\name{MLP}
\alias{MLP}
\title{This function calculates p-values for each gene set based on row permutations
of the gene p values or column permutations of the expression matrix; the p values
can be obtained either as individual gene set p values or p values based on smoothing
across gene sets of similar size.}
\usage{
  MLP(geneSet, geneStatistic, minGenes = 5, maxGenes = 100,
    rowPermutations = TRUE, nPermutations = 100,
    smoothPValues = TRUE,
    probabilityVector = c(0.5, 0.9, 0.95, 0.99, 0.999, 0.9999, 0.99999),
    df = 9, addGeneSetDescription = TRUE)
}
\arguments{
  \item{geneSet}{is the input list of gene sets
  (components) and gene IDs (character vectors). A gene set
  can, for example, be a GO category with for each category
  Entrez Gene identifiers; The \link{getGeneSets} function
  can be used to construct the geneSet argument for
  different pathway sources.}

  \item{geneStatistic}{is either a named numeric vector (if
  rowPermutations is TRUE) or a numeric matrix of pvalues
  (if rowPermutations is FALSE). The names of the numeric
  vector or row names of the matrix should represent the
  gene IDs.}

  \item{minGenes}{minimum number of genes in a gene set for
  it to be considered (lower threshold for gene set size)}

  \item{maxGenes}{maximum number of genes in a gene set for
  it to be considered (upper threshold for gene set size)}

  \item{rowPermutations}{logical indicating whether to use
  row permutations (TRUE; default) or column permutations
  (FALSE)}

  \item{nPermutations}{is the number of simulations. By
  default 100 permutations are conducted.}

  \item{smoothPValues}{logical indicating whether one wants
  to calculate smoothed cut-off thresholds (TRUE; default)
  or not (FALSE).}

  \item{probabilityVector}{vector of quantiles at which p
  values for each gene set are desired}

  \item{df}{degrees of freedom for the smooth.spline
  function used in getSmoothedPValues}

  \item{addGeneSetDescription}{logical indicating whether a
  column with the gene set description be added to the
  output data frame; defaults to TRUE.}
}
\value{
  data frame with four (or five) columns: totalGeneSetSize,
  testedGeneSetSize, geneSetStatistic and geneSetPValue and
  (if addDescription is set to TRUE) geneSetDescription;
  the rows of the data frame are ordered by ascending
  geneSetPValue.
}
\description{
  This function calculates p-values for each gene set based
  on row permutations of the gene p values or column
  permutations of the expression matrix; the p values can
  be obtained either as individual gene set p values or p
  values based on smoothing across gene sets of similar
  size.
}
\examples{
if (require(GO.db)){
  pathExampleGeneSet <- system.file("exampleFiles", "exampleGeneSet.rda", package = "MLP")
  pathExamplePValues <- system.file("exampleFiles", "examplePValues.rda", package = "MLP")
  load(pathExampleGeneSet)
  load(pathExamplePValues)
  head(examplePValues)
  head(exampleGeneSet)
  mlpResult <- MLP(geneSet = exampleGeneSet, geneStatistic = examplePValues)
  head(mlpResult)
}
}
\references{
  Raghavan, Nandini et al. (2007). The high-level
  similarity of some disparate gene expression measures,
  Bioinformatics, 23, 22, 3032-3038.
}