MineICA (version 1.12.0)

writeProjByComp: writeProjByComp

Description

This function writes in an html file the description of the features, or genes, that contribute to each component. It also writes an html file containing, for each feature or gene, its projection value on every component.

Usage

writeProjByComp(icaSet, params, mart = useMart(biomart = "ensembl",  dataset = "hsapiens_gene_ensembl"), typeRetrieved = NULL, addNbOcc = TRUE, selectionByComp = NULL, level = c("features", "genes"), typeId, selCutoffWrite=2.5)

Arguments

icaSet

An object of class IcaSet

params

An object of class MineICAParams containing the parameters of the analysis. The files are written in the path genesPath(params). selCutoff(params) is used to select the features or genes by component.

mart

An output of function useMart containing the database used for annotation.

typeRetrieved

The annotations biomaRt is queried about. They describe the feature or gene IDs of the argument icaSet, see listFilters.

addNbOcc

If TRUE, the number of components the features/genes contribute to is added to the output. A gene/feature is considered as a contributor of a component if its absolute scaled projection value is higher than selCutoff(icaSet).

selectionByComp

A list containing the feature/gene projections on each component, already restricted to the ones considered as contributors.

level

The data level of icaSet that will be annotated: either the feature projections ("features"), or the gene projections ("genes").

typeId

The type of ID the features or the genes of icaSet correspond to. By default typeID(icaSet) is used. It must be provided in the biomaRt way (type listFilters(mart) to choose the appropriate value).

selCutoffWrite

The cutoff applied to the absolute projection values to select the features/genes that will be annotated using package biomaRt, default is 2.5.

Value

listAnnotComp:

a list with the output of writeGenes for each component

nbOccInComp:

a data.frame storing the projection values of each feature/gene (row) across all the components (columns).

Details

One file is created by component, each file is named by the index of the components (indComp(icaSet)) and located in the path genePath(params).

In case you are interested in writing the description of features and their annotations, please remember to modify codegenesPath(params), or the previous files will be overwritten.

The genes are ranked according to their absolute projection values.

This function also writes an html file named "genes2comp" providing, for each feature or gene, the number of components it contributes to (according to the threshold cutoffSel(params)), and its projection value on all the components. The projection values are scaled.

See function writeGenes for details.

See Also

writeGenes, getBM, listFilters, listAttributes, useMart, selectContrib, nbOccInComp

Examples

## Not run: 
# ## load IcaSet object
# ## We will use 'icaSetCarbayo', whose features are hgu133a probe sets
# ## and feature annotations are Gene Symbols. 
# data(icaSetCarbayo)
# 
# ## define database to be used by biomaRt
# mart <- useMart(biomart="ensembl", dataset="hsapiens_gene_ensembl")
# 
# ## define the parameters of the analysis
# params <- buildMineICAParams(resPath="~/resMineICACarbayo/", selCutoff=0)
# 
# ## Make sure the elements "_biomaRt" of attribute 'typeID' are defined
# typeID(icaSetCarbayo) 
# 
# ### Query biomaRt and write gene descriptions in HTML files
# ### The files will be located in the directory 'genesPath(params)'
# 
# ## 1. Write description of genes 
# res <- writeProjByComp(icaSet=icaSetCarbayo, params=params, mart=mart,
#            level="genes") #, typeId="hgnc_symbol")
# 
# ## 2. Write description of features 
# # change attribute 'genesPath' of params to preserve the gene descriptions
# genesPath(params) <- paste(resPath(params),"comp2features/",sep="")
# res <- writeProjByComp(icaSet=icaSetCarbayo, params=params, mart=mart,
#            level="features") #, typeId="affy_hg_u133a")
# 
# ## End(Not run)