[BioC] question on ChIPpeakAnno::getEnrichedGO

[Zhu, Lihua (Julie)]

Cam,

Thanks for the feedback and your kind word!

Please see my response below. BTW, could you please cc the Bioconductor list
bioconductor at r-project.org next time so that others could contribute or/and
benefit? Thanks!

Best regards,

Julie



On 8/28/13 1:29 PM, "Camila Lopez-anido" <lopezanido at wisc.edu> wrote:

> Hi again, 
> 
> Thanks for putting together the ChIPpeakAnno package! It's been really nice to
> use. 
> 
> 1. I have a question about getEnrichedGO. I was wondering what you would
> recommend to make the test more relaxed? Currently, I'm using maxP=0.05,
> multiAdj=TRUE, minGOterm = 1, multiAdjMethod="BH". Should I not run multiAdj,
> or use another multiAdjMethod?

I usually use maxP = 0.05, multiAdj=TRUE, minGOterm = 5, multiAdjMethod="BH"

You could try to set maxP = 0.1 and see if there are some terms with
adjusted p-values on the boarder line.

> 
> 2. Also, I wanted to check to see if getEnrichedGO filters the list of genes
> prior to analysis?.... since more than 1 peak can be associated with the same
> gene, which means that the same gene may be counted twice or more. Do
> researchers usually use that data as is (i.e. multiples of the same gene), or
> do they filter for a unique list prior to GO analyses?

Very good question. Yes, the gene list is filtered to contain a unique gene
list prior to GO anlayis.

> 
> Thanks,
> Cam