R/runScrParal.R

In IFAA: Robust Inference for Absolute Abundance in Microbiome Analysis

#---------------------------------------------------------------------------------------
## parallel screen function to select independent reference taxa
#---------------------------------------------------------------------------------------

runScrParal <- function(data,
                        testCovInd,
                        testCovInOrder,
                        testCovInNewNam,
                        nRef,
                        paraJobs,
                        refTaxa,
                        maxDimensionScr = 0.8 * 434 * 10 * 10^4,
                        sequentialRun,
                        allFunc = allFunc,
                        refReadsThresh,
                        SDThresh,
                        SDquantilThresh,
                        balanceCut,
                        Mprefix,
                        covsPrefix,
                        binPredInd,
                        adjust_method) {
  results <- list()
  # load data info
  basicInfo <- dataInfo(
    data = data,
    qualifyRefTax = TRUE,
    refReadsThresh = refReadsThresh,
    SDThresh = SDThresh,
    SDquantilThresh = SDquantilThresh,
    balanceCut = balanceCut,
    Mprefix = Mprefix,
    covsPrefix = covsPrefix,
    binPredInd = binPredInd
  )

  taxaNames <- basicInfo$taxaNames
  nTaxa <- basicInfo$nTaxa
  nPredics <- basicInfo$nPredics
  nSub <- basicInfo$nSub
  predNames <- basicInfo$predNames

  results$goodRefTaxaCandi <- basicInfo$goodRefTaxaCandi
  rm(basicInfo)
  gc()

  nNorm <- nTaxa - 1
  nAlphaNoInt <- nPredics * nNorm
  nAlphaSelec <- nPredics * nTaxa

  
  if (nRef > (length(taxaNames))) {
    nRef <- length(taxaNames)
    }

  # make reference taxa list
  if (length(refTaxa) < nRef) {
    taxon_to_be_sample <-
      (results$goodRefTaxaCandi[!(results$goodRefTaxaCandi %in% refTaxa)])
    num_to_be_sample <- (nRef - length(refTaxa))
    if (num_to_be_sample >= length(taxon_to_be_sample)) {
      num_to_be_sample <- length(taxon_to_be_sample)
    }

    refTaxa_extra <- sample(taxon_to_be_sample, num_to_be_sample)
    refTaxa <- c(refTaxa, refTaxa_extra)

    if (length(refTaxa) < nRef) {
      taxa_to_be_sample2 <- (taxaNames[!(taxaNames %in% refTaxa)])
      num_to_be_sample2 <- (nRef - length(refTaxa))
      refTaxa_extra2 <-
        sample(taxa_to_be_sample2, num_to_be_sample2)
      refTaxa <- c(refTaxa, refTaxa_extra2)
      message("Reference taxa are randomly picked in Phase 1.")
    }
  } else if (length(refTaxa) > nRef) {
    refTaxa <- sample(refTaxa, nRef)
  }

  results$refTaxa <- refTaxa


  #
  ## run original data screen
  #
  gc(reset = TRUE)

  if (suppressMessages(dataSparsCheck(data, Mprefix))[[1]] == 0) {
    screen1 <- originDataScreen_imputed(
      data = data,
      testCovInd = testCovInd,
      nRef = nRef,
      refTaxa = refTaxa,
      paraJobs = paraJobs,
      allFunc = allFunc,
      Mprefix = Mprefix,
      covsPrefix = covsPrefix,
      binPredInd = binPredInd,
      sequentialRun = sequentialRun
    )
  } else {
    screen1 <- originDataScreen(
      data = data,
      testCovInd = testCovInd,
      nRef = nRef,
      refTaxa = refTaxa,
      paraJobs = paraJobs,
      allFunc = allFunc,
      Mprefix = Mprefix,
      covsPrefix = covsPrefix,
      binPredInd = binPredInd,
      sequentialRun = sequentialRun,
      adjust_method = adjust_method
    )
  }


  maxMemUsedInMb <- sum(gc()[, 6])

  results$countOfSelecForAPred <- screen1$countOfSelecForAPred
  results$estOfSelectForAPred <- screen1$estOfSelectForAPred
  results$testCovCountMat <- screen1$testCovCountMat
  results$testEstMat <- screen1$testEstMat

  rm(screen1)
  gc()

  nTestCov <- length(testCovInd)
  results$nTestCov <- nTestCov
  results$nTaxa <- nTaxa
  results$nPredics <- nPredics

  results$taxaNames <- taxaNames
  rm(taxaNames)
  return(results)
}