Package 'MOSClip'

Description

Given the hierarchy of the pathways, this formula finds the fathers of the respective pathway (e.g. pathway: 'PI3K Cascade'; father: 'Signaling Pathways'). This function is necessary for calculating the contribution of different omics to survival prediction in different biological processes, grouping the pathways by hierarchy.

Usage

annotePathwayToFather(pathways, graphiteDB, hierarchy)

Arguments

Value

a vector of the pathway fathers' names

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOM_list <- lapply(reactSmall[1:2], function(g) {
    fcl <- multiOmicsSurvivalModuleTest(multiOmics, g,
        survFormula = "Surv(days, status) ~",
        autoCompleteFormula = TRUE,
        useTheseGenes = genesToUse
    )
    fcl
})

moduleSummary <- multiPathwayModuleReport(MOM_list)

pathHierarchy <- downloadPathwayRelationFromReactome()
pathHierarchyGraph <- igraph::graph.data.frame(
    d = pathHierarchy,
    directed = TRUE
)

omicsClasses2pathways <- computeOmicsIntersections(
    moduleSummary,
    pvalueThr = 1, zscoreThr = 1,
    excludeColumns = c("pathway", "module")
)

omicsClasses2pathways <- lapply(
    omicsClasses2pathways,
    stripModulesFromPathways
)

# This step requires to download the whole reactome graph, which usually
# takes a lot of time.
# reactome <- graphite::pathways('hsapiens', 'reactome')
# reactome <- graphite::convertIdentifiers(reactome, 'entrez')
# omicsClasses2fathers <- lapply(omicsClasses2pathways, 
#                                annotePathwayToFather, 
#                                graphiteDB = reactome,
#                                hierarchy = pathHierarchyGraph)

Description

Gives a vector of the available methods to summarize omics.

Usage

availableOmicMethods()

Value

character vector with the implemented methods.

Examples

availableOmicMethods()

Description

For internal use only

For internal use only

For internal use only

For internal use only

For internal use only

Prepare subset of patients for permutations

Usage

checkOrder(li)

resolveAndOrder(li)

mergeCol(li, col = "PC1", resolve = FALSE)

filterExpr(exp, samples)

filterMultiOmicsForSamples(MO, samples)

preparePerms(fullMultiOmics, nperm = 100, nPatients = 3)

Arguments

Value

a matrix

a filtered matrix

a filtered MultiOmics objects

list of sampled patients for each permutation

Description

Regular PCA

Usage

compPCs(exp, shrink, k)

Arguments

Value

a list with the following elements:

Description

Compute frequencies in a named list. This function is necessary for plotFrequencies, in which it will calculate the frequency of each pathway father for every omics intersection.

Usage

computeFreqs(elementsIntersections)

Arguments

Value

a data.frame of the frequencies

Examples

omicsIntersection <- list(
    "exp;met" = c("PathwayA", "PathwayB", "PathwayC"),
    "exp;mut" = c("PathwayA", "PathwayC"),
    "cnv;mut" = c("PathwayB")
)
freqDf <- computeFreqs(omicsIntersection)

Description

Finds the modules that have any intersection among the available omics

Usage

computeOmicsIntersections(
  multiPathwayReportData,
  pvalueThr = 0.05,
  zscoreThr = 0.05,
  resampligThr = NULL,
  excludeColumns = NULL
)

Arguments

Value

a list of pathway modules present for every intersection of omics present

Examples

df <- data.frame(
    pvalue = c(0.06, 0.04, 0.04, 0.03, 0.02),
    cnv = c(0.07, 0.03, 0.02, 0.04, 0.01),
    mut = c(0.08, 0.02, 0.01, 0.04, 0.04),
    row.names = c(
        "PathwayA", "PathwayB", "PathwayC",
        "PathwayD", "PathwayE"
    )
)

omicsClasses2Pathways <- computeOmicsIntersections(df,
    pvalueThr = 0.1,
    zscoreThr = 0.1
)

Description

For internal use only. Performs Principal Componenent analysis.

Usage

computePCs(
  exp,
  shrink = FALSE,
  method = c("regular", "topological", "sparse"),
  cliques = NULL,
  maxPCs = 3
)

Arguments

Details

Three methods are implemented:

• regular: a regular PCA ('prcomp')

• topological: PCA using a pathway topology.

