ISA.GO: Calculate Gene Ontology enrichment for transcription modules

In eisa: Expression data analysis via the Iterative Signature Algorithm

Description

Gene Ontology enrichment is calculated for each ISA module separately. In the end the result is corrected for multiple hypothesis testing.

Usage

ISAGO (modules, ann = annotation(modules), features = featureNames(modules),      
    hgCutoff = 0.05, correction = TRUE, correction.method = "holm")      

Arguments

modules	An ISAModules object, a set of ISA modules.
ann	Character scalar. The annotation package to be used. By default it is taken from the modules argument.
features	Character vector. The names of the features. By default it is taken from the modules argument.
hgCutoff	Numeric scalar. The cutoff value to be used for the enrichment significance. This can be changed later, without recalculating the test.
correction	Logical scalar, whether to perform multiple hypothesis testing correction.
correction.method	Character scalar, the multiple testing correction method to use. Possible values: “holm”, “hochberg”, “hommel”, “bonferroni”, “BH”, “BY”, “fdr”, “none”. See the p.adjust function for details on these.

Details

The Gene Ontology is a database of gene annotation. The annotating labels (these are called terms) are standardized and organized into a directed acyclic graph. In other words terms may have more specific sub-terms, that can have even more specific sub-sub-terms, and so on.

The Gene Ontology database has three big sub-graphs, the root nodes (the most general terms) of these are the direct children of the root term of the whole ontology: biological process, cellular component, molecular function. They are usually referred to as ontologies.

The hypergeometric test, a version Fisher's exact test, takes a GO term and a gene set (in our case coming from an ISA module) and asks whether the number of genes in the set annotated by the term is significantly more (or less) than what one would expect by chance.

ISAGO performs the hypergeometric test for every module, for all GO terms of the three GO ontologies. The GO data is taken from the GO.db package and the annotation package of the chip.

ISAGO currently cannot test for under-representation and the conditional test, as implemented in the GOstats package, is not available either.

Value

A list with three GOListHyperGResult objects, for the three Gene Ontologies, named

BP	aka Biological Processes
CC	aka Cellular Components
MF	aka Molecular Function

Author(s)

Gabor Csardi [e-mail]

References

The Gene Ontology Consortium. Gene ontology: tool for the unification of biology. Nat. Genet. May 2000;25(1):25-9.

Bergmann S, Ihmels J, Barkai N: Iterative signature algorithm for the analysis of large-scale gene expression data Phys Rev E Stat Nonlin Soft Matter Phys. 2003 Mar;67(3 Pt 1):031902. Epub 2003 Mar 11.

See Also

ISAKEGG, ISACHR, ISAmiRNA for other enrichment calculations.

The GO.db, GOstats and Category packages.

Examples

data(ALLModulesSmall)
GO <- ISAGO(ALLModulesSmall)
GO
summary(GO$BP)[[1]][,1:5]

eisa documentation built on Nov. 8, 2020, 6:47 p.m.