R/internPostProcess.R

In cbg-ethz/TiMEx: Finding Mutually Exclusive Groups of Alterations in Cancer Datasets

# Author: Simona Cristea; scristea@@jimmy.harvard.edu

############################################################################## 
getFreqsForCliquesAfterTest <- function(cliqueIdxList, matrix) {
    # (internal) function to compute the frequencies of the genes part of a group
    
    # inputs: cliqueIdxList: a list of indices of significant groups, as part of the signifGroups structure,
    # returned by TiMEx. 'fdr' and 'bonf' represent the different correction methods used for the testing of
    # groups matrix: binary input matrix of alterations
    
    # outputs: freqs: a list with two elements, 'fdr' and 'bonf' (corresponding to the respective multiple
    # correction method), each consisting of the vector frequencies of the genes in the group
    
    l <- length(cliqueIdxList)
    freqs <- list()
    for (i in 2:l) {
        freqs[[i]] <- list()
        
        cliqueIdx.bonf <- (cliqueIdxList[[i]])$bonf
        if (length(dim(cliqueIdx.bonf)[1]) > 0) 
            freqs[[i]]$bonf <- apply(cliqueIdx.bonf, 2, function(x) {
                apply(matrix[, x], 2, sum)/dim(matrix)[1] * 100
            }) else freqs[[i]]$bonf <- sum(matrix[, cliqueIdx.bonf])/(dim(matrix)[1] * 100)
        
        
        cliqueIdx.fdr <- (cliqueIdxList[[i]])$fdr
        if (length(dim(cliqueIdx.fdr)[1]) > 0) 
            freqs[[i]]$fdr <- apply(cliqueIdx.fdr, 2, function(x) {
                apply(matrix[, x], 2, sum)/dim(matrix)[1] * 100
            }) else freqs[[i]]$fdr <- sum(matrix[, cliqueIdx.fdr])/(dim(matrix)[1] * 100)
    }
    return(freqs)
}



############################################################################## 
combineNameWithFreq <- function(genesSignifMat, ffMat, muMat, pvalMat) {
    # (internal) function to adjacently print the name of the genes in a group, their frequency, the estimated mu,
    # and the pvalue of the group
    
    # inputs: genesSignifMat: matrix in which each row consists of the gene names (as characters) of one
    # identified significant maximal clique of a given size. there are as many rows as identified maximal cliques
    # ffMat: matrix with same structure as genesSignifMat, consisting however of positive real numbers between 0
    # and 100, representing the frequency in the input dataset of each gene present in genesSignifMat muMat:
    # vector with as many elements as identified maximal cliques, consisting of the estimated mu values for each
    # maximal clique (real values between 0 and 1) pvalMat: vector with as many elements as identified maximal
    # cliques, consisting of the adjusted pvalue for each maximal clique (real values between 0 and 1)
    
    # outputs: newMat: matrix with as many rows as identified maximal clique.  each row consists of the names of
    # the genes part of each clique (followed, in brackets, by their frequency in the input dataset), the
    # intensity of mutual exclusivity of the group, and the corrected p-value of the group
    
    d <- dim(genesSignifMat)[1]
    if (!is.null(d) && !d == 0) {
        newMat <- array(NA, dim = c(d, (dim(genesSignifMat)[2]) + 2))
        for (j in 1:d) {
            newMat[j, ] <- c(paste(genesSignifMat[j, ], " (", round(ffMat[j, ], 2), "%)", sep = ""), round(muMat[j], 
                2), round(pvalMat[j], 6))
        }
    } else {
        newMat <- rep(NA, (length(genesSignifMat) + 2))
        newMat <- c(paste(genesSignifMat, " (", round(ffMat, 2), "%)", sep = ""), round(muMat, 2), round(pvalMat, 
            6))
    }
    
    return(newMat)
}