[Genome] Gene ontology and the genome browser

[Ann Zweig]

Hello Avi,

	There is not a ready-made table that includes hg18 (and mm8) known gene names 
and go annotations.  But there is a 'go' database which has a relationship with 
the hg18 (and mm8) database.  Since you are working with large tables, I would 
suggest that you download the 'go' tables that you will need and then write some 
SQL joins to get the information you want.

	You can download tables from the 'go' database at our ftp server:
ftp://hgdownload.cse.ucsc.edu/goldenPath/go/database/

	In this directory, the files are structured like so:

   - *.sql files: the MySQL commands used to create the tables.
   - *.txt.gz files: the database tables in a tab-delimited format
     compressed with gzip.

	Tables of particular interest to you might be the goaPart table and the term 
table.  The tables in the go database are related to the knownGene table like so:

go.goaPart.dbObjectSymbol (via hg18.knownGene.proteinID)

go.term.acc (via go.goaPart.goId)

go.goaPart.goId (via go.term.acc)
go.graph_path.term1_id (via go.term.id)
go.graph_path.term2_id (via go.term.id)
go.term2term.term1_id (via go.term.id)
go.term2term.term2_id (via go.term.id)
go.term_dbxref.term_id (via go.term.id)
go.term_definition.term_id (via go.term.id)
go.term_synonym.term_id (via go.term.id)

	If you are interested in the table structure and other table relationships, 
please visit our table browser (press the "Tables" link in the blue navigation 
bar across the top of the browser).

	This should get you started in your quest.  If you need more detailed help, 
please feel free to write back.


Regards,

----------
Ann Zweig
UCSC Genome Bioinformatics Group
http://genome.ucsc.edu



fungazid fungazid wrote:
> Dear all,
>    
>   Hello to you, 
>    
>   I downloaded knownGene mysql-tables of mm8 and hg18, and I want to attach each gene name (‘name’ column) to its gene ontology annotations (Go). I would like to know it there is any simple way to achieve this task either in the UCSC site or in other sites. If there is a ready-made table somewhere, it is also a good idea. If you think it is better to use othe gene table than knownGene, I can listen.
>    
>   The table should be something like 
>    
>   -----------------------------------------------------------------------------------
>   fields:
>   Identifier ,Symbol ,Gene name,Category of term , GO ID, GO Term, Source Evidence, PubMed
>    
>    example - content of a record:
>    
>   NM_001005484          
>   OR4F5            
>   Olfactory receptor, family 4, subfamily F, member 5      
>   molecular_function            
>   GO:0004872    
>   receptor activity                        
>   IEA      
>   -
>    
>   ------------------------------------------------------------------
>    
>    
>   I tried genetools software (http://www.genetools.microarray.ntnu.no/egon/index.php), which is designed to do exactly that, but I believe I need to verify what I got.
>    
>    
>   Thanks
>   Avi
>      
>    
>    
> 
> 
>  __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around 
> http://mail.yahoo.com 
> _______________________________________________
> Genome maillist  -  Genome at soe.ucsc.edu
> http://www.soe.ucsc.edu/mailman/listinfo/genome