test_MIGSA.R
In MIGSA: Massive and Integrative Gene Set Analysis

# library(covr);
library(RUnit)
library(MIGSA)
# MIGSA_coverage <- package_coverage("MIGSA"); MIGSA_coverage
# shine(MIGSA_coverage);
# BiocGenerics:::testPackage("MIGSA")

rOpts <- getOption("RUnit")
rOpts$silent <- !FALSE
options("RUnit" = rOpts)
library(Biobase)
hasInternet <- testBioCConnection()
https_enabled <- !inherits(
  try(readLines(url("https://www.bioconductor.org"))[1]),
  "try-error"
)
testAll <- FALSE
library(BiocParallel)
if (.Platform$OS.type == "unix") {
  bp_param <- MulticoreParam(workers = 1)
} else if (.Platform$OS.type == "windows") {
  bp_param <- SnowParam(workers = 1)
}
register(bp_param)
######## FitOptions tests

###### FitOptions-class tests

# It doesnt have any function to test

###### FitOptions-constructor tests

#### Correct ones

test_FitOptions.default_ok <- function() {
  conditions <- c(rep("C1", 4), rep("C2", 7))

  checkTrue(validObject(FitOptions(conditions)))
}

test_FitOptions.data.frame_ok <- function() {
  myData <- data.frame(cond = c(rep("C1", 4), rep("C2", 7)))
  myFormula <- ~ cond - 1
  myContrast <- c(-1, 1)

  checkTrue(validObject(FitOptions(myData, myFormula, myContrast)))
}

test_FitOptions_bothConstructorEqual <- function() {
  conditions <- as.factor(c(rep("C1", 4), rep("C2", 7)))
  myData <- data.frame(cond = conditions)
  myFormula <- ~ cond - 1
  myContrast <- c(-1, 1)

  fitDefault <- FitOptions(conditions)
  fitDataFrame <- FitOptions(myData, myFormula, myContrast)
  checkEquals(fitDefault, fitDataFrame)
}

#### Incorrect ones

test_FitOptions.default_wrong_oneCond <- function() {
  conditions <- rep("C1", 1)

  checkException(FitOptions(conditions))
}

test_FitOptions.default_wrong_threeCond <- function() {
  conditions <- c(rep("C1", 4), rep("C2", 7), rep("C3", 1))

  checkException(FitOptions(conditions))
}

###### FitOptions tests

# It doesnt have any exported function to test

###### Genesets-geneSetsFromFile tests

#### Correct ones

test_Genesets_geneSetsFromFile_ok_simpleFile <- function() {
  # create gene sets file
  genes <- paste("gene", 1:(10 * 20))
  gsets <- data.frame(
    IDs = paste("set", 1:10),
    Names = rep("", 10),
    matrix(genes, nrow = 10, byrow = TRUE)
  )

  geneSetsFile <- paste(tempdir(), "/fakeGsets.tsv", sep = "")
  write.table(gsets,
    file = geneSetsFile, sep = "\t",
    col.names = FALSE, row.names = FALSE, quote = FALSE
  )

  # Now lets load this tsv file as a Genesets object.
  myGsets <- geneSetsFromFile(geneSetsFile)
  myGsetsGO <- geneSetsFromFile(geneSetsFile, is_GO = TRUE)

  # Lets delete this tsv file
  unlink(geneSetsFile)

  library(GSEABase)

  checkTrue(all(unlist(lapply(myGsetsGO, function(x) {
    is(collectionType(x), "GOCollection")
  }))))
  checkTrue(all(unlist(lapply(myGsets, function(x) {
    is(collectionType(x), "NullCollection")
  }))))
  checkEquals(length(myGsets), 10)
  checkEquals(names(MIGSA:::asList(myGsets)), as.character(gsets$IDs))
  checkTrue(all(unlist(MIGSA:::asList(myGsets)) == genes))
}

#### Incorrect ones

test_Genesets_geneSetsFromFile_wrong_emptyPath <- function() {
  filePath <- ""
  name <- "myGenesets"

  checkException(
    geneSetsFromFile(filePath, name)
  )
}

test_Genesets_geneSetsFromFile_wrong_dirPath <- function() {
  filePath <- list.dirs("..")[[2]]
  name <- "myGenesets"

  checkException(
    geneSetsFromFile(filePath, name)
  )
}

test_Genesets_geneSetsFromFile_wrong_wrongPath <- function() {
  filePath <- "./notExistingPath/notExistingFile.csv"
  name <- "myGenesets"

  checkException(
    geneSetsFromFile(filePath, name)
  )
}

###### Genesets-asGenesets tests

#### Correct ones

test_Genesets_asGenesets_ok_complete <- function() {
  library(GSEABase)

  myGs1 <- GeneSet(as.character(1:10), setName = "fakeId1", setIdentifier = "")
  myGs2 <- GeneSet(as.character(7:15), setName = "fakeId2", setIdentifier = "")
  myGs3 <- GeneSet(as.character(20:28), setName = "fakeId3", setIdentifier = "")

  constGsets <- GeneSetCollection(list(myGs1, myGs2, myGs3))

  listGsets <- list(geneIds(myGs1), geneIds(myGs2), geneIds(myGs3))
  names(listGsets) <- c(setName(myGs1), setName(myGs2), setName(myGs3))

  asGsets <- as.Genesets(listGsets)

  checkEquals(constGsets, asGsets)
}

test_Genesets_asGenesets_wrong_emptyGsetRemoved <- function() {
  library(GSEABase)
  myGs1 <- as.character(1:10)
  myGs2 <- ""
  myGs3 <- as.character(20:28)

  listGsets <- list(myGs1, myGs2, myGs3)
  names(listGsets) <- c("myGs1", "myGs2", "myGs3")

  asGsets <- as.Genesets(listGsets)
  checkEqualsNumeric(length(asGsets), 2)

  fstGset <- asGsets[[1]]
  sndGset <- asGsets[[2]]

  checkEquals(setName(fstGset), "myGs1")
  checkEquals(setName(sndGset), "myGs3") # myGs2 must be deleted

  checkEquals(geneIds(fstGset), myGs1)
  checkEquals(geneIds(sndGset), myGs3) # myGs2 must be deleted
}

#### Incorrect ones

test_Genesets_asGenesets_wrong_noIds <- function() {
  myGs1 <- as.character(1:10)
  myGs2 <- as.character(7:15)

  listGsets <- list(myGs1, myGs2)

  checkException(
    as.Genesets(listGsets)
  )
}

test_Genesets_asGenesets_wrong_repeatedIds <- function() {
  myGs1 <- as.character(1:10)
  myGs2 <- as.character(7:15)

  listGsets <- list(myGs1, myGs2)
  names(listGsets) <- c("myGs1", "myGs1")

  checkException(
    as.Genesets(listGsets)
  )
}

test_Genesets_asGenesets_wrong_noGenes <- function() {
  myGs1 <- ""
  myGs2 <- ""

  listGsets <- list(myGs1, myGs2)
  names(listGsets) <- c("myGs1", "myGs2")

  checkException(
    as.Genesets(listGsets)
  )
}

###### Genesets-enrichrGeneSets tests

if (testAll) {
  test_Genesets_enrichrGeneSets_ok_wellListed <- function() {
    # if we dont have internet then dont fail the test
    library(Biobase)
    if (!hasInternet || !https_enabled) {
      return(checkTrue(TRUE))
    }

    enrichrList <- enrichrGeneSets()

    checkTrue(is(enrichrList, "data.frame"))
    checkTrue(ncol(enrichrList) > 0)
    checkTrue(nrow(enrichrList) > 0)
  }
}

