[DAS2] DAS1 TYPE attribute "category" is what in DAS2?

[Andrew Dalke]

Several questions listed here.

DAS1 TYPE elements had an attribute "category".  Here are some
of the categories listed in DAS1 servers

     ALPHA-BETA-MOTIF, ASX-MOTIF, ASX-TURN, BETA-BULGE, BETA-BULGE-LOOP,
     BETA-TURN, CATMAT-3, CATMAT-4, GAMMA-TURN, HELIX-L, NEST,
     SCHELLMANN-LOOP, ST-MOTIF, ST-STAPLE, ST-TURN, enzyme, miscellaneous
     motif, pathway, rRNA, repeat, rfam, structural, tRNA
     transcription, transmembrane prediction


DAS1 says:
    category (optional, recommended) attribute, which provides
    functional grouping to related types.


*TOPIC*: What should I do for automated conversion in my proxy system?

Currently I have:
   DAS1 "id" used to make DAS2 "uri" (via url encoding)

   DAS1 "method" copied into DAS2 "method"

   DAS1 (non-standard extension) "description" copied into DAS1 
"description"

   DAS1 (non-standard extensions) "ontology" and "evidence" used to fake
       a DAS2 "ontology" uri

*Q1*: If there is a DAS1 "category" should I use it to make a DAS2 
"title"?

Gregg's viewer merges types into a single track based on the title, so
I that feels correct to me.

*Q2*: If the title is not given, should I use the DAS1 "id" as the DAS2 
"title"?
I think that is correct.

*Q3*: If there's no DAS1 "description" extension to use for DAS2's 
"description"
should I copy DAS1's "title" instead (which in turn might come from the
"category" and/or the "id" fields).  My feeling is no, that is not 
appropriate.

*Q4*:  I fake an ontology if I can.  Does anyone know examples of
DAS1 extensions with to support ontologies other than TMHMM, which has

<TYPE id="INSIDE" method="TMHMM-2.0" category="transmembrane prediction"
  description="Prediction of transmembrane helices in proteins, here
         intracellular part of protein"
   evidence="IEA" 
ontology="http://www.geneontology.org/GO.evidence.shtml">
  1766831
</TYPE>

For now I convert that into
     http://www.geneontology.org/GO.evidence.shtml#IEA



					Andrew
					dalke at dalkescientific.com