• sparse: sparse PCA analysis implemented by 'elasticnet'

Value

a list with the following elements:

Description

A generic function to convert pathway

Usage

convertPathway(graph, useTheseGenes)

Arguments

Value

NULL. No value is returned

Description

Create the coxObj from the covariates used in the test

Usage

createCoxObj(colData, moView)

Arguments

Value

data.frame, samples in the rows, covariates in the columns

Description

Extract sub-matrix for the genes of a module or pathway from data matrix of a specific omic

Usage

createDataModule(omic, multiOmicObj)

Arguments

Value

matrix, genes in the rows, samples in the columns

Description

Create the list of covariates that are going to be tested

Usage

createMOMView(omicsObj, genes)

Arguments

Value

list with 1 reduced representation of the omics 2 sdev 3 loadings or eigenvector 4 usedGenes 5 method 6 namesCov 7 omicName

Description

Download Pathway Relations from Reactome. The file is retrieved from the url

Usage

downloadPathwayRelationFromReactome(url = NULL, speciesAbbr = "HSA")

Arguments

Value

A data frame with 2 columns:

Examples

downloadPathwayRelationFromReactome()

Description

For internal use only. Estimate Covariance from one matrix

Usage

estimateExprCov(expr, shrink)

Arguments

Value

a covariance matrix

Description

For internal use only. Extract the cliques.

For internal use only. Force Moralization

Usage

extractCliquesFromDag(dag, root = NULL)

mmmoralize(graph)

Arguments

Value

list of nodes cliques

Description

Given an omic summarized by 'summarizeToBinaryEvents' extract the most important features.

Usage

extractSummaryFromBinary(omic, multiOmicObj, n = 3)

Arguments

Value

Meant for internal use only. The summary for omic summarized using binary events.

Description

Given an omic summarized by 'summarizeInCluster' extract the most important features.

Usage

extractSummaryFromCluster(omic, multiOmicObj, n = 3)

Arguments

Value

summary for omic summarized using clusters

Description

Given an omic summarized by 'summarizeToNumberOfEvents' extract the most important features.

Usage

extractSummaryFromNumberOfEvents(
  omic,
  multiOmicObj,
  moduleCox,
  analysis,
  n = 3,
  minprop = 0.1,
  labels = c("few", "many")
)

Arguments

Value

Meant for internal use only. The summary for omic summarized using counting of events.

Description

Given an omic summarized by 'summarizeWithPca' extract the most important features.

Usage

extractSummaryFromPCA(
  omic,
  multiOmicObj,
  moduleCox,
  analysis,
  loadThr = 0.6,
  atleast = 1,
  minprop = 0.1
)

Arguments

Value

summary for omic summarized using pca

Description

For internal use only. Retrieves relatives given a pathway id.

Usage

getPathFathers(pathway, hierarchyGraph, ord = 3, plot = FALSE)

Arguments

Details

Pathway Hierarchy is needed as igraph object.

Value

a character vector with the relatives

Description

Two-classes glm test.

Usage

glmTest(data, fullModelFormula, nullModelFormula)

Arguments

Value

Two-classes glm test results

Description

Given a pathway analyzed by multiOmicsModuleSurvivalTest or multiOmicsTwoClassModuleTest, it retrieves for each omic the most influent features.

Usage

guessInvolvement(
  pathway,
  moduleNumber,
  loadThr = 0.6,
  n = 3,
  atleast = 1,
  min_prop_pca = 0.1,
  min_prop_events = 0.1,
  ...
)

Arguments

Value

a list. Each item of the list corresponds to an omic that is summarized with the specific 'extractSummary' functions. Each item is the summary for an omic summarized using the setted method: pvalues are present only for cluster method.

Description

Given a pathway analyzed by multiOmicsSurvivalPathwayTest or multiOmicsTwoClassPathwayTest, it retrieves for each omic the most influent features.

Usage

guessInvolvementPathway(
  pathway,
  loadThr = 0.6,
  n = 3,
  atleast = 1,
  min_prop_pca = 0.1,
  min_prop_events = 0.1,
  ...
)

Arguments

Value

a list. Each item of the list corresponds to an omic that is summarized with the specific 'extractSummary' functions. Each item is the summary for an omic summarized using the setted method: pvalues are present only for cluster method.

Description

For internal use only. Retrieves name from pathway id.

Usage

id2name(idList, namedVect)

Arguments

Details

You must provide a namedVect to be used as translator.

Value

a character vector with the names

Description

makeOmics creates the Omics object necessary to perform most of the analyses of this package. It contains all the omics data in the format of a ExperimentList, the clinical data, and all the information necessary for the dimensionality reduction step.

