Convert GO numbers to GO terms

[Ramet]

Hi all!

Which programme is most suitable for converting a list of GO numbers to their corresponding GO terms (including assignment to MF, CC, BP and lvl)?

I have something like:
73562
01276
89003
00271
...
...
...

And I need an output like
73562 --> MF: lvl3: oxygen transporter activity
01276 --> CC: lvl2: mitochondrial membrane
89003 --> CC: lvl3: inner mitochondrial membrane
00271 --> BP: lvl4: oxygen detoxification
...
...
...

Thanks for hints!

[svl]

Hi Ramet,

At this link you can find txt files with the info you need, although I'm not sure what the "lvl" is...

http://www.geneontology.org/GO.downloads.files.shtml

You can write a small perl script to parse and combine this with your own file.

-Stef

[robertorun]

I think, the key is the caculation of level!

[malachig]

Another option is to create a local instance of the MySQL database and use an SQL query to map ids to descriptions. Instructions for doing this are here

http://www.geneontology.org/GO.database.shtml#build-your-own

[malachig]

Yes calculating the level (or 'depth') of a term is the main challenge to obtain the information 'Ramet' seeks. A large number of MySQL queries of the GO schema are discussed at the following URL. Included on this page are various mentions of the difficulty and caveats associated with depth. Each term can have multiple associations to other terms. This means that one term may have more than one depth depending on the path you take through the tree. You are therefore faced with the problem of deciding what depth to use (average, minimum, maximum?). This all leads to the questions: Do you need the level? Does it really mean what you hope it does?

From the GO documentation: "The depth of a node or distance between nodes is not necessarily biologically meaningful. Other more advanced queries are required to get biologically meaningful information-theory based distances between terms "

http://wiki.geneontology.org/index.php/Example_Queries#Find_terms_by_GO_ID

Good luck!

[simonandrews]

There's quite a nice Perl API to the GO databases at:

http://search.cpan.org/~cmungall/go-perl/go-perl.pod

The example code at the top pretty much does what you're after.

[Ramet]

Thanks to everyone!

You have convinced me that I maybe do not need the level (depth). Thanks for that conceptual help!

And parsing the textfiles from http://www.geneontology.org/GO.downloads.files.shtml with a perl script seems to be the solution for my problem.
Shame on me to have missed them.

I'll later on will have a look on the MySQL database solution and then decide which way to go.

Again, thank you very much!