[BioC] hyperGTest interpretation of 'under' and finding all Entrez associated with a GO term

[James W. MacDonald]

Hi Mary,

On 11/12/2012 12:28 PM, Mary Kindall wrote:
> Hi Steve
> Thanks for reply.
> I am able to find all Entrez for a particular GO term, however, the 'count'
> and 'size' of HyperGTest result do not match the numbers.
>
> For example, I got this as result of my HyperGTest.
>
> "GOBPID" "Pvalue" "OddsRatio" "ExpCount" "Count" "Size" "Term"
> "GO:0032501" "9.62590502703856e-31" "2.84380308273042" "116.003908034309"
> "235" "3975" "multicellular organismal process"
>
> count = 235
> Size = 3975
>
> When I do :
>> length(org.Mm.egGO2ALLEGS[["GO:0032501"]])
> [1] 8163
>> length(org.Mm.egGO2EG[["GO:0032501"]])
> [1] 0
>
> None of them equal to 3975.

Did you use a conditional test? If so, the number of terms under 
consideration is conditional on those that were significant at a 'lower' 
level in the DAG for that term.

Best,

Jim


>
> Thanks
>
>
>
>
>
>
> On Mon, Nov 12, 2012 at 10:34 AM, Steve Lianoglou<
> mailinglist.honeypot at gmail.com>  wrote:
>
>> Hi,
>>
>> On Mon, Nov 12, 2012 at 9:30 AM, Mary Kindall<mary.kindall at gmail.com>
>> wrote:
>>>     1.
>>>
>>>     How do I interpret the result of 'hyperGTest' if the argument
>>>     "testDirection='under'" is supplied? The hypergtest result is given
>> below.
>>
>> Are you looking for help in interpreting this test in a statistical
>> sense, or a biological one?
>>
>>>     2.
>>>
>>>     Is there a way to find all Entrez IDs associated with a particular GO
>>>     term (For example, GO:0034660).
>> Here is the "old school" way -- I bet there is now a nicer way to do
>> this with the Homo.sapiens package (for instance). But let's say you
>> wanted all human genes that are annotated with SMAD binding
>> ("GO:0046332")
>>
>> R>  library(org.Hs.eg.db)
>> R>  head(org.Hs.egGO2ALLEGS[["GO:0046332"]])
>> R>    IDA   IDA   IDA   IDA   IDA   IDA
>>   "90"  "91"  "94" "650" "657" "658"
>>
>> The names of the vector tell you the evidence code for the term ->
>> gene association
>>
>>>     3.
>>>
>>>     How do I find list of all Entrez IDs which has a GO BP/CC/MF term? My
>>>     gene universe 'uGenes' initially had a total of ~22,000 genes, where
>> as the
>>>     result below show it to be 19,703. This mean that some of my genes in
>> the
>>>     universe did not have an associated GO term. How to find those which
>> were
>>>     left out?
>> Once you have your GO term of interest, this is essentially the same
>> question (2), no?
>>
>> HTH,
>> -steve
>>
>> --
>> Steve Lianoglou
>> Graduate Student: Computational Systems Biology
>>   | Memorial Sloan-Kettering Cancer Center
>>   | Weill Medical College of Cornell University
>> Contact Info: http://cbio.mskcc.org/~lianos/contact
>>
>
>

-- 
James W. MacDonald, M.S.
Biostatistician
University of Washington
Environmental and Occupational Health Sciences
4225 Roosevelt Way NE, # 100
Seattle WA 98105-6099