Usage

makeOmics(
  experiments = ExperimentList(),
  colData = S4Vectors::DataFrame(),
  sampleMap = S4Vectors::DataFrame(assay = factor(), primary = character(), colname =
    character()),
  metadata = list(),
  drops = list(),
  modelInfo = character(),
  specificArgs = list()
)

Arguments

Value

an Omics class object

Examples

data(ovarianDataset)

myColData <- data.frame(
    status = sample(c(0, 1), 50, replace = TRUE),
    days = sample(c(0, 500), 50, replace = TRUE),
    row.names = colnames(ovarianDataset$exp)
)

myOmicsObj <- makeOmics(
    experiments = ovarianDataset,
    colData = myColData,
    modelInfo = c(
        "summarizeWithPca",
        "summarizeInCluster",
        "summarizeToNumberOfEvents",
        "summarizeToNumberOfDirectionalEvents"
    ),
    specificArgs = list(
        pcaArgs = list(
            name = "exp", shrink = "FALSE",
            method = "sparse", maxPCs = 3
        ),
        clusterArgs = list(
            name = "met",
            max_cluster_number = 3
        ),
        countEvent = list(name = "mut", min_prop = 0.05),
        cnvAgv = list(name = "cnv", min_prop = 0.05)
    )
)

Description

Make positive and definite covariance matrix

Usage

makePositiveDefinite(m1, m2 = NULL, m3 = NULL, threshold = 0.1)

Arguments

Value

list with

Description

For internal use only. Retrieve pathway id and names from Pathways object.

Usage

mapPathwaysIDfromGraphite(pathways, pathwayNames = NULL)

Arguments

Value

a data frame, id and pathway name

Description

For internal use only. Get back minimum or NA.

Usage

minOrNA(x)

Arguments

Value

a numeric. The minimum or NA

Examples

# minOrNA(c(1,5,0.1,NA))
# minOrNA(c(NA,NA,NA))

Description

MOSClip R package implements a statistical approach able to integrate multi-omic data and look for survival associated gene modules. It integrates multiple omics - trascriptomics, methylomics, genomic mutations, and genomic copy number variations - using various data dimensionality reduction strategies and multivariate models. Exploiting graph theory, pathways can be decomposed into their connected components, that we call modules. The analysis can then be performed at the level of entire pathways or pathway modules. MOSClip pathway analysis serves two primary purposes: testing the survival association of pathways or modules using the Cox proportional hazard model, and conducting a two-class analysis with a generalized linear model. Additionally, the package offers valuable graphical tools to visualize and interpret the results.

Details

To conduct a multi-omic survival analysis on pathways or modules use:

• multiOmicsSurvivalPathwayTest

• multiOmicsSurvivalModuleTest

To perform a two-class comparison enrichment analysis on pathways or modules use:

• multiOmicsTwoClassPathwayTest

• multiOmicsTwoClassModuleTest

Author(s)

Maintainer: Paolo Martini [email protected] (ORCID)

Authors:

• Anna Bortolato [email protected] (ORCID)

• Anna Tanada [email protected] (ORCID)

• Enrica Calura [email protected] (ORCID)

• Stefania Pirrotta [email protected] (ORCID)

• Federico Agostinis [email protected]

References

Paolo Martini, Monica Chiogna, Enrica Calura, and Chiara Romualdi. 2019. “MOSClip: Multi-Omic and Survival Pathway Analysis for the Identification of Survival Associated Gene and Modules.” Nucleic Acids Research 47 (14): e80. https://doi.org/10.1093/nar/gkz324

See Also

Useful links:

• https://github.com/CaluraLab/MOSClip/

• Report bugs at https://github.com/CaluraLab/MOSClip/issues

Description

An Omics class object containing data from TCGA ovarian cancer. The TCGA data was manually selected and preprocessed. It contains 4 omics: expression, methylation, mutation, and copy number variation. Additionally, it contains specific arguments to perform the dimensionality reduction. The datasets were downloaded from TCGA using TCGABiolink R package, selecting only patients with primary solid tumors. Expression matrix was processed first, converting gene identifiers into Entrez IDs. The profiles of genes present more than once were averaged. Genes with at least 100 counts in at least one patients were selected, to avoid data sparsity. Mutation matrix was filtered, keeping only genes with expression data available. We chose to consider only missense and nonsense mutations and mutation impact was also considered following Mutect2 pipeline. CNV values were transformed into numeric values. Methylation data were processed with Methyl Mix R package. Patients that had both normal and primary tumors samples were selected. With the help of a dictionary array probes were connected to CpG clusters, and finally CpG clusters were mapped to genes (Entrez ID). Survival annotation curated by Liu et al. (2018) was used to extract PFS information. Only patients with matched data across the four omics were considered. After the selection of patients and genes, we performed expression normalization and log2 of the counts+1 transformation. This will ensure us to work with expression data approximately close to a normal distribution, the most suitable distribution for the subsequent MOSClip tests. Genes and samples were manually selected to create this small example dataset for demonstration purposes.

Usage

data('multiOmics')

Format

multiOmics

An Omics with 4 omics:

exp

Matrix with 151 rows and 50 columns of RNA expression values

met

A matrix with 178 rows and 50 columns of methylation data with probes clustered

mut

A matrix with 107 rows and 50 columns of mutation counts

cnv

A matrix with matrix with 145 rows and 50 columns of copy number

...