###### Genesets-downloadEnrichrGeneSets tests

#### Correct ones

if (testAll) {
  test_Genesets_downloadEnrichrGeneSets_ok_goCC <- function() {
    # if we dont have internet then dont fail the test
    library(Biobase)
    if (!hasInternet || !https_enabled) {
      return(checkTrue(TRUE))
    }

    goCcName <- "GO_Cellular_Component_2013"
    goCc <- downloadEnrichrGeneSets(goCcName)

    checkEquals(length(goCc), 1)
    checkEquals(names(goCc), goCcName)

    checkEquals(names(goCc)[[1]], goCcName)
    goCc <- goCc[[1]]
    #         checkTrue(goCc@is_GO);
    checkTrue(length(goCc) > 0)
  }
}

if (testAll) {
  test_Genesets_downloadEnrichrGeneSets_ok_kegg <- function() {
    # if we dont have internet then dont fail the test
    library(Biobase)
    if (!hasInternet || !https_enabled) {
      return(checkTrue(TRUE))
    }

    bioCartaName <- "BioCarta_2015"
    bioCarta <- downloadEnrichrGeneSets(bioCartaName)

    checkEquals(length(bioCarta), 1)
    checkEquals(names(bioCarta), bioCartaName)

    bioCarta <- bioCarta[[1]]
    #         checkTrue(!kegg@is_GO);
    checkTrue(length(bioCarta) > 0)
  }
}

if (testAll) {
  test_Genesets_downloadEnrichrGeneSets_ok_keggGoCC <- function() {
    # if we dont have internet then dont fail the test
    library(Biobase)
    if (!hasInternet || !https_enabled) {
      return(checkTrue(TRUE))
    }

    libNames <- c("KEGG_2016", "GO_Cellular_Component_2013")
    libs <- downloadEnrichrGeneSets(libNames)

    checkEquals(length(libs), 2)
    checkEquals(names(libs), libNames)
    kegg <- libs[[1]]
    goCc <- libs[[2]]

    checkEquals(names(libs)[[1]], libNames[[1]])
    #         checkTrue(!kegg@is_GO);
    checkTrue(length(kegg) > 0)

    checkEquals(names(libs)[[2]], libNames[[2]])
    #         checkTrue(goCc@is_GO);
    checkTrue(length(goCc) > 0)
  }
}

if (testAll) {
  test_Genesets_downloadEnrichrGeneSets_ok_keggOneFakeLib <- function() {
    # if we dont have internet then dont fail the test
    library(Biobase)
    if (!hasInternet || !https_enabled) {
      return(checkTrue(TRUE))
    }

    libNames <- c("KEGG_2013", "fakeLib")
    libs <- downloadEnrichrGeneSets(libNames)

    checkEquals(length(libs), 1)
    checkEquals(names(libs), libNames[[1]])

    checkEquals(names(libs)[[1]], libNames[[1]])
    kegg <- libs[[1]]
    #         checkTrue(!kegg@is_GO);
    checkTrue(length(kegg) > 0)
  }
}

#### Incorrect ones

if (testAll) {
  test_Genesets_downloadEnrichrGeneSets_wrong_noLibs <- function() {
    # if we dont have internet then dont fail the test
    library(Biobase)
    if (!hasInternet || !https_enabled) {
      return(checkTrue(TRUE))
    }

    libNames <- ""
    checkTrue(
      length(downloadEnrichrGeneSets(libNames)) == 0
    )
  }
}

if (testAll) {
  test_Genesets_downloadEnrichrGeneSets_wrong_fakeLibs <- function() {
    # if we dont have internet then dont fail the test
    library(Biobase)
    if (!hasInternet || !https_enabled) {
      return(checkTrue(TRUE))
    }

    libNames <- c("fakeLib1", "fakeLib2")
    checkTrue(
      length(downloadEnrichrGeneSets(libNames)) == 0
    )
  }
}

###### Genesets-loadGo tests

#### Correct ones

if (testAll) {
  test_Genesets_loadGo_ok_cc <- function() {
    ccName <- "CC"
    goCc <- loadGo(ccName)

    #         checkEquals(goCc@name, ccName);
    #         checkTrue(goCc@is_GO);
    checkTrue(length(goCc) > 0)
  }
}

#### Incorrect ones

test_Genesets_loadGo_wrong_kegg <- function() {
  checkException(
    loadGo("KEGG")
  )
}

###### Genesets tests

# It doesnt have any exported function to test

######## GSEAparams tests

###### GSEAparams-class tests

#### Correct ones

test_GSEAparams_ok_default <- function() {
  checkTrue(validObject(GSEAparams()))
}

#### Incorrect ones

test_GSEAparams_wrong_1perms <- function() {
  checkException(
    GSEAparams(perm_number = 1)
  )
}

###### GSEAparams tests

# It doesnt have any exported function to test

######## SEAparams tests

###### SEAparams-class tests

#### Correct ones

test_SEAparams_ok_default <- function() {
  checkTrue(validObject(SEAparams()))
}

test_SEAparams_ok_bri <- function() {
  checkTrue(validObject(SEAparams(br = "bri")))
}

test_SEAparams_ok_briii <- function() {
  checkTrue(validObject(SEAparams(br = "briii")))
}

test_SEAparams_ok_ownbr <- function() {
  checkTrue(validObject(SEAparams(
    br = as.character(1:10)
  )))
}

test_SEAparams_ok_over1Lfc <- function() {
  checkTrue(validObject(SEAparams(
    treat_lfc = 1.1
  )))
}

#### Incorrect ones

test_SEAparams_wrong_negLfc <- function() {
  checkException(
    SEAparams(treat_lfc = -1)
  )
}

test_SEAparams_wrong_negDe <- function() {
  checkException(
    SEAparams(de_cutoff = -1)
  )
}

test_SEAparams_wrong_over1De <- function() {
  checkException(
    SEAparams(de_cutoff = 1.1)
  )
}

test_SEAparams_wrong_fakeAdj <- function() {
  checkException(
    SEAparams(adjust_method = "fakeAdj")
  )
}

test_SEAparams_wrong_emptybr <- function() {
  checkException(
    SEAparams(br = "")
  )
}

test_SEAparams_wrong_twoEmptybr <- function() {
  checkException(
    SEAparams(br = c("", ""))
  )
}

###### SEAparams tests

# It doesnt have any exported function to test

######## GenesetRes tests

###### GenesetRes tests

# It doesnt have any exported function to test

######## GSEAres tests

###### GSEAres tests

# It doesnt have any exported function to test

######## SEAres tests

###### SEAres tests

# It doesnt have any exported function to test

######## GenesetsRes tests

###### GenesetsRes tests

# It doesnt have any exported function to test

######## IGSAinput tests

###### IGSAinput-class tests

#### Correct ones

test_IGSAinput_ok_complete <- function() {
  name <- "myIgsaInput"
  nSamples <- 4
  nGenes <- 10
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  checkTrue(
    validObject(IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts
    ))
  )
}

test_IGSAinput_ok_oneGSet <- function() {
  library(GSEABase)

  name <- "myIgsaInput"
  nSamples <- 4
  nGenes <- 10
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  myGs <- GeneSet(as.character(1:10),
    setName = "fakeName1",
    setIdentifier = "fakeId1"
  )

  gsets <- list(
    myGenesets =
      GeneSetCollection(list(myGs))
  )

  checkTrue(
    validObject(IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts,
      gene_sets_list = gsets
    ))
  )
}

test_IGSAinput_ok_twoGSets <- function() {
  library(GSEABase)

  name <- "myIgsaInput"
  nSamples <- 4
  nGenes <- 10
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  gsName1 <- "myGenesets1"
  gsName2 <- "myGenesets2"
  myGs <- GeneSet(as.character(1:10),
    setName = "fakeName1",
    setIdentifier = "fakeId1"
  )

  gsets <- list(
    myGenesets1 = GeneSetCollection(list(myGs)),
    myGenesets2 = GeneSetCollection(list(myGs))
  )

  checkTrue(
    validObject(IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts,
      gene_sets_list = gsets
    ))
  )
}

