R3CPET (version 1.4.2)

loadPPI-methods: Setting the background protein-protein interaction

Description

This method enables the user to define the PPI network as a background network. The user can provide his own PPI or use the HPRD or the Biogrid PPI incorporated in the package.

Usage

"loadPPI"(object,type=c("HPRD","Biogid"),customPPI= NULL,  filter=FALSE, term ="GO:0005634", annot=NULL,  RPKM= NULL, threshold=1 )

Arguments

object

a ChiapetExperimentData to load the PPI into as an igraph object

type

if customPPI = NULL, this parameter indicats which of the available PPI to use HPRD or Biogrid (build 3.2.100). by default, the HPRD network is used.

customPPI

If the user wants to use his own PPI interaction network (for example for another species), he can provide an igraph object or a path to an "ncol" formatted graph (a file with two columns indicating the interacting parts and an optional third column to indicate the weight). It is preferred that the user provides an igraph object, to avoid any problems when parsing the "ncol" file.

filter

This parameter indicates whether the user want to filter the selected PPI or not. filter = FALSE means that the PPI (provided by the user or by the HPRD and Biogrid networks) will be used as is. if filter = TRUE,the proteins in the PPI will be filtered according to their position in cell ( by default proteins located in the nucleus are kept and the other removed). In addition to the filtering by location, if the user wants just to keep the proteins that show a certain amount of expression he can provide a gene expression table to the RPKM parameter and set the threshold for filtering.

term

The GO term used to for filtering. By default, only protein that are located in the nucleus are kept (term = "GO:0005634"). If the user want to use another annotation table he can pass a data.frame object to the parameter annot.

annot

If the user wants to provide his own annotation data-set, he can pass a data.frame object to this parameter. The passed data.frame should have at least two columns named respectively: geneSymbol that contains the gene names, and cellular_component_term that contain the term.

    geneSymbol                    cellular_component_term
    FAU Ribosome      (GO:0005840); Nucleolus (GO:0005730)
    ALDH3A1   Cytoplasm (GO:0005737); Nucleus (GO:0005634)
    ASCL1     Nucleus (GO:0005634); Cytoplasm (GO:0005737)
  

RPKM

A data.frame that contains the expression of each gene in the PPI. The data.frame should at least have 2 columns. The first one contains the gene symbol (should be the same as the one used in the PPI) and the second gives the expression.

threshold

Threshold value used to filter gene expression. All genes with expression value less than threshold are removed.

Value

A ChiapetExperimentData object in which the ppi slot is populated as an igraph object filtered according to the specified conditions.

References

Prasad, T. S. K. et al. (2009) Human Protein Reference Database - 2009 Update. Nucleic Acids Research. 37, D767-72.

Chatr-Aryamontri A, Breitkreutz BJ et al. The BioGRID Interaction Database: 2013 update. Nucleic Acids Res. 2012 Nov 30

M.N Djekidel et al,3CPET: Finding Co-factor Complexes in Chia-PET experiment using a Hierarchical Dirichlet Process, in press, 2015

See Also

ChiapetExperimentData, loadTFBS , loadPETs, createIndexes

Examples

## Create a ChiapetExperimentData object
x <- ChiapetExperimentData(ppiType= "HPRD")

## Loading the default HPRD network without filtering

x <- loadPPI(x,type="HPRD")
ppi(x)

## Using the HPRD network and filtering using the GO:0005634
x <- loadPPI(x,type="HPRD", filter=TRUE)
ppi(x)

data(RPKMS)
x <- loadPPI(x,type="HPRD",filter=TRUE,annot= NULL, RPKM= RPKMS, threshold = 5)
ppi(x)