Description

This class organizes the results of the Multi Omics Module Test analysis, in which corresponds to one pathway decomposed into modules.

Usage

## S4 method for signature 'MultiOmicsModules'
showModule(object)

Arguments

Methods (by generic)

• showModule(MultiOmicsModules): shows module info

Slots

alphas

a numeric vector of the pvalues of all the modules.

zlists

a list of numeric vectors with the zs of the covariates for each module.

modulesView

a list of module information: for each omic, the name of the omic, the genes used, the method, the name of the covariates analyzed and other specific information based on the omic.

modules

a list with the genes that belong to the module.

title

the name of the pathway.

analysis

the type of analysis done: survival or two-class.

Description

This class organize the results of the Multi-Omics Pathway Survival Test analysis.

Usage

## S4 method for signature 'MultiOmicsPathway'
showPathway(object)

Arguments

Methods (by generic)

• showPathway(MultiOmicsPathway): shows module info

Slots

pvalue

a numeric vector of the pvalues of the pathways.

zlist

a numeric vector with the zs of all the covariates.

pathView

a list of pathway information: for each omic, the name of the omic, the genes used, the method, the name of the covariates analyzed and other specific information based on the omic.

title

the name of the pathway.

analysis

the type of analysis done: survival or two-class.

Description

Performs survival analysis using an Omics object. The pathway (graph) used is decomposed in modules (cliques) using graph theory.

Usage

multiOmicsSurvivalModuleTest(
  omicsObj,
  graph,
  survFormula = "Surv(days, status) ~",
  autoCompleteFormula = TRUE,
  useTheseGenes = NULL,
  pathName = NULL,
  robust = FALSE,
  include_from_annot = FALSE
)

Arguments

Value

MultiOmicsModules object

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOM_survival <- multiOmicsSurvivalModuleTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

Description

Performs topological survival analysis using an Omics object.

Usage

multiOmicsSurvivalPathwayTest(
  omicsObj,
  graph,
  survFormula = "Surv(days, status) ~",
  autoCompleteFormula = TRUE,
  useTheseGenes = NULL,
  pathName = NULL,
  robust = FALSE,
  include_from_annot = FALSE
)

Arguments

Value

MultiOmicsPathway object

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOP_survival <- multiOmicsSurvivalPathwayTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

Description

An Omics class object containing data from TCGA ovarian cancer. The data are the same as in multiOmics object. Arguments in specificArgs slot have been set to efficiently run a topological pathway analysis, i.e., the topological method is used for PCA and shrink parameter is set to TRUE. This method can't be used for analyses on modules.

Usage

data('multiOmicsTopo')

Format

multiOmicsTopo

An Omics with 4 omics:

exp

Matrix with 151 rows and 50 columns of RNA expression values

met

A matrix with 178 rows and 50 columns of methylation data with probes clustered

mut

A matrix with 107 rows and 50 columns of mutation counts

cnv

A matrix with matrix with 145 rows and 50 columns of copy number

...

Description

Performs topological two-class analysis using an Omics object. It decomposes graphs (pathways) into modules.

Usage

multiOmicsTwoClassModuleTest(
  omicsObj,
  graph,
  classAnnot,
  baseFormula = "classes ~",
  autoCompleteFormula = TRUE,
  useTheseGenes = NULL,
  nullModel = "classes ~ 1",
  pathName = NULL
)

Arguments

Value

MultiOmicsModule object

Examples

data("multiOmics")
data("reactSmall")

genesToUse <- row.names(multiOmics[[1]])

classAnnot <- data.frame(
    "treatment" = c(rep("A", 25), rep("B", 25)),
    row.names = colnames(multiOmics[[1]])
)

MOM_twoclasses <- multiOmicsTwoClassModuleTest(
    multiOmics, reactSmall[[1]], classAnnot,
    baseFormula = "treatment ~ ", nullModel = "treatment ~ 1",
    useTheseGenes = genesToUse
)

Description

Performs topological two-class analysis using an Omics object.

Usage

multiOmicsTwoClassPathwayTest(
  omicsObj,
  graph,
  classAnnot,
  baseFormula = "classes ~ ",
  autoCompleteFormula = TRUE,
  useTheseGenes = NULL,
  nullModel = "classes ~ 1",
  pathName = NULL
)

Arguments

Value

MultiOmicsPathway object

Examples

data("multiOmics")
data("reactSmall")

genesToUse <- row.names(multiOmics[[1]])

classAnnot <- data.frame(
    "treatment" = c(rep("A", 25), rep("B", 25)),
    row.names = colnames(multiOmics[[1]])
)

MOP_twoClasses <- multiOmicsTwoClassPathwayTest(
    multiOmics, reactSmall[[1]], classAnnot,
    baseFormula = "treatment ~ ", nullModel = "treatment ~ 1",
    useTheseGenes = genesToUse
)

Description

Summarizes the results of a multi omics module test given a list of MultiOmicsModules objects

Usage

multiPathwayModuleReport(multiPathwayModuleList, priority_to = NULL)

Arguments

Value

a data.frame class object. Rows correspond to the modules, and the columns to the overall and covariates pvalues of the test.