#### Incorrect ones

test_IGSAinput_wrong_noName <- function() {
  nSamples <- 4
  nGenes <- 10
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  checkException(
    IGSAinput(
      expr_data = exprData,
      fit_options = fitOpts
    )
  )
}

test_IGSAinput_wrong_emptyName <- function() {
  name <- ""
  nSamples <- 4
  nGenes <- 10
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  checkException(
    IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts
    )
  )
}

test_IGSAinput_wrong_oneSubj <- function() {
  name <- "myIgsaInput"
  nSamples <- 1
  nGenes <- 10
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c("C1", "C2"))

  checkException(
    IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts
    )
  )
}

test_IGSAinput_wrong_oneGene <- function() {
  name <- "myIgsaInput"
  nSamples <- 4
  nGenes <- 1
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  checkException(
    IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts
    )
  )
}

test_IGSAinput_wrong_badFitExprData <- function() {
  name <- "myIgsaInput"
  nSamples <- 4
  nGenes <- 10

  # one more subject in exprData than in fit
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples + 1, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  checkException(
    IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts
    )
  )
}

test_IGSAinput_wrong_repGSets <- function() {
  library(GSEABase)
  name <- "myIgsaInput"
  nSamples <- 4
  nGenes <- 10
  exprData <- new("MAList", list(M = matrix(0, ncol = nSamples, nrow = nGenes)))
  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))

  myGs <- GeneSet(as.character(1:10),
    setIdentifier = "fakeId1",
    setName = "fakeName1"
  )

  gsets <- list(
    myGenesets = GeneSetCollection(list(myGs)),
    myGenesets = GeneSetCollection(list(myGs))
  )

  checkException(
    IGSAinput(
      name = name,
      expr_data = exprData,
      fit_options = fitOpts,
      gene_sets_list = gsets
    )
  )
}

###### IGSAinput-getterSetters tests

test_IGSAinput_getterSetters_ok <- function() {
  exprData1 <- new("MAList", list(M = matrix(0, ncol = 6, nrow = 6)))
  fitOpts <- FitOptions(c(1, 1, 1, 2, 2, 2))
  igsaInput <- IGSAinput(
    name = "fakeName",
    expr_data = exprData1,
    fit_options = fitOpts
  )

  checkEquals(name(igsaInput), "fakeName")
  name(igsaInput) <- "fakeName2"
  checkEquals(name(igsaInput), "fakeName2")

  checkEquals(fitOptions(igsaInput), fitOpts)
  fitOpts2 <- FitOptions(c(1, 1, 1, 1, 1, 2))
  fitOptions(igsaInput) <- fitOpts2
  checkEquals(fitOptions(igsaInput), fitOpts2)

  checkEquals(exprData(igsaInput), exprData1)
  exprData2 <- new("MAList", list(M = matrix(1, ncol = 6, nrow = 6)))
  exprData(igsaInput) <- exprData2
  checkEquals(exprData(igsaInput), exprData2)

  checkEquals(length(MIGSA::geneSetsList(igsaInput)), 0)
  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) {
    sample(as.character(1:100),
      size = 10
    )
  })
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)
  geneSetsList(igsaInput) <- list(myGeneSets = myGSs)
  checkEquals(length(MIGSA::geneSetsList(igsaInput)), 1)

  checkEquals(gseaParams(igsaInput), GSEAparams())
  gseaParams(igsaInput) <- GSEAparams(perm_number = 10)
  checkEquals(gseaParams(igsaInput), GSEAparams(perm_number = 10))

  checkEquals(seaParams(igsaInput), SEAparams())
  seaParams(igsaInput) <- SEAparams(treat_lfc = 1)
  checkEquals(seaParams(igsaInput), SEAparams(treat_lfc = 1))
}

###### IGSAinput-getDEGenes tests

test_IGSAinput_getDEGenes_ok_deGenes <- function() {
  name <- "myIgsaInput"
  nSamples <- 40
  nGenes <- 100

  set.seed(8818)
  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  # lets make some DE genes
  exprData[1:5, 1:(nSamples / 2)] <-
    matrix(6 * abs(rnorm(5 * nSamples / 2)), ncol = nSamples / 2)
  rownames(exprData) <- 1:nrow(exprData)
  exprData <- new("MAList", list(M = exprData))

  fitOpts <- FitOptions(c(rep("C1", nSamples / 2), rep("C2", nSamples / 2)))
  igsaInput <- IGSAinput(
    name = name,
    expr_data = exprData,
    fit_options = fitOpts
  )

  newIgsaInput <- getDEGenes(igsaInput)
  deGenes <- seaParams(newIgsaInput)@de_genes

  checkEquals(deGenes, as.character(1:5))
}

###### IGSAinput tests

# It doesnt have any exported function to test

######## MGSZ tests

###### MGSZ tests

# It doesnt have any exported function to test

######## MIGSAmGSZ tests

###### MIGSAmGSZ tests

# this check is failing in bioconductor servers
# test_MIGSAmGSZ_ok_sameAsMgsz <- function() {
#     library(mGSZ);
#     perms <- 4;
#     nGenes <- 100;
#     nSamples <- 6;
#     geneNames <- paste("g", 1:nGenes, sep = "");
#
#     ## Create random gene expression data matrix.
#     set.seed(8818);
#     exprData <- matrix(rnorm(nGenes*nSamples),ncol=nSamples);
#     rownames(exprData) <- geneNames;
#
#     ## There will be nGenes/25 differentialy expressed genes.
#     nDeGenes <- nGenes/25;
#     ## Lets generate the offsets to sum to the differentialy expressed genes.
#     deOffsets <- matrix(2*abs(rnorm(nDeGenes*nSamples/2)), ncol=nSamples/2);
#
#     ## Randomly select which are the DE genes.
#     deIndexes <- sample(1:nGenes, nDeGenes, replace=FALSE);
#     exprData[deIndexes, 1:(nSamples/2)] <-
#         exprData[deIndexes, 1:(nSamples/2)] + deOffsets;
#
#     ## half of each condition.
#     conditions <- rep(c("C1", "C2"),c(nSamples/2,nSamples/2));
#
#     nGSets <- 4;
#     gSets <- lapply(1:nGSets, function(i) sample(geneNames, size=10));
#     names(gSets) <- paste("set", as.character(1:nGSets), sep="");
#
#     set.seed(8818);
#     mGSZres <- mGSZ(exprData, gSets, conditions, p=perms);
#     mGSZres <- mGSZres$mGSZ;
#
#     set.seed(8818);
#     MIGSAmGSZres <- MIGSAmGSZ(exprData, gSets, conditions, p=perms);
#
#     mergedRes <- merge(mGSZres, MIGSAmGSZres, by="gene.sets",
#     suffixes=c("mGSZ", "MIGSAmGSZ"))
#
#     # this check is failing in bioconductor servers
#     checkTrue(all.equal(round(mergedRes$gene.set.scores, 5),
#         round(abs(mergedRes$mGszScore), 5)));
#     checkTrue(all.equal(mergedRes$pvaluemGSZ, mergedRes$pvalueMIGSAmGSZ));
# }

test_MIGSAmGSZ_ok_validWithVoom <- function() {
  perms <- 10
  nGenes <- 500
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "") # with names g1 ... g1000

  ## Create random gene expression data matrix.
  set.seed(8818)
  exprData <- matrix(rnbinom(nGenes * nSamples, mu = 5, size = 2), ncol = nSamples)
  rownames(exprData) <- geneNames

  ## There will be 40 differentialy expressed genes.
  nDeGenes <- nGenes / 25
  ## Lets generate the offsets to sum to the differentialy expressed genes.
  deOffsets <- matrix(2 * abs(rnbinom(nDeGenes * nSamples / 2, mu = 5, size = 2)),
    ncol = nSamples / 2
  )

  ## Randomly select which are the DE genes.
  deIndexes <- sample(1:nGenes, nDeGenes, replace = FALSE)
  exprData[deIndexes, 1:(nSamples / 2)] <-
    exprData[deIndexes, 1:(nSamples / 2)] + deOffsets

