nbOccInComp: Select components the features contribute to

In MineICA: Analysis of an ICA decomposition obtained on genomics data

Description

For each feature/gene, this function returns the components they contribute to and their projection values across all the components.

Usage

  nbOccInComp(icaSet, params, selectionByComp = NULL,
    level = c("features", "genes"), file = NULL)

Arguments

icaSet	An object of class IcaSet
params	An object of class MineICAParams containing the parameters of the analysis, the attribute cutoffSel is used as a threshold on the absolute projections to determine which genes contribute to the components.
selectionByComp	The list of components already restricted to the contributing genes
level	The attribute of icaSet to be used, are reported the occurences of either the "features" or the "genes".
file	The file where the output data.frame and plots are written.

Details

A feature/gene is considered as a contributor when its scaled projection value exceeds the threshold selCutoff(icaSet).

This function plots the number of times the feature/gene is a contributor as a function of the standard deviation of its expression profile.

The created files are located in genePath(params). An extensiom '.htm' and '.pdf' is respectively added to the file name for the data.frame and the plot outputs.

Value

Returns a data.frame whose columns are: 'gene' the feature or gene ID, 'nbOcc' the number of components on which the gene contributes according to the threshold, 'components' the indices of these components, and then the component indices which contain its projection values.

Author(s)

Anne Biton

Examples

data(icaSetCarbayo)
params <- buildMineICAParams(resPath="carbayo/")
nbOcc <- nbOccInComp(icaSet=icaSetCarbayo, params=params, level="genes", file="gene2MixingMatrix")

MineICA documentation built on Nov. 8, 2020, 5:35 p.m.