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOM_list <- lapply(reactSmall[1:2], function(g) {
    fcl <- multiOmicsSurvivalModuleTest(multiOmics, g,
        survFormula = "Surv(days, status) ~",
        autoCompleteFormula = TRUE,
        useTheseGenes = genesToUse
    )
    fcl
})

moduleSummary <- multiPathwayModuleReport(MOM_list)

Description

Given the list of MOPs, it creates the table.

Usage

multiPathwayReport(multiPathwayList, priority_to = NULL)

Arguments

Value

a data.frame, pathways in rows, overall pvalue of the coxph, followed by covariates pvalues, in columns.

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOP_list <- lapply(reactSmall, function(g) {
    fcl <- multiOmicsSurvivalPathwayTest(multiOmics, g,
        survFormula = "Surv(days, status) ~",
        autoCompleteFormula = TRUE,
        useTheseGenes = genesToUse
    )
    fcl
})

pathwaysSummary <- multiPathwayReport(MOP_list)

Description

This class is the storage for the different omic datasets that we need to analyze. It is based on MultiAssayExperiment.

Usage

## S4 method for signature 'Omics'
showOmics(object)

Arguments

Methods (by generic)

• showOmics(Omics): shows model parameters

Slots

modelInfo

a list with length equal to length(data) that are modelInfo to process each dataset.

specificArgs

a list with length equal to length(data) to set additional parameters specific of the modelInfo.

Description

An ExperimentList class object containing data from TCGA ovarian cancer. The TCGA data was manually selected and preprocessed. It contains 4 omics: expression, methylation, mutation, and copy number variation.

Usage

data('ovarianDataset')

Format

ExperimentList

An ExperimentList with 4 omics:

exp

Matrix with 101 rows and 50 columns of RNA expression values

met

A matrix with 97 rows and 50 columnsof methylation data with probes clustered

mut

A matrix with 55 rows and 50 columns of mutation counts

cnv

A matrix with matrix with 101 rows and 50 columns of copy number

...

Description

Plots the frequencies of the pathway fathers by every omics intersection from a data.frame of the frequencies returned with the function computeFreqs.

Usage

plotFrequencies(
  frequencies,
  manualColors = NULL,
  minSize = 4,
  maxSize = 20,
  width = 20,
  relMagnificationOfLegend = 0.5,
  lineSize = 1
)

Arguments

Value

a circular plot of the frequencies of pathway fathers

Examples

df <- data.frame(
    category = c("PathwayA", "PathwayB", "PathwayC", "PathwayD"),
    frequencies = c(1, 2, 1, 3),
    class = rep("Mut", 4), stringsAsFactors = FALSE
)
plotFrequencies(df)

Description

It creates a heatmap of the most involved genes of each omic of a specific module from a MultiOmicsModule object.

Usage

plotModuleHeat(
  moduleobj,
  moduleNumber,
  sortBy = NULL,
  paletteNames = NULL,
  additionalAnnotations = NULL,
  additionalPaletteNames = NULL,
  withSampleNames = TRUE,
  fontsize_row = 10,
  fontsize_col = 1,
  nrowsHeatmaps = 3,
  orgDbi = "org.Hs.eg.db",
  discr_prop_pca = 0.15,
  discr_prop_events = 0.05,
  ...
)

Arguments

Value

A heatmap of a pathway module (results of the module test)

Examples

data(multiOmics)
data(reactSmall)

survAnnot <- data.frame(
    status = multiOmics$status,
    days = multiOmics$days,
    row.names = colnames(multiOmics[[1]])
)

genesToUse <- row.names(multiOmics[[1]])

MOM_survival <- multiOmicsSurvivalModuleTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

plotModuleHeat(MOM_survival, 1,
    sortBy = c("mut", "expPC1", "status", "days"),
    additionalAnnotations = survAnnot,
    additionalPaletteNames = list(status = "teal", days = "violet"),
    withSampleNames = F
)

Description

From a MultiOmicsModules object, it plots the position of a given module in the pathway. The omics are also represented in the graph.

