minHG: Do a minimum hyper-geometric test

In Carldeboer/BrockmanR: K-mer-based analysis of chromatin state.

Description

Applies the minimum hypergeometric test to given data, returning the maximal enrichment achieved.

Usage

minHG(x, n_max = length(x) - 1)

Arguments

x	A sorted vector of binary values, where 1 is a "hit" (e.g. cognate k-mer) and 0 is a "miss" (e.g. non-cognate k-mer).
n_max	The maximum number of k-mers to consider for each minimum hypergeometric test. Defaults to testing all possible entries in the vector for maximal enrichment.

Value

Returns a list containing the lnMinP (natural log(min(p))), k (the number of vector entries included for maximal enrichment), log2OR (log2 odds ratio (observed hits/expected hits)), and n_max (as given to method).

Examples

kmerMat = inputKMerFreqs(sprintf("kMerFiles/%s.freq.gz",sampleDesc$id), IDs = sampleDesc$id)
myPCA = doKMerPCA(kmerMat, nPCs = "jackstraw")
treatmentPCs = findDistinguishingPCs(myPCA$rotation[,1:myPCA$nPCs], sampleDesc[c("id","treated")])
testResultsLow = minHG(as.logical(cisbp$binaryPBMZScores[order(pcs$rotation[row.names(cisbp$binaryPBMZScores),treatmentPCs$PC[1]]),"M0312_1.02.PBM"])); # lowly-weighted k-mers
testResultsHigh = minHG(as.logical(cisbp$binaryPBMZScores[order(pcs$rotation[row.names(cisbp$binaryPBMZScores),treatmentPCs$PC[1]],decreasing = T),"M0312_1.02.PBM"])); # highly-weighted k-mers

Carldeboer/BrockmanR documentation built on May 31, 2019, 2:19 p.m.