[Bioperl-l] Re: [SO-devel] Re: GFF3 preliminary

Richard Durbin rd at sanger.ac.uk
Wed Mar 5 12:57:44 EST 2003


Ah.  I did say what I was doing when I emailed the list.  Michael made 
SOFA.ontology from SO.ontology using grep and a bit of editing.  I very 
strongly feel that SOFA should be made from SO this way, i.e. you should 
not edit SOFA.ontology directly, but just the SOFA parts of SO. 
Otherwise they will never be in synch.  Sorry that I failed to write the 
perl script I promised to automatically derive SOFA from SO.

Richard

PS If you are all working on this together, then my highest priority 
plea is not to elevate start_codon and stop_codon to significant status. 
  If you want to use endpoints then use interbase junctions 
(coding_start and coding_end).  In fact I have wobbled backwards in the 
great debate, and would not support Jim Kent and Lincoln to stick with 
cds or (perhaps a better name) coding_exon regions joined together by a 
common parent, as has been most widely used in earlier GFF and GTF.

Suzanna Lewis wrote:
> Hi Richard,
> 
> Looks like you committed your changes to so.ontology and I
> committed changes to sofa.ontology. Michael, Chris, and I
> are here in Chicago together and we'll try to work on
> resolving the Singapore changes to sofa back into so proper.
> Not sure where the differences will be, but it will be
> interesting I'm sure. Chris has added some definitions
> as well for sofa which should help in revealing what our
> thinking was.
> 
> -S
> 
> 
> On Friday, February 21, 2003, at 04:21 AM, Richard Durbin wrote:
> 
>> I have an edited version of SO which I will check back in shortly if 
>> it does not clash with too many other SO changes since Monday, in 
>> which the
>> SOFA:SOFA tags that Michael and I put in last week are replaced by 
>> either SOFA:junction or SOFA:region.  This makes explicit which SOFA 
>> terms are junctions and which are regions.
>>
>> Note that SO/SOFA themselves don't mandate whether you use base 
>> coordinates (as in GFF) or interbase coordinates (as perhaps in some 
>> wizzy new format from Berkeley).  i.e. SO/SOFA are not formats, but 
>> ontologies of terms, and you can and should use them with other 
>> formats than GFF.  The distinction between regions and junctions is 
>> real however, and will have to be reflected in any format.  Some 
>> things inherently have extent covering bases (regions) and some things 
>> are boundaries between neighbouring bases (junctions).
>>
>> Richard
>>
>> Chris Mungall wrote:
>>
>>> On Fri, 21 Feb 2003, Richard Durbin wrote:
>>>
>>>> length
>>>> just before the start of the ATG and after the stop codon.  There are SO
>>>> terms for the start_codon and stop_codon but you really don't want to
>>>> use them to define the start and end.  e.g. Michael's 4 base start
>>>> codons, but also many start and stop codons are broken by introns.
>>>> There is no reason not to use junctions.
>>>
>>> Ah, ok - spaced on the distinction between codon/junction there
>>> I'm still not comfortable with the hack to get round not having interbase
>>> coordinates, where certain subgraphs of SOFA are treated as flanking 
>>> bases
>>> and the rest is treated as inclusive bases. I guess it's necessary 
>>> for GFF
>>> though.
>>> should we make this explicit in the names; eg all types for features have
>>> flanking-base coordinates have the suffix '_junction'?
>>>
>>>> Richard
>>>>
>>>>
>>>>
>>>>
>>
>>
>>
>>
>> -------------------------------------------------------
>> This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
>> The most comprehensive and flexible code editor you can use.
>> Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
>> www.slickedit.com/sourceforge
>> _______________________________________________
>> SOng-devel mailing list
>> SOng-devel at lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/song-devel
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: Etnus, makers of TotalView, The 
> debugger for complex code. Debugging C/C++ programs can leave you 
> feeling lost and disoriented. TotalView can help you find your way. 
> Available on major UNIX and Linux platforms. Try it free. www.etnus.com
> _______________________________________________
> SOng-devel mailing list
> SOng-devel at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/song-devel
> 




More information about the Bioperl-l mailing list