Hello, Anton,

I don't know why the Table Browser output is tossing
out the relationship between the individual items and
their go.goaPart.aspect values.  In the database they
are separate:

mysql> SELECT * FROM goaPart WHERE dbObjectSymbol = "TAB2_RAT";
+------------+----------------+-------+------------+--------+
| dbObjectId | dbObjectSymbol | notId | goId       | aspect |
+------------+----------------+-------+------------+--------+
| Q5U303     | TAB2_RAT       |       | GO:0005622 | C      |
| Q5U303     | TAB2_RAT       |       | GO:0005634 | C      |
| Q5U303     | TAB2_RAT       |       | GO:0005737 | C      |
| Q5U303     | TAB2_RAT       |       | GO:0008270 | F      |
| Q5U303     | TAB2_RAT       |       | GO:0046872 | F      |
+------------+----------------+-------+------------+--------+

I see two ways to solve your problem.  If you are technically
capable of querying the database directly, you could use the
public mysql server, genome-mysql.  See this page for details:

http://genome.ucsc.edu/FAQ/FAQdownloads.html#download29

Otherwise, you could apply a filter on the query and repeat it
for each of the three aspect types (F,P,C).  To do that:

filter [create]

Then you would follow the same table-intersection logic as
before, to access the goaPart table via the relationship,

rn4.rgdGene2 -> rn4.rgdGene2ToUniProt -> go.goaPart

then set:

aspect does match C (etc, in separate queries)

this will give you separate output for each of the three
aspect types.

regards,

                        --b0b kuhn      
                        ucsc genome bioinformatics group


On 4/10/2012 9:11 PM, Anton Kratz wrote:
> Hi Vanessa,
> 
> thank you very much, this is almost what I was looking for, I have one
> follow-up question though. I made a table following your instructions with
> these fields:
> 
> #rn4.rgdGene2.name    go.goaPart.dbObjectId    go.goaPart.dbObjectSymbol
> go.goaPart.notId    go.goaPart.goId    go.goaPart.aspect
> 
> Here is one example entry, for RGD:1309527 (Map3k7ip2):
> 
> RGD:1309527    Q5U303,    TAB2_RAT,    ,
> GO:0005622,GO:0005634,GO:0005737,GO:0008270,GO:0046872,    C,F,
> 
> RGD:1309527 has five GO terms, two of them are Molecular Function
> (GO:0008270,GO:0046872) and three of them are Cellular Component
> (GO:0005622,GO:0005634,GO:0005737). But, from the table which I retrieve I
> do not see how one can parse this information, because it just says "C,F"
> it is unclear which GO terms are C and which ones are F. Shouldn't it be
> something like "C,C,C,F,F"? Or maybe two tab-separated blocks splitting the
> GO terms into C and F?
> 
> I would like to ask how to get a table with all RGD genes (rn4 assembly)
> and the associated Gene Ontology terms, in tab-separated ASCII, such that
> it is still clear which GO term belongs to which of the three categories.
> 
> Thank you very much,
> Anton
> 
> 
> On Tue, Apr 10, 2012 at 5:48 AM, Vanessa Kirkup Swing
> <[email protected]>wrote:
> 
>> Hi Anton,
>>
>> You will want to select the "rgdGene2ToUniProt" table. This will then give
>> you the option at the top of the linked tables list to select "go.goaPart"
>> which will then give you the option to select the individual fields from
>> go.goaPart.
>>
>> If you have further questions or comments, please email the list:
>> [email protected].
>>
>> Vanessa Kirkup Swing
>> UCSC Genome Bioinformatics Group
>>
>>
>> ---------- Forwarded message ----------
>> From: Anton Kratz <[email protected]>
>> Date: Thu, Apr 5, 2012 at 10:26 PM
>> Subject: [Genome] RGD Genes and Gene Ontology for rn4 assembly via Table
>> Browser
>> To: UCSC Genome Browser Mailing List <[email protected]>
>>
>>
>> Dear UCSC team,
>>
>> I would like to get a table with all RGD genes and the associated Gene
>> Ontology terms, in tab-separated ASCII, for the rn4 assembly.
>>
>> I tried using the Table Browser for that, my idea was to start with the
>> tack "RGD Genes" and the table "rgdGene2", then use "selected fields from
>> primary and related tables" to find linked tables which will lead me to the
>> GO terms. However, without success.
>>
>> Could you please explain how to get such a table?
>>
>> Thank you!
>>
>> Anton
>> _______________________________________________
>> Genome maillist  -  [email protected]
>> https://lists.soe.ucsc.edu/mailman/listinfo/genome
>>
>>
> _______________________________________________
> Genome maillist  -  [email protected]
> https://lists.soe.ucsc.edu/mailman/listinfo/genome
_______________________________________________
Genome maillist  -  [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome

Reply via email to