R/getTestsModTable.R

In montilab/K2Taxonomer: Creating annotated submodules of high-throughput data through recursive partitioning

#' Extract table of phenotypic variable tests from 'K2' object
#'
#' Create table phenotypic variable results from 'K2' object.
#' @param K2res An object of class K2 or K2results().
#' @return A data.frame object with the following columns:
#' \itemize{
#'  \item{value: }{The variable being tested}
#'  \item{node: }{The partition label}
#'  \item{edge: }{The subgroup for the given partition}
#'  \item{pval: }{Nominal p-value of test}
#'  \item{fdr: }{Benjamini-Hochberg FDR corrected p-value}
#'  \item{df: }{Degrees of freedom of test}
#'  \item{stat: }{Test statistic}
#'  \item{obsMean: }{Mean value across partition members}
#'  \item{altMean: }{Mean value for all other observations}
#'  \item{diffMean: }{Difference is mean}
#'  \item{nhits: }{The number of second label values in subgroup}
#'  \item{ncase: }{The total second-level label value}
#'  \item{nalt: }{The total first-level label value}
#'  \item{ndrawn: }{The total members in the subgroup}
#'  \item{hits: }{Members of subgroup with second-level label value}
#'  \item{test: }{The statistical test that produced this result}
#' }
#' @references
#'  \insertRef{reed_2020}{K2Taxonomer}
#'  \insertRef{bh}{K2Taxonomer}
#' @keywords clustering
#' @export
#' @examples
#' ## Read in ExpressionSet object
#' library(Biobase)
#' data(sample.ExpressionSet)
#'
#' ## Pre-process and create K2 object
#' K2res <- K2preproc(sample.ExpressionSet)
#'
#' ## Run K2 Taxonomer algorithm
#' K2res <- K2tax(K2res,
#'             stabThresh=0.5)
#'
#' infoClassVector <- c(sex='factor', score='numeric1')
#'
#' K2res <- runTestsMods(K2res, infoClass=infoClassVector)
#'
#' head(getTestsModTable(K2res))
#'

getTestsModTable <- function(K2res) {

    # Run checks
    .isK2(K2res)

    K2resList <- K2results(K2res)

    ## Format table
    K2modTestTable <- do.call(rbind, lapply(names(K2resList),
        function(x) {

            # Get phenotype results
            modTests <- K2resList[[x]]$modTests

            # Add edge ID to each subgroup
            modTests[[1]]$edge <- "1"
            modTests[[2]]$edge <- "2"

            # Concatenate
            modTests <- do.call(rbind, modTests)

            # Add node ID
            modTests$node <- x

            return(modTests)
        }))
    rownames(K2modTestTable) <- NULL

    ## Sort columns
    K2modTestTable <- K2modTestTable[, c("value", "node", "edge",
        "pval", "fdr", "stat", "df", "obsMean", "altMean", "diffMean",
        "nhits", "ncase", "nalt", "ndrawn", "hits", "test")]

    ## Sort by p-value
    K2modTestTable <- K2modTestTable[order(K2modTestTable$pval),
        ]

    return(K2modTestTable)
}