fea2incidMat: Transforms FEA output into incidence matrices.

In FGNet: Functional Gene Networks derived from biological enrichment analyses

Description

Transforms the Funtional Enrichment Analysis (FEA) results into cluster-gene incidence matrices.

Usage

fea2incidMat(feaResults, key = "Genes", sepChar = NULL, clusterColumn = NULL, 
    filterAttribute = NULL, filterOperator = "<", filterThreshold = 0, 
    removeFilteredGtl = NULL)

Arguments

feaResults	list or data.frame/matrix. Output from one of the FEA functions.
key	"Genes" or "Terms". To build gene- or term-based networks.
sepChar	character. Character separating genes or terms in the same field. By default: "," for genes and ";" for terms.
clusterColumn	character. Name of the column that contains the value to group gene-term sets. Only required if it is different than "Cluster" or "Metagroup".
filterAttribute	character or data.frame. Attribute to filter the clusters/metagroups. Filtered clusters/metagroups will not be included in the matrices (and subsequent networks). Its value should be the data.frame column to use for filtering. It can be provided as character (column name) or data.frame (subset of the data.frame with drop=FALSE).
filterOperator	character. Logical operator used for filtering. i.e. ">" (bigger than), "<=" (smaller or equal than), "==" (equal), "!=" (different), "%%" (included in),... The evaluation order is left to right: filterAttribute ">" filterThreshold, will filter out clusters with filter attribute bigger than the threshold.
filterThreshold	numeric. Sets the value to compare to.
removeFilteredGtl	logical. Only used by GeneTerm Linker term network. If FALSE, it includes generic terms filtered by GeneTerm Linker from final metagroups.

Value

List:

clustersMatrix or metagroupsMatrix	Incidende matrix with the genes or Terms in each cluster or metagroup.
gtSetsMatrix	Incidende matrix with the genes or Terms in each gene-term set
filteredOut	Clusters or metagroups which where filtered out and therefore not included in the incidence matrices. NULL if none.

See Also

Next step in the workflow: functionalNetwork()

Overview of the package: FGNet

Package tutorial: vignette("FGNet-vignette")

Examples

## Not run: 
jobID <- 3907019
results <- fea_gtLinker_getResults(jobID)
incidMat <- fea2incidMat(results)

# Filtering (threshold)
incidMat <- fea2incidMat(results, 
    filterAttribute="Silhouette Width", filterThreshold=0.2)

incidMat$filteredOut
head(incidMat$metagroupsMatrix)
head(incidMat$gtSetsMatrix)

functionalNetwork(incidMat)
    
# Term-based network
incidMatTerms <- fea2incidMat(results, key="Terms")
functionalNetwork(incidMatTerms, plotOutput="dynamic")

# Including generic terms filterd by GtLinker from final metagroups:
incidMatTerms <- fea2incidMat(results, key="Terms",removeFilteredGtl=FALSE)
functionalNetwork(incidMatTerms, plotOutput="dynamic", plotType="bipartite")

# Filtering by keyword
keywords <- c("rna")
selectedGroups <- sapply(getTerms(results),
    function(x) 
    any(grep(paste("(", paste(keywords, collapse="|") ,")",sep=""), tolower(x))))

resultsCbind <- results
resultsCbind$metagroups <- cbind(results$metagroups,
    selectedKeywords=as.numeric(selectedGroups))

matSelectedGroups <- fea2incidMat(resultsCbind,
    filterAttribute="selectedKeywords", filterThreshold=1)

functionalNetwork(matSelectedGroups)
getTerms(results)[selectedGroups]

## End(Not run)

FGNet documentation built on Nov. 8, 2020, 5:43 p.m.