Usage

plotModuleInGraph(
  modulesobj,
  pathList,
  moduleNumber,
  orgDbi = "org.Hs.eg.db",
  paletteNames = NULL,
  legendLabels = NULL,
  fileName = NULL,
  discr_prop_pca = 0.15,
  discr_prop_events = 0.05,
  pathTitle = NULL,
  ...
)

Arguments

Value

a MOSClip plot in form of a list class object

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOM_survival <- multiOmicsSurvivalModuleTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

plotModuleInGraph(MOM_survival, reactSmall,
    moduleNumber = 1,
    paletteNames = c(exp = "red", met = "green", 
                     mut = "blue", cnv = "yellow")
)

Description

Given a MultiOmicsModule class object and a specific module number, it plots Kaplan-Meier curves, in which the strata corresponds to the omics

Usage

plotModuleKM(
  MOM,
  moduleNumber,
  formula = "Surv(days, status) ~ PC1",
  fileName = NULL,
  paletteNames = NULL,
  h = 9,
  w = 7,
  risk.table = TRUE,
  pval = TRUE,
  size = 1,
  inYears = FALSE,
  discr_prop_pca = 0.15,
  discr_prop_events = 0.05,
  additional_discrete = NULL,
  additional_continuous = NULL,
  ...
)

Arguments

Value

a ggsurvplot class object

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOM_survival <- multiOmicsSurvivalModuleTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

plotModuleKM(MOM_survival, 1,
    formula = "Surv(days, status) ~ mut + expPC2",
    paletteNames = "Paired", inYears = TRUE
)

Description

Given a MultiOmicsModules object, it plots its results in a tabular fashion

Usage

plotModuleReport(
  modulesObj,
  MOcolors = NULL,
  priority_to = NULL,
  fontsize = 12,
  ...
)

Arguments

Value

a Heatmap list object from ComplexHeatmap package of the results contained in the MultiOmicsModules object provided

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOM_survival <- multiOmicsSurvivalModuleTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

plotModuleReport(MOM_survival,
    MOcolors = c(
        exp = "red", met = "green", mut = "blue",
        cnv = "yellow"
    )
)

Description

Given the list of MOPs, it plots a table of its results.

Usage

plotMultiPathwayReport(
  multiPathwayList,
  top = 25,
  MOcolors = NULL,
  priority_to = NULL,
  fontsize = 6,
  ...
)

Arguments

Value

a Heatmap list object from ComplexHeatmap package of the results contained in the MultiOmicsPathway object provided

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOP_list <- lapply(reactSmall, function(g) {
    fcl <- multiOmicsSurvivalPathwayTest(multiOmics, g,
        survFormula = "Surv(days, status) ~",
        autoCompleteFormula = TRUE,
        useTheseGenes = genesToUse
    )
    fcl
})

plotMultiPathwayReport(MOP_list,
    MOcolors = c(
        exp = "red", met = "green", mut = "blue",
        cnv = "yellow"
    ),
    fontsize = 12
)

Description

Given the pathway, it creates the heatmaps of the mostly involved genes for each omic.

Usage

plotPathwayHeat(
  pathway,
  sortBy = NULL,
  paletteNames = NULL,
  additionalAnnotations = NULL,
  additionalPaletteNames = NULL,
  discr_prop_pca = 0.15,
  discr_prop_events = 0.05,
  withSampleNames = TRUE,
  nrowsHeatmaps = 3,
  orgDbi = "org.Hs.eg.db",
  ...
)

Arguments

Value

An object of class ggplot plotted with ComplexHeatMap package.

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

survAnnot <- data.frame(
    status = multiOmics$status,
    days = multiOmics$days,
    row.names = colnames(multiOmics[[1]])
)

# Creating the MultiOmicsPathway object
MOP_survival <- multiOmicsSurvivalPathwayTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

# Plotting
plotPathwayHeat(MOP_survival,
    sortBy = c("expPC2", "mut", "status", "days"),
    paletteNames = c(exp = "red", met = "green", 
                     mut = "blue", cnv = "yellow"),
    additionalAnnotations = survAnnot,
    additionalPaletteNames = list(status = "teal", days = "violet"),
    nrowsHeatmaps = 2, withSampleNames = F
)

Description

Given a MultiOmicsPathway class object, it plots Kaplan-Meier curves, in which the strata corresponds to the chosen omics

Usage

plotPathwayKM(
  pathway,
  formula = "Surv(days, status) ~ PC1",
  fileName = NULL,
  paletteNames = NULL,
  h = 9,
  w = 7,
  risk.table = TRUE,
  pval = TRUE,
  size = 1,
  inYears = FALSE,
  discr_prop_pca = 0.15,
  discr_prop_events = 0.05,
  additional_discrete = NULL,
  additional_continuous = NULL,
  ...
)

Arguments

Value

a ggsurvplot class object

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

# Creating the MultiOmicsPathway object
MOP_survival <- multiOmicsSurvivalPathwayTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

plotPathwayKM(MOP_survival,
    formula = "Surv(days, status) ~ mut + expPC2",
    paletteNames = "Paired", inYears = TRUE
)

Description

Compute pvalue Summary

Usage

pvalueSummary(multiPathwayReportData, excludeColumns = NULL, as.list = FALSE)

Arguments

Value

a list

Description

A PathwayList with three pathways necessary for the analysis: 'Activation of Matrix Metalloproteinases', 'FGFR1 mutant receptor activation', and 'VEGFA-VEGFR2 Pathway'. Pathways were downloaded using graphite package and the names of the nodes were converted into Entrez IDs.

