ex.genes)) )
#  nFocusedSet <- 41398  # pre-computed using above.
  
#  res <- ganet.hyperg(nOverlap=nrow(Validated), nPredicted=nrow(ganet.ex.net), 
#nFocusedSet=nFocusedSet, nUniverse=nUniverse)
  
#  res$p.val.hyperg
#  res$precision

altayg/ganet documentation built on May 11, 2019, 11:29 p.m.