[BioC] ChrMapHyperGParams and linearMTest in Category

[Marc Carlson]

Hi Daren,

We aren't trying to ignore you.  Everyone is just extra busy right now
because we are trying to do a release.  We really appreciate you letting
us know that there is a problem however.  If you could please provide me
with a specific testable example that does not work, this will greatly
help me to expedite the process of making sure that it does.  Otherwise
I have to guess about what happened and that takes A LOT longer.  Also,
please always post your sessionInfo() so that we know what version of
things you are working with.  You might also find our posting guide to
be helpful.

http://www.bioconductor.org/docs/postingGuide.html


It's my hope that you might find some information for your other
questions in the Category vignettes here:

http://www.bioconductor.org/packages/devel/bioc/html/Category.html


Thanks again for your patience,


  Marc




Daren Tan wrote:
> Hope someone or the developer of GOStats/Category can answer my queries.
>
> On Sat, Apr 4, 2009 at 11:28 AM, Daren Tan <darentan76 at gmail.com> wrote:
>   
>> Beyond the scope of the manual and vigenettes of Category and GOStats,
>> I would like to understand more about ChrMapHyperGParams and
>> linearMTest.
>>
>> After successfully using "GOHyperGParams" and "KEGGHyperGParams" with
>> platform-independent annotation database "org.Hs.eg.db", I tried to do
>> the same with "ChrMapHyperGParams" but kept getting error messages
>> that "org.Hs.egENTREZID" couldn't be found. This was unexpected as I
>> couldn't fanthom why ChrMapHyperGParams would need probes information.
>> I would like to get around this problem even with "org.Hs.eg.db" so
>> that my enrichment analysis would be complete for GO/Pathway/CHR.
>>
>> hyperGTest and linearMTest use hypergeometric distribution and linear
>> model respectively as their names implied. What situations will either
>> tests be more suitable for ? Would both tests at same p-value cutoffs
>> give similar results ? Should I even favour hyperGTest or linearMTest
>> ?
>>
>> Previously, I used david and panther and wondered is it appropriate to
>> get FDR based on bonferroni-tests from the p-values of hyperGTest and
>> linearMTest. Would FDR be meaningful for GO/Pathway/CHR enrichment
>> analysis ?
>>
>>     
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
>
>