  ## 15 subjects with condition C1 and 15 with C2.
  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10
  ## Lets create randomly 200 gene sets, of 10 genes each
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")

  set.seed(8818)
  MIGSAmGSZres <- MIGSAmGSZ(exprData, gSets, conditions,
    use.voom = !F,
    p = perms
  )

  checkEquals(ncol(MIGSAmGSZres), 4)
  checkEquals(nrow(MIGSAmGSZres), nGSets)
}

######## DEnricher tests

###### DEnricher tests

# It doesnt have any exported function to test

######## IGSAres tests

###### IGSAres tests

# It doesnt have any exported function to test

######## IGSA tests

###### IGSA tests

# It doesnt have any exported function to test

###### IGSAinput-common tests

test_IGSAinput_common_ok_summary <- function() {
  library(GSEABase)
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get some DE genes
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInputName <- "igsaInput"
  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = GeneSetCollection(myGSs))
  )

  igsaSumm <- summary(igsaInput)
  checkTrue(all(igsaSumm == c(
    "igsaInput", "6", "C1VSC2", "3", "3", "1",
    "200", "0", "0.3", "fdr", "1", "briii", "5", "0.5"
  )))
}

######## MIGSA tests

###### MIGSA tests

#### Correct ones

test_MIGSA_ok_oneExp <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get some DE genes
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInputName <- "MIGSA_ok_oneExp"
  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)

  checkEquals(ncol(migsaRes), 4)
  checkEquals(nrow(migsaRes), nGSets)

  checkTrue(length(unique(migsaRes$GS_Name)) == 1)
  checkTrue(unique(migsaRes$GS_Name) == "myGeneSets")

  checkTrue(all(names(gSets) %in% migsaRes$id))

  checkTrue(colnames(migsaRes)[[4]] == igsaInputName)
}

test_MIGSA_ok_twoExp <- function() {
  set.seed(8818)
  nGenes <- 100
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData1 <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData1) <- geneNames
  exprData1 <- new("MAList", list(M = exprData1))

  exprData2 <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData2) <- geneNames
  exprData2 <- new("MAList", list(M = exprData2))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 4
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get some DE genes
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInput1Name <- "MIGSA_ok_twoExp1"
  igsaInput1 <- IGSAinput(
    name = igsaInput1Name, expr_data = exprData1,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )

  igsaInput2Name <- "MIGSA_ok_twoExp2"
  igsaInput2 <- IGSAinput(
    name = igsaInput2Name, expr_data = exprData2,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )

  experiments <- list(igsaInput1, igsaInput2)

  migsaRes <- MIGSA(experiments)

  checkEquals(ncol(migsaRes), 5)
  checkEquals(nrow(migsaRes), nGSets)

  checkTrue(length(unique(migsaRes$GS_Name)) == 1)
  checkTrue(unique(migsaRes$GS_Name) == "myGeneSets")

  checkTrue(all(names(gSets) %in% migsaRes$id))

  checkTrue(colnames(migsaRes)[[4]] == igsaInput1Name)
  checkTrue(colnames(migsaRes)[[5]] == igsaInput2Name)
}

test_MIGSA_ok_twoGSs <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10

  gSets1 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets1) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs1 <- as.Genesets(gSets1)

  gSets2 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets2) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs2 <- as.Genesets(gSets2)

  fitOpts <- FitOptions(conditions)

  # to get some DE genes
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInputName <- "MIGSA_ok_twoGSs"
  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets1 = myGSs1, myGeneSets2 = myGSs2)
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)

  checkEquals(ncol(migsaRes), 4)
  checkEquals(nrow(migsaRes), 2 * nGSets)

  checkTrue(length(unique(migsaRes$GS_Name)) == 2)
  checkTrue(unique(migsaRes$GS_Name)[[1]] == "myGeneSets1")
  checkTrue(unique(migsaRes$GS_Name)[[2]] == "myGeneSets2")

  checkTrue(all(names(gSets1) %in% migsaRes$id))
  checkTrue(all(names(gSets2) %in% migsaRes$id))

  checkTrue(colnames(migsaRes)[[4]] == igsaInputName)
}

test_MIGSA_ok_twoExptwoGSs <- function() {
  set.seed(8818)
  nGenes <- 100
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData1 <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData1) <- geneNames
  exprData1 <- new("MAList", list(M = exprData1))

  exprData2 <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData2) <- geneNames
  exprData2 <- new("MAList", list(M = exprData2))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 4

  gSets1 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets1) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs1 <- as.Genesets(gSets1)

  gSets2 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets2) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs2 <- as.Genesets(gSets2)

  fitOpts <- FitOptions(conditions)

  # to get some DE genes
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInput1Name <- "MIGSA_ok_twoExptwoGSs1"
  igsaInput1 <- IGSAinput(
    name = igsaInput1Name, expr_data = exprData1,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets1 = myGSs1, myGeneSets2 = myGSs2)
  )

  igsaInput2Name <- "MIGSA_ok_twoExptwoGSs2"
  igsaInput2 <- IGSAinput(
    name = igsaInput2Name, expr_data = exprData2,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets1 = myGSs1, myGeneSets2 = myGSs2)
  )

  experiments <- list(igsaInput1, igsaInput2)

  migsaRes <- MIGSA(experiments)

  checkEquals(ncol(migsaRes), 5)
  checkEquals(nrow(migsaRes), 2 * nGSets)

  checkTrue(length(unique(migsaRes$GS_Name)) == 2)
  checkTrue(unique(migsaRes$GS_Name)[[1]] == "myGeneSets1")
  checkTrue(unique(migsaRes$GS_Name)[[2]] == "myGeneSets2")

  checkTrue(all(names(gSets1) %in% migsaRes$id))
  checkTrue(all(names(gSets2) %in% migsaRes$id))

  checkTrue(colnames(migsaRes)[[4]] == igsaInput1Name)
  checkTrue(colnames(migsaRes)[[5]] == igsaInput2Name)
}

test_MIGSA_ok_twoExpFourGSs <- function() {
  set.seed(8818)
  nGenes <- 100
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData1 <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData1) <- geneNames
  exprData1 <- new("MAList", list(M = exprData1))

  exprData2 <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData2) <- geneNames
  exprData2 <- new("MAList", list(M = exprData2))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 4

  gSets1 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets1) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs1 <- as.Genesets(gSets1)

  gSets2 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets2) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs2 <- as.Genesets(gSets2)

  gSets3 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets3) <- paste("set", as.character((nGSets + 1):(2 * nGSets)), sep = "")
  myGSs3 <- as.Genesets(gSets3)

  gSets4 <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets4) <- paste("set", as.character((nGSets + 1):(2 * nGSets)), sep = "")
  myGSs4 <- as.Genesets(gSets4)

  fitOpts <- FitOptions(conditions)

