[BioC] GOHyperG pvalue and expression value,

[Sean Davis]

Saurin,

GOHyperG computes the p-value from the hypergeometric distribution.  You 
need four numbers to compute such a p-value:
1) total genes on chip annotated in the ontology
2) total genes on chip annotated with specific term of interest
3) total genes in diff. expressed list annotated in the ontology
4) total genes in diff. expressed list annotated with specific term of 
interest.

See pretty much any paper on calculating "enrichment" of gene ontology terms 
for a review of the process.  Also, have you looked at the help for 
GOHyperG.  There is a discussion about the methods and a reference to 
phyper, the function that does the hypergeometric p-value calculation.

Sean

----- Original Message ----- 
From: "Saurin Jani" <saurin_jani at yahoo.com>
To: "Bioconductor Bioconductor" <bioconductor at stat.math.ethz.ch>
Sent: Monday, March 07, 2005 10:35 AM
Subject: [BioC] GOHyperG pvalue and expression value,


> Hi BioC,
>
> May be this is basic question,please forgive me for
> that...but anyone can help me understand,how GOHyperG
> pvalue consider expression intensities in order to
> calculate pvalue to make GO Terms more important that
> other GO Terms exists on chip or differential
> expressed gene set? for instance: Bio Process:
>
> [1]
> GO:0006333
> chromatin assembly or disassembly
> 2 Genes associated out of my differ. Exp. Genes
> pvalue: 0.00110121111270826
> total genes from chip : 16
>
> [2]
> GO:0030199
> collagen fibril organization
> 2 Genes associated out of my differ. Exp. Genes
> pvlaue: 0.0929493556475916
> total genes from chip : 14
>
> GOHyperG function takes unique LLids..and environment
> and what GO Term we need information for..! there is
> no  argument about expression set?
>
> Thank you so much in advance,
> Saurin
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>