Re: [SO-devel] SOFA

[Chris M. <cj...@fr...>]

There is a description of the relations used in OBO ontologies here:
http://genomebiology.com/2005/6/5/R46

If I understand correctly, Payan, previously you were only using the  
directly asserted relations?

Previously there may have been

X part_of Y

now there may be

X is_a Z part_of Y

none of this should affect any software provided they have  
implemented the relation transitivity rules correctly.

Perhaps it would make things easier for people if we deployed a  
version of SO with all the implied links realized. This would save  
people writing their own rules (and possibly implementing them wrongly)

If you already have oboedit you can do this from the command line:

	obo2obo -o -saveallimpliedlinks so-with-implied-links.obo so.obo

As an aside, I'm wondering when it is necessary to traverse non-isa  
links on col9 of GFF. I thought this column was purely for sequence  
attributes, which do not stand in part_of relations? I may have  
misunderstood gff here.

On Apr 30, 2007, at 12:43 PM, Payan Canaran wrote:

> Hi Karen,
>
> Thanks for the info. Yes, this clarifies it. My question was in  
> relation
> to column 9 (ID and Parent relationships) rather than column 3. For  
> each
> term I walk up the tree towards the top using is_a, part_of and
> member_of relationships, identify parents (for ID-Parent  
> relationships)
> and also determine features that have a "is_a" relationship with the
> parents.
>
> Since, the only change is addition of "is_a" relationships from each
> node to the top-level node, I can simply disregard is_a relationships
> from top-level nodes when calculating parent-child pairs for use in
> column 9.
>
> Regards,
>
> Payan
>
> On Mon, 2007-04-30 at 11:11 -0600, Karen Eilbeck wrote:
>> Dear Payan,
>> I think what you are referring to is the the fact that between the
>> release 2.0 and 2.1, the ontology became 'is_a complete'.
>> Previous to this, terms like exon were indeed part of transcripts etc
>> but they had no is_a relationship to anything. The OBO ontologies
>> such as GO are mandated to be is_a complete as it is good ontological
>> practice and essential for reasoning and future interoperability.
>>
>> What is an exon?
>> Technically it is a region just like a transcript is a region. All we
>> did was fill in these relationships, and add in a couple of terms.
>> A region can be part of another region. Think of it as a substring.
>> I don't see how this complicates gff3 at all.
>> Everything in col 3 is a located_sequence_feature, so either a region
>> or a junction.
>>
>> Hope this helps,
>> --Karen
>> On Apr 30, 2007, at 9:53 AM, Canaran, Payan wrote:
>>
>>> Hi,
>>>
>>> I'm looking for any documentation on changes from SOFA release 2.0
>>> (revision 1.12) to SOFA release 2.1 (revision 1.14)?
>>>
>>> I believe additional terms ("transcript_region", "mRNA_region",
>>> etc.) that brigde between top-level terms such as "region" have
>>> been in this transition. One effect of this seems to be that now
>>> some terms have a recursive "is_a" relationship to region. For
>>> example, "exon" is_a "region" through "transcript_region" in
>>> release 2.1.
>>>
>>> This makes interpretation of part_of relationships somewhat tricky
>>> from a GFF3 perspective. For example, in release 2.0, exon is
>>> part_of a transcript, which is a member_of gene. In release 2.1,
>>> exon can still be a part_of (through recursive is_a and part_of
>>> relationships) of a gene. However, both exon and gene are "is_a"
>>> region through recursive "is_a" relationships. Since they are of
>>> the same type of feature, I'm not sure how they can be part of each
>>> other.
>>>
>>> I might be misinterpreting the relationships. I would very much
>>> appreciate any input or pointer to documentation that would help me
>>> understand this. Thank you.
>>>
>>> Payan
>>>
>>> -------------------------------------------------------------------- 
>>> --
>>> ---
>>> This SF.net email is sponsored by DB2 Express
>>> Download DB2 Express C - the FREE version of DB2 express and take
>>> control of your XML. No limits. Just data. Click to get it now.
>>> http://sourceforge.net/powerbar/db2/
>>> _______________________________________________
>>> SOng-devel mailing list
>>> SOn...@li...
>>> https://lists.sourceforge.net/lists/listinfo/song-devel
>>>
>>
>>
>> --------------------------------------------------------------------- 
>> ----
>> This SF.net email is sponsored by DB2 Express
>> Download DB2 Express C - the FREE version of DB2 express and take
>> control of your XML. No limits. Just data. Click to get it now.
>> http://sourceforge.net/powerbar/db2/
>> _______________________________________________
>> SOng-devel mailing list
>> SOn...@li...
>> https://lists.sourceforge.net/lists/listinfo/song-devel
>
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> SOng-devel mailing list
> SOn...@li...
> https://lists.sourceforge.net/lists/listinfo/song-devel
>