  # to get some DE genes
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInput1Name <- "MIGSA_ok_twoExpFourGSs1"
  igsaInput1 <- IGSAinput(
    name = igsaInput1Name, expr_data = exprData1,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets1 = myGSs1, myGeneSets3 = myGSs3)
  )

  igsaInput2Name <- "MIGSA_ok_twoExpFourGSs2"
  igsaInput2 <- IGSAinput(
    name = igsaInput2Name, expr_data = exprData2,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets2 = myGSs2, myGeneSets4 = myGSs4)
  )

  experiments <- list(igsaInput1, igsaInput2)

  migsaRes <- MIGSA(experiments)

  checkEquals(ncol(migsaRes), 5)
  checkEquals(nrow(migsaRes), 4 * nGSets)

  checkTrue(length(unique(migsaRes$GS_Name)) == 4)
  checkTrue(unique(migsaRes$GS_Name)[[1]] == "myGeneSets1")
  checkTrue(unique(migsaRes$GS_Name)[[2]] == "myGeneSets2")
  checkTrue(unique(migsaRes$GS_Name)[[3]] == "myGeneSets3")
  checkTrue(unique(migsaRes$GS_Name)[[4]] == "myGeneSets4")

  checkTrue(all(names(gSets1) %in% migsaRes$id))
  checkTrue(all(names(gSets2) %in% migsaRes$id))
  checkTrue(all(names(gSets3) %in% migsaRes$id))
  checkTrue(all(names(gSets4) %in% migsaRes$id))

  checkTrue(colnames(migsaRes)[[4]] == igsaInput1Name)
  checkTrue(colnames(migsaRes)[[5]] == igsaInput2Name)
}

test_MIGSA_ok_noIssueWNoDE <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get 0 DE genes
  seaParams <- SEAparams(de_cutoff = 0)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInputName <- "MIGSA_ok_noIssueWNoDE"
  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)
  checkTrue(validObject(migsaRes))
  checkEquals(ncol(migsaRes), 4)
  checkEquals(nrow(migsaRes), nGSets)

  checkTrue(all(migsaRes@migsa_res_all$SEA_pval == 1))
  checkTrue(all(is.na(migsaRes@migsa_res_all$SEA_score)))
  checkTrue(all(migsaRes@migsa_res_all$SEA_enriching_genes == ""))
}

test_MIGSA_ok_noIssueWtwoPerm <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get at least one DE gene
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 2)

  igsaInputName <- "MIGSA_ok_noIssueWtwoPerm"
  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)
  checkTrue(validObject(migsaRes))
  checkEquals(ncol(migsaRes), 4)
  checkEquals(nrow(migsaRes), nGSets)
}

test_MIGSA_ok_manuallyDEGenes <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # de_cutoff=1 so if we use fit then all genes are DE. but we just set as DE
  # the ones from gSet #1
  seaParams <- SEAparams(de_cutoff = 1, de_genes = gSets[[1]])
  gseaParams <- GSEAparams(perm_number = 2)

  igsaInputName <- "MIGSA_ok_manuallyDEGenes"
  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)
  checkEquals(
    sort(unique(unlist(strsplit(
      migsaRes@migsa_res_all$SEA_enriching_genes, ", "
    )))),
    sort(gSets[[1]])
  )
}

test_MIGSA_ok_usebri <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  # generate genes that are not in our experiment (so we get them in gSets).
  geneNames <- paste("g", 1:(10 * nGenes), sep = "")
  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get at least one DE gene
  seaParams1 <- SEAparams(de_cutoff = 0.5) # with briii
  seaParams2 <- SEAparams(de_cutoff = 0.5, br = "bri") # with bri
  gseaParams <- GSEAparams(perm_number = 2, min_sz = 0)

  igsaInputName1 <- "MIGSA_ok_usebri1"
  igsaInputName2 <- "MIGSA_ok_usebri2"

  igsaInput1 <- IGSAinput(
    name = igsaInputName1, expr_data = exprData,
    sea_params = seaParams1, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments1 <- list(igsaInput1)

  igsaInput2 <- IGSAinput(
    name = igsaInputName2, expr_data = exprData,
    sea_params = seaParams2, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments2 <- list(igsaInput2)

  set.seed(8818)
  migsaRes1 <- MIGSA(experiments1)

  set.seed(8818)
  migsaRes2 <- MIGSA(experiments2)

  # genes ranks are equal
  checkTrue(all(migsaRes1@genes_rank[[1]] == migsaRes2@genes_rank[[1]]))

  migsaRes1All <- as.data.frame(migsaRes1@migsa_res_all)
  migsaRes2All <- as.data.frame(migsaRes2@migsa_res_all)

  # results except for SEA are equal
  checkEquals(migsaRes1All[, -(c(1, 6:10))], migsaRes2All[, -(c(1, 6:10))])

  # results for SEA are different
  checkTrue(any(migsaRes1All[, 8] != migsaRes2All[, 8]))
}

test_MIGSA_ok_useOwnbr <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  # generate genes that are not in our experiment (so we get them in gSets).
  geneNames <- paste("g", 1:(10 * nGenes), sep = "")
  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get at least one DE gene
  seaParams1 <- SEAparams(de_cutoff = 0.5) # with briii
  seaParams2 <- SEAparams(
    de_cutoff = 0.5,
    br = geneNames[1:(length(geneNames) / 2)]
  )
  gseaParams <- GSEAparams(perm_number = 2, min_sz = 0)

  igsaInputName1 <- "MIGSA_ok_useOwnbr1"
  igsaInputName2 <- "MIGSA_ok_useOwnbr2"

  igsaInput1 <- IGSAinput(
    name = igsaInputName1, expr_data = exprData,
    sea_params = seaParams1, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments1 <- list(igsaInput1)

  igsaInput2 <- IGSAinput(
    name = igsaInputName2, expr_data = exprData,
    sea_params = seaParams2, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments2 <- list(igsaInput2)

  set.seed(8818)
  migsaRes1 <- MIGSA(experiments1)

  set.seed(8818)
  migsaRes2 <- MIGSA(experiments2)

  # genes ranks are equal
  checkTrue(all(migsaRes1@genes_rank[[1]] == migsaRes2@genes_rank[[1]]))

  migsaRes1All <- as.data.frame(migsaRes1@migsa_res_all)
  migsaRes2All <- as.data.frame(migsaRes2@migsa_res_all)

  # results except for SEA are equal
  checkEquals(migsaRes1All[, -(c(1, 6:10))], migsaRes2All[, -(c(1, 6:10))])

  # results for SEA are different
  checkTrue(any(migsaRes1All[, 8] != migsaRes2All[, 8]))
}

test_MIGSA_ok_useDifferenttests <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  # generate genes that are not in our experiment (so we get them in gSets).
  geneNames <- paste("g", 1:(10 * nGenes), sep = "")
  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get at least one DE gene
  seaParams1 <- SEAparams(de_cutoff = 0.5)
  seaParams2 <- SEAparams(de_cutoff = 0.5, test = "HypergeoTest")
  seaParams3 <- SEAparams(de_cutoff = 0.5, test = "BinomialTest")
  gseaParams <- GSEAparams(perm_number = 2, min_sz = 0)

  igsaInputName1 <- "MIGSA_ok_useDifferenttests1"
  igsaInputName2 <- "MIGSA_ok_useDifferenttests2"
  igsaInputName3 <- "MIGSA_ok_useDifferenttests3"

  igsaInput1 <- IGSAinput(
    name = igsaInputName1, expr_data = exprData,
    sea_params = seaParams1, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments1 <- list(igsaInput1)

  igsaInput2 <- IGSAinput(
    name = igsaInputName2, expr_data = exprData,
    sea_params = seaParams2, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments2 <- list(igsaInput2)

  igsaInput3 <- IGSAinput(
    name = igsaInputName3, expr_data = exprData,
    sea_params = seaParams3, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  # to test also putting gene sets in MIGSAinput
  experiments3 <- list(igsaInput3)

  set.seed(8818)
  migsaRes1 <- MIGSA(experiments1)

  set.seed(8818)
  migsaRes2 <- MIGSA(experiments2)


  set.seed(8818)
  migsaRes3 <- MIGSA(experiments3, geneSets = list(myGeneSets = myGSs))

  # genes ranks are equal
  checkTrue(all(migsaRes1@genes_rank[[1]] == migsaRes2@genes_rank[[1]]))
  checkTrue(all(migsaRes1@genes_rank[[1]] == migsaRes3@genes_rank[[1]]))

  migsaRes1All <- as.data.frame(migsaRes1@migsa_res_all)
  migsaRes2All <- as.data.frame(migsaRes2@migsa_res_all)
  migsaRes3All <- as.data.frame(migsaRes3@migsa_res_all)

