Pathways: Pathway analysis for multiple clustering results
In IntClust: Integration of Multiple Data Sets with Clustering Techniques

Description Usage Arguments Details Value Examples

A pathway analysis per cluster per method is conducted.

Pathways(List, Selection = NULL, geneExpr = NULL, nrclusters = NULL,
  method = c("limma", "MLP"), geneInfo = NULL, geneSetSource = "GOBP",
  topP = NULL, topG = NULL, GENESET = NULL, sign = 0.05,
  fusionsLog = TRUE, weightclust = TRUE, names = NULL)

`List`	A list of clustering outputs or output of the`DiffGenes` function. The first element of the list will be used as the reference in `ReorderToReference`. The output of `ChooseFeatures` is also accepted.
`Selection`	If pathway analysis should be conducted for a specific selection of objects, this selection can be provided here. Selection can be of the type "character" (names of the objects) or "numeric" (the number of specific cluster). Default is NULL.
`geneExpr`	The gene expression matrix or ExpressionSet of the objects. The rows should correspond with the genes.
`nrclusters`	Optional. The number of clusters to cut the dendrogram in. The number of clusters should not be specified if the interest lies only in a specific selection of objects which is known by name. Otherwise, it is required. Default is NULL.
`method`	The method to applied to look for differentially expressed genes and related pathways. For now, only the limma method is available for gene analysis and the MLP method for pathway analysis. Default is c("limma","MLP").
`geneInfo`	A data frame with at least the columns ENTREZID and SYMBOL. This is necessary to connect the symbolic names of the genes with their EntrezID in the correct order. The order of the gene is here not in the order of the rownames of the gene expression matrix but in the order of their significance. Default is NULL.
`geneSetSource`	The source for the getGeneSets function, defaults to "GOBP".
`topP`	Overrules sign. The number of pathways to display for each cluster. If not specified, only the significant genes are shown. Default is NULL.
`topG`	Overrules sign. The number of top genes to be returned in the result. If not specified, only the significant genes are shown. Defaults is NULL.
`GENESET`	Optional. Can provide own candidate gene sets. Default is NULL.
`sign`	The significance level to be handled. Default is 0.05.
`fusionsLog`	Logical. To be handed to `ReorderToReference`: indicator for the fusion of clusters. Default is TRUE
`weightclust`	Logical. To be handed to `ReorderToReference`: to be used for the outputs of CEC, WeightedClust or WeightedSimClust. If TRUE, only the result of the Clust element is considered. Default is TRUE.
`names`	Optional. Names of the methods. Default is NULL.

After finding differently expressed genes, it can be investigated whether pathways are related to those genes. This can be done with the help of the function Pathways which makes use of the MLP function of the MLP package. Given the output of a method, the cutree function is performed which results into a specific number of clusters. For each cluster, the limma method is performed comparing this cluster to the other clusters. This to obtain the necessary p-values of the genes. These are used as the input for the MLP function to find interesting pathways. By default the candidate gene sets are determined by the AnnotateEntrezIDtoGO function. The default source will be GOBP, but this can be altered. Further, it is also possible to provide own candidate gene sets in the form of a list of pathway categories in which each component contains a vector of Entrez Gene identifiers related to that particular pathway. The default values for the minimum and maximum number of genes in a gene set for it to be considered were used. For MLP this is respectively 5 and 100. If a list of outputs of several methods is provided as data input, the cluster numbers are rearranged according to a reference method. The first method is taken as the reference and ReorderToReference is applied to get the correct ordering. When the clusters haven been re-appointed, the pathway analysis as described above is performed for each cluster of each method.

The returned value is a list with an element per cluster per method. This element is again a list with the following four elements:

`objects`	A list with the elements LeadCpds (the objects of interest) and OrderedCpds (all objects in the order of the clustering result)
`Characteristics`	The found (top) characteristics of the feauture data
`Genes`	A list with the elements TopDE (a table with information on the top genes) and AllDE (a table with information on all genes)
`Pathways`	A list with the element ranked.genesets.table which is a data frame containing the genesets, their p-values and their descriptions. The second element is nr.genesets and contains the used and total number of genesets.

## Not run: 
data(fingerprintMat)
data(targetMat)
data(geneMat)
data(GeneInfo)

MCF7_F = Cluster(fingerprintMat,type="data",distmeasure="tanimoto",normalize=FALSE,
method=NULL,clust="agnes",linkage="flexible",gap=FALSE,maxK=55,StopRange=FALSE)
MCF7_T = Cluster(targetMat,type="data",distmeasure="tanimoto",normalize=FALSE,
method=NULL,clust="agnes",linkage="flexible",gap=FALSE,maxK=55,StopRange=FALSE)

L=list(MCF7_F,MCF7_T)
names=c('FP','TP')

MCF7_PathsFandT=Pathways(List=L, geneExpr = geneMat, nrclusters = 7, method = c("limma", 
"MLP"), geneInfo = GeneInfo, geneSetSource = "GOBP", topP = NULL, 
topG = NULL, GENESET = NULL, sign = 0.05,fusionsLog = TRUE, weightclust = TRUE, 
 names =names)
 
## End(Not run)