Re: [Mged-mage] How to tell SeqFeature type?

[Valerie W. <val...@sr...>]

Fair enough--thanks for your assistance!

cheers,
Valerie

Miller, Michael D (Rosetta) wrote:
> Hi Valerie,
> 
> it is intended for a higher level type (from the documentation):
> 
> "type: The type of biosequence, i.e. gene, exon, UniGene cluster, 
>     fragment, BAC, EST, etc."
> 
>> Are you saying that this BioSequence Type association doesn't 
>> actually 
>> tell us the types of the SeqFeatures because people can use 
>> Descriptions 
>> or identifiers to determine the type?  (And in that case, 
>> what is meant 
>> by the BioSequence Type association?)
> 
> I really wasn't saying anything more than I was saying.
> 
>> So 
>> for arbitrary input, it would be hard to parse an arbitrary 
>> description 
>> with any certainty of what might be in it.
> 
> I've seen very few applications use SeqFeature--as an importer/consumer
> without knowledge of the source, one can only provide a place to hold
> what's in the MAGE elements/objects.  How does an importer actually know
> much about the BioSequence itself?
> 
> My suggestion was (apparently wrongly) more geared towards an exporter
> or in-house application.
> 
> cheers,
> Michael
> 
>> -----Original Message-----
>> From: Valerie Wagner [mailto:val...@sr...] 
>> Sent: Friday, September 29, 2006 1:51 PM
>> To: Miller, Michael D (Rosetta)
>> Cc: Trish Whetzel; mge...@li...
>> Subject: Re: [Mged-mage] How to tell SeqFeature type?
>>
>>
>> I'm writing a loader for the BioWarehouse that takes MAGE-ML 
>> input.  So 
>> for arbitrary input, it would be hard to parse an arbitrary 
>> description 
>> with any certainty of what might be in it.  If it's the case that the 
>> BioSequence Type association determines the (single) type of 
>> SeqFeatures, then that's at least something I can parse.
>>
>> Are you saying that this BioSequence Type association doesn't 
>> actually 
>> tell us the types of the SeqFeatures because people can use 
>> Descriptions 
>> or identifiers to determine the type?  (And in that case, 
>> what is meant 
>> by the BioSequence Type association?)
>>
>> thanks,
>> Valerie
>>
>> Miller, Michael D (Rosetta) wrote:
>>> Hi Trish and Valerie,
>>>
>>> SeqFeature is Describable so one can add a Description to 
>> it with text =
>>> "FeatureType" and a URI to the appropriate documentation for that
>>> particular Feature then add the OntologyEntry using a SO term.
>>>
>>> But...in truth, when we designed MAGEv1 (and have pretty 
>> much carried
>>> this into FuGE and MAGEv2), to truly model annotating a Sequence was
>>> decided to be out of scope.  Many users of MAGE form a pseudo-LSID
>>> identifier for their genes based on GenBank, RefSeq, 
>> Entrez, whatever,
>>> accession/identifiers then use that sequence database as the
>>> Database_ref. Later some other tool/software can come back 
>> and annotate
>>> the genes.
>>>
>>> But to add limited annotation, I think the above would work 
>> reasonably
>>> well.
>>>
>>> cheers,
>>> Michael
>>>
>>>> -----Original Message-----
>>>> From: mge...@li... 
>>>> [mailto:mge...@li...] On Behalf Of 
>>>> Valerie Wagner
>>>> Sent: Friday, September 29, 2006 12:41 PM
>>>> To: Trish Whetzel
>>>> Cc: mge...@li...
>>>> Subject: Re: [Mged-mage] How to tell SeqFeature type?
>>>>
>>>>
>>>> Thanks, Trish, that's exactly what I was asking.
>>>>
>>>> Forgive my ignorance in this area, but can't a BioSequence 
>> have many 
>>>> different features associated with it?  So are you saying that the 
>>>> BioSequence in question is only meant to have one type of 
>> SeqFeature 
>>>> associated with it?
>>>>
>>>> Valerie
>>>>
>>>> Trish Whetzel wrote:
>>>>> Hi Valerie,
>>>>>
>>>>> If I understand the question correctly, the MAGE class 
>>>> SeqFeature has an 
>>>>> association to BioSequence which has 4 associations to 
>>>> OntologyEntry, one 
>>>>> of which is called Type. The values for this were initially 
>>>> derived from 
>>>>> the MGED Ontology (MO) BioSequenceType subclasses, but were 
>>>>> deprecated in favor of using terms from the Sequence 
>>>> Ontology (SO). The 
>>>>> mapping of MO to SO terms can be found at:
>>>>> http://song.sourceforge.net/so_mappings.shtml
>>>>>
>>>>> Trish
>>>>>
>>>>>
>>>>>> How does one discern what type of feature a MAGE 
>>>> SeqFeature is (gene,
>>>>>> binding site, etc.)?
>>>> --------------------------------------------------------------
>>>> -----------
>>>>> Take Surveys. Earn Cash. Influence the Future of IT
>>>>> Join SourceForge.net's Techsay panel and you'll get the 
>>>> chance to share your
>>>>> opinions on IT & business topics through brief surveys -- 
>>>> and earn cash
>>>> http://www.techsay.com/default.php?page=join.php&p=sourceforge
>>> &CID=DEVDEV
>>>> _______________________________________________
>>>> Mged-mage mailing list
>>>> Mge...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/mged-mage
>>>
>>>
>> --------------------------------------------------------------
>> ----------
>>> -
>>> Take Surveys. Earn Cash. Influence the Future of IT
>>> Join SourceForge.net's Techsay panel and you'll get the 
>> chance to share
>>> your
>>> opinions on IT & business topics through brief surveys -- 
>> and earn cash
>> http://www.techsay.com/default.php?page=join.php&p=sourceforge
> &CID=DEVDE
>> V
>> _______________________________________________
>> Mged-mage mailing list
>> Mge...@li...
>> https://lists.sourceforge.net/lists/listinfo/mged-mage
>>
>>
>>
>>
> ------------------------------------------------------------------------
> -
>> Take Surveys. Earn Cash. Influence the Future of IT
>> Join SourceForge.net's Techsay panel and you'll get the chance to
> share your
>> opinions on IT & business topics through brief surveys -- and earn
> cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDE
> V
>> _______________________________________________
>> Mged-mage mailing list
>> Mge...@li...
>> https://lists.sourceforge.net/lists/listinfo/mged-mage
> 
> 
>