  # results except for SEA are equal
  checkEquals(migsaRes1All[, -(c(1, 6:10))], migsaRes2All[, -(c(1, 6:10))])
  checkEquals(migsaRes1All[, -(c(1, 6:10))], migsaRes3All[, -(c(1, 6:10))])

  # results for SEA are different
  checkTrue(any(migsaRes1All[, 8] != migsaRes2All[, 8]))
  checkTrue(any(migsaRes1All[, 8] != migsaRes3All[, 8]))
  checkTrue(any(migsaRes2All[, 8] != migsaRes3All[, 8]))
}

test_MIGSA_ok_onlySeaOrGsea <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get some DE genes
  seaParams <- SEAparams(de_cutoff = 0.3)
  gseaParams <- GSEAparams(perm_number = 5)

  igsaInput1 <- IGSAinput(
    name = "MIGSA_ok_onlySeaOrGsea1", expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  set.seed(8818)
  migsaRes1 <- MIGSA(list(igsaInput1))

  igsaInput2 <- IGSAinput(
    name = "MIGSA_ok_onlySeaOrGsea2", expr_data = exprData,
    sea_params = seaParams, gsea_params = NULL, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  set.seed(8818)
  migsaRes2 <- MIGSA(list(igsaInput2))

  # todo: uncomment this and pass tests on Windows i386
  #     igsaInput3 <- IGSAinput(name="MIGSA_ok_onlySeaOrGsea3", expr_data=exprData,
  #         sea_params=NULL, gsea_params=gseaParams, fit_options=fitOpts,
  #         gene_sets_list=list(myGeneSets=myGSs));
  #     set.seed(8818);
  #     migsaRes3 <- MIGSA(list(igsaInput3));

  # check results from both SEA are equal, and the rest NA
  checkEquals(migsaRes1@migsa_res_all[, 2:10], migsaRes2@migsa_res_all[, 2:10])
  checkTrue(all(is.na(migsaRes2@migsa_res_all[, 11:15])))

  # check results from both GSEA are equal, and the rest NA
  #     checkEquals(migsaRes1@migsa_res_all[,c(2:5, 11:15)],
  #         migsaRes3@migsa_res_all[,c(2:5, 11:15)]);
  #     checkTrue(all(is.na(migsaRes3@migsa_res_all[,6:10])));
}

test_MIGSA_ok_useOwnbrNoExprData <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  # generate genes that are not in our experiment (so we get them in gSets).
  geneNames <- paste("g", 1:(10 * nGenes), sep = "")
  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get at least one DE gene
  seaParams1 <- SEAparams(de_cutoff = 0.5) # with briii

  igsaInputName1 <- "MIGSA_ok_useOwnbrNoExprData1"
  igsaInputName2 <- "MIGSA_ok_useOwnbrNoExprData2"

  igsaInput1 <- IGSAinput(
    name = igsaInputName1, expr_data = exprData,
    sea_params = seaParams1, gsea_params = NULL, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments1 <- list(igsaInput1)

  myDeGenes <- seaParams(getDEGenes(igsaInput1))
  igsaInput2 <- IGSAinput(
    name = igsaInputName2,
    sea_params = SEAparams(
      de_genes = myDeGenes@de_genes,
      br = rownames(exprData)
    ),
    gsea_params = NULL,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments2 <- list(igsaInput2)

  set.seed(8818)
  migsaRes1 <- MIGSA(experiments1)

  set.seed(8818)
  migsaRes2 <- MIGSA(experiments2)

  checkEquals(migsaRes1@migsa_res_all[, -1], migsaRes2@migsa_res_all[, -1])
  checkTrue(all(migsaRes1@migsa_res_summary == migsaRes2@migsa_res_summary))
}

#### Incorrect ones

test_MIGSA_wrong_noExperiments <- function() {
  checkException(MIGSA(list()))
}

test_MIGSA_wrong_wrongExperiments <- function() {
  checkException(MIGSA(list(1:10)))
}

test_MIGSA_wrong_repExpNames <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))
  fitOpts <- FitOptions(conditions)

  igsaInput <- IGSAinput(
    name = "exp", expr_data = exprData,
    fit_options = fitOpts
  )

  checkException(MIGSA(list(
    igsaInput,
    igsaInput
  )))
}

test_MIGSA_wrong_noGSets <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  fitOpts <- FitOptions(conditions)

  igsaInputName <- "MIGSA_wrong_noGSets"
  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    fit_options = fitOpts, gene_sets_list = list()
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)

  checkTrue(is.na(migsaRes))
}

# test_MIGSA_ok_wrongbr <- function() {
#     set.seed(8818);
#     nGenes <- 200;
#     nSamples <- 6;
#     geneNames <- paste("g", 1:nGenes, sep = "");
#
#     exprData <- matrix(rnorm(nGenes*nSamples),ncol=nSamples);
#     rownames(exprData) <- geneNames;
#     exprData <- new("MAList",list(M=exprData));
#
#     conditions <- rep(c("C1", "C2"),c(nSamples/2,nSamples/2));
#
#     # generate genes that are not in our experiment (so we get them in gSets).
#     geneNames <- paste("g", 1:(10*nGenes), sep = "");
#     nGSets <- 10;
#     gSets <- lapply(1:nGSets, function(i) sample(geneNames, size=10));
#     names(gSets) <- paste("set", as.character(1:nGSets), sep="");
#     myGSs <- as.Genesets(gSets);
#
#     fitOpts <- FitOptions(conditions);
#
#     # to get at least one DE gene
#     seaParams <- SEAparams(de_cutoff=0.5,
#         br=paste("gene", 1:100));
#     gseaParams <- GSEAparams(perm_number=2, min_sz=0);
#
#     igsaInputName <- "igsaInput";
#
#     igsaInput <- IGSAinput(name=igsaInputName, expr_data=exprData,
#         sea_params=seaParams, gsea_params=gseaParams, fit_options=fitOpts,
#         gene_sets_list=list(myGeneSets=myGSs));
#     experiments <- list(igsaInput);
#
#     migsaRes <- MIGSA(experiments);
#     checkTrue(is.na(migsaRes));
# }

test_MIGSA_ok_wrongbrOption <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  # generate genes that are not in our experiment (so we get them in gSets).
  geneNames <- paste("g", 1:(10 * nGenes), sep = "")
  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  # to get at least one DE gene
  seaParams <- SEAparams(de_cutoff = 0.5, br = "wrongbr")
  gseaParams <- GSEAparams(perm_number = 2, min_sz = 0)

  igsaInputName <- "MIGSA_ok_wrongbrOption"

  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = seaParams, gsea_params = gseaParams, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)
  checkTrue(is.na(migsaRes))
}

test_MIGSA_ok_paramsCantbebothNull <- function() {
  set.seed(8818)
  nGenes <- 200
  nSamples <- 6
  geneNames <- paste("g", 1:nGenes, sep = "")

  exprData <- matrix(rnorm(nGenes * nSamples), ncol = nSamples)
  rownames(exprData) <- geneNames
  exprData <- new("MAList", list(M = exprData))

  conditions <- rep(c("C1", "C2"), c(nSamples / 2, nSamples / 2))

  # generate genes that are not in our experiment (so we get them in gSets).
  geneNames <- paste("g", 1:(10 * nGenes), sep = "")
  nGSets <- 10
  gSets <- lapply(1:nGSets, function(i) sample(geneNames, size = 10))
  names(gSets) <- paste("set", as.character(1:nGSets), sep = "")
  myGSs <- as.Genesets(gSets)

  fitOpts <- FitOptions(conditions)

  igsaInputName <- "MIGSA_ok_paramsCantbebothNull"

  igsaInput <- IGSAinput(
    name = igsaInputName, expr_data = exprData,
    sea_params = NULL, gsea_params = NULL, fit_options = fitOpts,
    gene_sets_list = list(myGeneSets = myGSs)
  )
  experiments <- list(igsaInput)

  migsaRes <- MIGSA(experiments)
  checkTrue(is.na(migsaRes))
}

######## MIGSAres tests

###### MIGSAres-class tests

# It doesnt have any exported function to test

###### MIGSAres-common tests

test_MIGSAres_common_ok_summaryMinimPval <- function() {
  data(migsaRes)

  migsaResShow <- show(migsaRes)
  migsaResPvals <- as.data.frame(migsaRes@migsa_res_all)

  checkEquals(rep(migsaResShow$id, 2), migsaResPvals$id)
  checkEquals(rep(migsaResShow$Name, 2), migsaResPvals$name)
  checkEquals(rep(migsaResShow$GS_Name, 2), migsaResPvals$gene_set_name)

  checkEquals(migsaResShow[, 4], as.numeric(
    apply(migsaResPvals[1:nrow(migsaResShow), c(8, 13)], 1, min)
  ))
  checkEquals(migsaResShow[, 5], as.numeric(
    apply(migsaResPvals[
      (nrow(migsaResShow) + 1):nrow(migsaResPvals),
      c(8, 13)
    ], 1, min)
  ))
}

test_MIGSAres_common_ok_tailHead <- function() {
  data(migsaRes)

  checkTrue(class(head(migsaRes)) == "MIGSAres")
  checkTrue(class(tail(migsaRes)) == "MIGSAres")

  checkEquals(nrow(head(migsaRes, 10)), 10)
  checkEquals(nrow(tail(migsaRes, 10)), 10)

  checkEquals(head(migsaRes, 10), migsaRes[1:10, ])
  checkEquals(
    tail(migsaRes, 10),
    migsaRes[(nrow(migsaRes) - 9):nrow(migsaRes), ]
  )
}

test_MIGSAres_common_ok_summary <- function() {
  data(migsaRes)

  migsaResSummary <- summary(migsaRes)
  checkEquals(c(migsaResSummary), c(6, 15, 31, 8, 23, 37))
  migsaRes <- setEnrCutoff(migsaRes, 0.01)
  migsaResSummary <- summary(migsaRes)

  checkEquals(mean(migsaResSummary$consensusGeneSets), 100)
  checkEquals(
    unlist(c(migsaResSummary$enrichmentIntersections)),
    c(6, 0, 0, 8)
  )
}

test_MIGSAres_common_ok_asDframe <- function() {
  data(migsaRes)
  migsaResDframe <- as.data.frame(migsaRes)
  checkEquals(dim(migsaResDframe), c(200, 5))
}

test_MIGSAres_common_ok_merge <- function() {
  data(migsaRes)
  data(bcMigsaRes)
  migsaResMerge <- merge(migsaRes, bcMigsaRes)

  checkEquals(nrow(migsaResMerge), nrow(migsaRes) + nrow(bcMigsaRes))
  checkEquals(ncol(migsaResMerge), ncol(migsaRes) + ncol(bcMigsaRes) - 3)

  migsaResMergeDframe <- as.data.frame(migsaResMerge)
  migsaResDframe <- as.data.frame(migsaRes)
  bcMigsaResDframe <- as.data.frame(bcMigsaRes)

  checkTrue(all(migsaResDframe$id %in% migsaResMergeDframe$id))
  checkTrue(all(bcMigsaResDframe$id %in% migsaResMergeDframe$id))

  checkTrue(all(migsaResDframe$Name %in% migsaResMergeDframe$Name))
  checkTrue(all(bcMigsaResDframe$Name %in% migsaResMergeDframe$Name))

  checkTrue(all(migsaResDframe$GS_Name %in% migsaResMergeDframe$GS_Name))
  checkTrue(all(bcMigsaResDframe$GS_Name %in% migsaResMergeDframe$GS_Name))

  checkEquals(
    sort(migsaResMergeDframe[, 4]),
    sort(migsaResDframe[, 4])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 5]),
    sort(migsaResDframe[, 5])
  )

  checkEquals(
    sort(migsaResMergeDframe[, 6]),
    sort(bcMigsaResDframe[, 4])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 7]),
    sort(bcMigsaResDframe[, 5])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 8]),
    sort(bcMigsaResDframe[, 6])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 9]),
    sort(bcMigsaResDframe[, 7])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 10]),
    sort(bcMigsaResDframe[, 8])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 11]),
    sort(bcMigsaResDframe[, 9])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 12]),
    sort(bcMigsaResDframe[, 10])
  )
  checkEquals(
    sort(migsaResMergeDframe[, 13]),
    sort(bcMigsaResDframe[, 11])
  )
}

test_MIGSAres_common_wrong_merge <- function() {
  data(migsaRes)

  # experiment names can not be the same
  checkException(merge(migsaRes, migsaRes))
}

###### MIGSAres-setEnrCutoff tests

test_MIGSAres_setEnrCutoff_ok <- function() {
  data(migsaRes)

  pvals <- migsaRes[, 4:5]

  checkTrue(all(unlist(lapply(seq(0, 1, by = 0.05), function(actCoff) {
    actMigsaRes <- setEnrCutoff(migsaRes, actCoff)
    actEnr <- actMigsaRes[, 4:5]
    all(actEnr == (pvals < actCoff))
  }))))
}

test_MIGSAres_setEnrCutoff_ok_cOffNoEnr <- function() {
  data(migsaRes)

  noEnrMigsaRes <- setEnrCutoff(migsaRes, 0)
  checkTrue(all(!noEnrMigsaRes[, 4:5]))
}

test_MIGSAres_setEnrCutoff_ok_cOffAllEnr <- function() {
  data(migsaRes)

  allEnrMigsaRes <- setEnrCutoff(migsaRes, 1)
  checkTrue(all(!!allEnrMigsaRes[, 4:5]))
}

test_MIGSAres_setEnrCutoff_ok_NACoff <- function() {
  data(migsaRes)

  checkEquals(class(migsaRes[, 4, drop = !FALSE]), "numeric")
  checkEquals(class(migsaRes[, 5, drop = !FALSE]), "numeric")

  migsaResCoff <- setEnrCutoff(migsaRes, 0.01)

  checkEquals(class(migsaResCoff[, 4, drop = !FALSE]), "logical")
  checkEquals(class(migsaResCoff[, 5, drop = !FALSE]), "logical")
}

###### MIGSAres-genesInSets tests

test_MIGSAres_genesInSets_ok <- function() {
  data(migsaRes)
  migsaRes <- migsaRes[1:10, ]

  additionalInfo <- getAdditionalInfo(migsaRes)
  genesInfo <- additionalInfo[
    ,
    c(
      "experiment_name", "gene_set_name", "id", "SEA_pval", "GSEA_pval",
      "SEA_enriching_genes", "GSEA_enriching_genes"
    )
  ]

  checkTrue(all(unlist(lapply(seq(0, 1, by = 0.25), function(actCoff) {
    actMigsaRes <- setEnrCutoff(migsaRes, actCoff)
    actGInSets <- genesInSets(actMigsaRes)

    totalRes <- apply(
      unique(genesInfo[, c("gene_set_name", "id")]), 1,
      function(actGSet) {
        actGSetInfo <- genesInfo[
          genesInfo$gene_set_name == actGSet[[1]] &
            genesInfo$id == actGSet[[2]],
        ]
        exprGenes <- table(unlist(apply(
          actGSetInfo, 1,
          function(actActGSetInfo) {
            enrGenes <- NA
            if (actActGSetInfo[["SEA_pval"]] < actCoff) {
              enrGenes <- union(enrGenes, unlist(
                strsplit(actActGSetInfo[["SEA_enriching_genes"]], ", ")
              ))
            }
            if (actActGSetInfo[["GSEA_pval"]] < actCoff) {
              enrGenes <- union(enrGenes, unlist(
                strsplit(actActGSetInfo[["GSEA_enriching_genes"]], ", ")
              ))
            }

            enrGenes <- enrGenes[!is.na(enrGenes)]
            return(enrGenes)
          }
        )))

        actGISEprGenes <- actGInSets[
          paste(actGSet[[1]], actGSet[[2]], sep = "_"),
        ]
        res <- all(actGISEprGenes[names(exprGenes)] == exprGenes)
        res <- res && sum(actGISEprGenes) == sum(exprGenes)
        return(res)
      }
    )
    return(totalRes)
  }))))
}

test_MIGSAres_genesInSets_ok_noEnriched <- function() {
  data(migsaRes)

  # So I get 0 enriched gene sets
  noEnrMigsaRes <- setEnrCutoff(migsaRes, 0)
  gInSets <- genesInSets(noEnrMigsaRes)

  checkEquals(nrow(gInSets), 200)
  checkEquals(ncol(gInSets), 0)
}

###### MIGSAres-filterByGenes tests

#### Correct ones

test_MIGSAres_filterByGenes_ok_someGenes <- function() {
  data(migsaRes)
  migsaRes <- migsaRes[1:40, ]

  # to enrich every gene set
  migsaRes <- setEnrCutoff(migsaRes, 1)
  gInSets <- genesInSets(migsaRes)
  impGenes <- colnames(gInSets)[1:10]

  filtMigsaRes <- filterByGenes(migsaRes, impGenes)
  checkEquals(ncol(filtMigsaRes), 5)
}

test_MIGSAres_filterByGenes_ok_oneGene <- function() {
  data(migsaRes)
  migsaRes <- migsaRes[1:40, ]

  # to enrich every gene set
  migsaRes <- setEnrCutoff(migsaRes, 1)
  gInSets <- genesInSets(migsaRes)
  impGenes <- colnames(gInSets)[[1]]

  filtMigsaRes <- filterByGenes(migsaRes, impGenes)
  checkEquals(ncol(filtMigsaRes), 5)
}

test_MIGSAres_filterByGenes_ok_emptyGenes <- function() {
  data(migsaRes)
  migsaRes <- migsaRes[1:40, ]

  # to enrich every gene set
  migsaRes <- setEnrCutoff(migsaRes, 1)

  filtMigsaRes <- filterByGenes(migsaRes, "")
  checkEquals(ncol(filtMigsaRes), 0)
  checkEquals(nrow(filtMigsaRes), 0)
}

test_MIGSAres_filterByGenes_ok_fakeGenes <- function() {
  data(migsaRes)

  # to enrich every gene set
  migsaRes <- setEnrCutoff(migsaRes, 1)

  filtMigsaRes <- filterByGenes(migsaRes, c("fakeGene1", "fakeGene2"))
  checkEquals(ncol(filtMigsaRes), 0)
  checkEquals(nrow(filtMigsaRes), 0)
}

test_MIGSAres_filterByGenes_ok_noEnriched <- function() {
  data(migsaRes)

  # to enrich every gene set
  migsaRes <- setEnrCutoff(migsaRes, 1)
  gInSets <- genesInSets(migsaRes)
  impGenes <- colnames(gInSets)

  # to enrich 0 gene sets
  migsaRes <- setEnrCutoff(migsaRes, 0)

  filtMigsaRes <- filterByGenes(migsaRes, impGenes)
  checkEquals(ncol(filtMigsaRes), 0)
  checkEquals(nrow(filtMigsaRes), 0)
}

###### MIGSAres-genesBarplot tests

test_MIGSAres_genesBarplot_ok_simplePlot <- function() {
  data(migsaRes)

  plotRes <- genesBarplot(migsaRes)
  checkTrue(is(plotRes, "gg"))
  #     checkEqualsNumeric(summary(plotRes$data$number), c(1,1,1,1.213,1,4));
  checkEqualsNumeric(length(unique(plotRes$data$id)), 47)
}

###### MIGSAres-genesHeatmap tests

test_MIGSAres_genesHeatmap_ok_simplePlot <- function() {
  data(migsaRes)

  plotRes <- genesHeatmap(migsaRes)
  checkTrue(is(plotRes, "list"))
  checkEqualsNumeric(
    plotRes$rowInd,
    c(13, 8, 5, 6, 2, 12, 11, 4, 14, 1, 10, 9, 3, 7)
  )
  checkEqualsNumeric(summary(plotRes$colInd), c(1, 12.5, 24, 24, 35.5, 47))
  #     checkEqualsNumeric(summary(c(plotRes$data)), c(0,0,0,0.08662614,0,1));
}

###### MIGSAres-geneSetBarplot tests

test_MIGSAres_geneSetBarplot_ok_simplePlot <- function() {
  data(migsaRes)

  plotRes <- geneSetBarplot(migsaRes)
  checkTrue(is(plotRes, "gg"))
  checkEqualsNumeric(summary(plotRes$data$number), c(1, 1, 1, 1, 1, 1))
  checkEqualsNumeric(length(unique(plotRes$data$id)), 14)
  checkEqualsNumeric(length(unique(plotRes$data$GS_Name)), 1)
}

###### MIGSAres-getAdditionalInfo tests

# quite a irrelevant test, as the function does the same.
test_MIGSAres_getAdditionalInfo_ok <- function() {
  data(migsaRes)

  additionalInfo <- getAdditionalInfo(migsaRes)
  allInfo <- migsaRes@migsa_res_all

  checkEquals(ncol(allInfo), ncol(additionalInfo))

  # to avoid NAs
  allInfo[is.na(allInfo)] <- 0
  additionalInfo[is.na(additionalInfo)] <- 0

  checkTrue(all(unlist(lapply(1:ncol(allInfo), function(j) {
    all(allInfo[[j]] == additionalInfo[, j])
  }))))
}

###### MIGSAres-migsaHeatmap tests

test_MIGSAres_migsaHeatmap_ok_simplePlot <- function() {
  data(migsaRes)
  migsaRes <- migsaRes[1:50, ]
  migsaRes <- setEnrCutoff(migsaRes, 0.01)
  plotRes <- migsaHeatmap(migsaRes)

  checkTrue(is(plotRes, "list"))
  checkEqualsNumeric(plotRes$rowInd, c(1, 3, 2, 4))
  checkEqualsNumeric(plotRes$colInd, 1:2)
  #     checkEqualsNumeric(summary(c(plotRes$data)), c(0,0,0.5,0.5,1,1));
}

###### MIGSAres tests

# It doesnt have any exported function to test

######## GoAnalysis tests

###### GoAnalysis-getHeights tests

if (testAll) {
  test_GoAnalysis_getHeights <- function() {
    heights <- getHeights(
      c("GO:0008150", "GO:0007610", "GO:0050789", "fakeId")
    )
    checkEquals(heights[1:3], c(0, 1, 1))
    checkTrue(is.na(heights[[4]]))
  }
}

# test_GoAnalysis_getHeights_maxHeights <- function() {
#   heights <- getHeights(
#     c("GO:0008150", "GO:0007610", "GO:0050789", "fakeId"),
#     minHeight = FALSE
#   )
#   checkEquals(heights[1:3], c(0, 1, 2))
#   checkTrue(is.na(heights[[4]]))
# }

###### GoAnalysis-migsaGoTree tests

#### Correct ones

test_GoAnalysis_migsaGoTree_ok_simplePlot <- function() {
  data(bcMigsaRes)
  bcMigsaRes <- bcMigsaRes[5:6, ]

  plotRes <- migsaGoTree(bcMigsaRes, ont = "MF")
  checkTrue(all(sort(unlist(lapply(plotRes$gotree, nrow))) ==
    sort(table(bcMigsaRes$GS_Name)[-3])))
}

#### Incorrect ones

test_GoAnalysis_migsaGoTree_wrong_noGO <- function() {
  data(bcMigsaRes)
  keggMigsaRes <- bcMigsaRes[bcMigsaRes$GS_Name == "KEGG_2015", ]

  checkException(migsaGoTree(keggMigsaRes))
}