Usage

data('reactSmall')

Format

reactSmall

A PathwayList with Reactome pathways for hsapiens

entries

Three Reactome pathways with their nodes

Description

Remove the self loops that a present in the graph graphNEL object

Usage

removeSelfLoops(graph)

Arguments

Value

a graphNEL object

#' @rdname graph-processing

Description

Resampling function for survival analysis on modules

Resampling function for pathways (survival analysis)

Usage

resamplingModulesSurvival(
  fullMultiOmics,
  pathdb,
  nperm = 100,
  pathwaySubset = NULL,
  nPatients = 3,
  genesToConsider = NULL
)

resamplingPathwaySurvival(
  fullMultiOmics,
  pathdb,
  nperm = 100,
  pathwaySubset = NULL,
  nPatients = 3,
  genesToConsider = NULL
)

Arguments

Value

list of the resampling tables of results

list of the resampling tables of results

Examples

data(multiOmics)
data(reactSmall)

perms <- resamplingModulesSurvival(
    fullMultiOmics = multiOmics, reactSmall,
    nperm = 10,
    pathwaySubset =
        "FGFR1 mutant receptor activation"
)

Description

Resampling function for two-class analysis on modules

Resampling function for pathways (two-class analysis)

Usage

resamplingModulesTwoClass(
  fullMultiOmics,
  classAnnot,
  pathdb,
  nperm = 100,
  pathwaySubset = NULL,
  nPatients = 3,
  genesToConsider = NULL
)

resamplingPathwayTwoClass(
  fullMultiOmics,
  classAnnot,
  pathdb,
  nperm = 100,
  pathwaySubset = NULL,
  nPatients = 3,
  genesToConsider = NULL
)

Arguments

Value

list of the resampling tables of results

list of the resampling tables of results

Examples

data(multiOmics)
data(reactSmall)

classAnnot <- data.frame(
    "treatment" = c(rep("A", 25), rep("B", 25)),
    row.names = colnames(multiOmics[[1]])
)

perms <- resamplingModulesTwoClass(
    fullMultiOmics = multiOmics,
    classAnnot, reactSmall,
    nperm = 10,
    pathwaySubset =
        "FGFR1 mutant receptor activation"
)

Description

This function performs a exact test implementing a theorical framework using the SuperExactTest package. It calculates the statistical distributions of multi omics set intersections. It can be used with both a MultiOmicsModules or MultiOmicsPathway class objects.

Usage

runSupertest(
  multiPathwayReportData,
  pvalueThr = 0.05,
  zscoreThr = 0.05,
  resampligThr = NULL,
  plot = c("circular", "landscape", "noplot"),
  sort.by = c("set", "size", "degree", "p-value"),
  excludeColumns = NULL,
  color.on = "#f6bb42",
  color.off = "#D3D3D3"
)

Arguments

Details

This function calculates intersection sizes between multiple set of pathways or modules and performs statistical test of the intersections using the total amout of analyzed pathways or modules as background. The super exact test of this function was described in Wang et al 2015.

Value

a data.frame containing all the numeric information of the plot included the pathways shared by different omics.

References

Minghui Wang, Yongzhong Zhao, and Bin Zhang (2015). Efficient Test and Visualization of Multi-Set Intersections. Scientific Reports 5: 16923.

Examples

df <- data.frame(
    pvalue = c(0.06, 0.04, 0.04, 0.03, 0.02),
    cnv = c(0.07, 0.03, 0.02, 0.04, 0.01),
    mut = c(0.08, 0.02, 0.01, 0.04, 0.04),
    row.names = c(
        "PathwayA", "PathwayB", "PathwayC",
        "PathwayD", "PathwayE"
    )
)

runSupertest(df, pvalueThr = 0.05, zscoreThr = 0.05)

Description

Select stable pathway modules

Count the resampling success

Add resampling counts to module summary

Usage

selectStablePathwaysModules(perms, moduleSummary, success = 90, col = "pvalue")

getPathwaysModulesSuccess(perms, moduleSummary, col = "pvalue", thr = 0.05)

addResamplingCounts(moduleSummary, resamplingCounts)

Arguments

Value

the subset of stable modules

the counts of success for each pathway or module

a module or pathway summary with resampling counts column appended

Examples

data("multiOmics")
data("reactSmall")

perms <- resamplingPathwaySurvival(multiOmics, reactSmall, nperm = 5)
res <- lapply(reactSmall, function(g) {
    multiOmicsSurvivalPathwayTest(multiOmics, g,
        useTheseGenes = row.names(multiOmics[[1]])
    )
})
pathSummary <- multiPathwayReport(res)
getPathwaysModulesSuccess(perms, pathSummary)

Description

A generic function showing pathway's module info

Usage

showModule(object)

Arguments

Value

NULL. No value is returned

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOM_survival <- multiOmicsSurvivalModuleTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

showModule(MOM_survival)

Description

This function shows the MOSClip palette. Each omic should be coupled to a color panel, this match will be preserved in plots.

Usage

showMOSpalette()

Value

NULL. No value is returned

Examples

showMOSpalette()

Description

A generic functions showing parameter associated with each omics

Usage

showOmics(object)

Arguments

Value

NULL. No value is returned

Examples

data(multiOmics)

showOmics(multiOmics)

Description

A generic function showing pathway info

Usage

showPathway(object)

Arguments

Value

NULL. No value is returned

Examples

data(multiOmics)
data(reactSmall)

genesToUse <- row.names(multiOmics[[1]])

MOP_survival <- multiOmicsSurvivalPathwayTest(multiOmics, reactSmall[[1]],
    survFormula = "Surv(days, status) ~", autoCompleteFormula = TRUE,
    useTheseGenes = genesToUse
)

showPathway(MOP_survival)

Description

Sparse PCA

Usage

sparseCompPCs(exp, shrink, k)

Arguments

Value

a list with the following elements:

Description

Function to remove the suffix corresponding to the module number of the pathway name. Necessary step for annotePathwayToFather and plotFrequencies

Usage

stripModulesFromPathways(pathways)

Arguments

Value

list of pathway names without the module number

Examples

pathwaysModules <- list(
    "Intrinsic Pathway for Apoptosis.1",
    "Intrinsic Pathway for Apoptosis.2",
    "Opioid Signalling.1", "Opioid Signalling.2"
)

resPathwayNames <- stripModulesFromPathways(pathwaysModules)

Description

Given a matrix it summarize in classes

Usage

summarizeInCluster(
  data,
  features,
  name = "clu",
  dictionary = NULL,
  max_cluster_number = 3,
  cliques = NULL
)

Arguments

Details

The user can define a maximum of classes. The function guess the optimal number of clusters using NbClust methods.

Value

a list with summary of the omic:

Description

For internal use only. for each line extrac 'col' and get the minimum.

Usage

summarizeOmicsResByMinPvalue(col, mat)

Arguments

Value

a summarized version of the matrix.

Examples

# summarizeOmicsResByMinPvalue(2:3, mat=matrix(c(1,2,4,1,2,5), nrow=2))

Description

Given a matrix it summarize the positive and negative to 0 or 1 in two vectors

Usage

summarizeToBinaryDirectionalEvents(
  data,
  features,
  name = "dirBin",
  binaryClassMin = 10,
  eventThr = 2,
  cliques = NULL
)

Arguments

Value

a list with summary of the omic:

Description

Given a matrix it summarize to a 0 or 1

Usage

summarizeToBinaryEvents(
  data,
  features,
  name = "bin",
  binaryClassMin = 10,
  cliques = NULL
)

Arguments

Value

a list with summary of the omic:

Description

Given a matrix it summarize the positive and negative in two vectors, with counts of the events

Usage

summarizeToNumberOfDirectionalEvents(
  data,
  features,
  name = "dCount",
  eventThr = 2,
  min_prop = 0.1,
  cliques = NULL
)

Arguments

Value

a list with summary of the omic:

Description

Given a matrix it summarize to a 0 or 1

Usage

summarizeToNumberOfEvents(
  data,
  features,
  name = "event",
  min_prop = 0.1,
  cliques = NULL
)

Arguments

Value

a list with summary of the omic:

Description

Given a matrix it summarize to principal components. The user can specify the number of principal components. Default 3.

Usage

summarizeWithPca(
  data,
  features,
  name = "pca",
  shrink = FALSE,
  method = "regular",
  cliques = NULL,
  maxPCs = 3,
  loadThr = 0.6
)

Arguments

Value

a list with summary of the omic:

Description

Cox Analysis

Usage

survivalcox(coxObj, formula)

Arguments

Details

For internal use only

Value

A list with

Description

Cox Robust Analysis

Usage

survivalcoxr(coxObj, formula)

coxrsummary(x)

Arguments

Details

For internal use only

Value

A list with

a list with wald test and robust and partial coefficients

Description

Topological PCA

Usage

topoCompPCs(exp, shrink, cliques, k)

Arguments

Value

a list with the following elements: