[DAS2] transitioning from DAS/1 to DAS/2

[Andrew Dalke]

> Andreas Prlic wrote:
>> I would therefore like to discuss to rename the "genome" domain to
>> "sequence".
>>
>> The information of which type of sequence, "genomic" or "protein
>> sequence"  should be provided via the
>> source description.

Steve:
> I like this idea. The only nucleotide specific stuff in the DAS/2 
> retrieval
> spec is the region request. Strand designation in a location specifier 
> is
> already optional.
>
> We'd may then want to change the 'sequence' request to something else,
> perhaps 'residues'?

I've been thinking the same thing.

Going one step further - what about dropping the name entirely?

Consider this, with some of the xml: attributes removed for clarity.
I've added a 'source_type' field in the <SOURCE> element.

   This is what you get from a SOURCES request

HTTP GET http://www.example.com/das2/

    (note lack of 'genome' in that URL)

<SOURCES>
   <SOURCE id="volvox" description="Volvox Example Database"
           taxon="http://www.ncbi.nlm.nih.gov/taxon-browser?id=29118"
           source_type="genome">
      <VERSION id="volvox/1" description="Build 1, October 2002">
        <ASSEMBLY id="http://www.ensembl.org/das/genome/vv116" />
      </VERSION>
   </SOURCE>

   <SOURCE id="yeast" description="S. cerevisae Genome via SGD"
           taxon="http://www.ncbi.nlm.nih.gov/taxon-browser?id=2186"
           source_type="genome">
      <VERSION id="yeast/1Feb2001" description="February 2001 build" />
      <VERSION id="yeast/1Mar2001" description="March 2001 build (minor 
changes)" />
   </SOURCE>


   <SOURCE id="swissprot" description="SWISS-PROT sequence"
           source_type="protein">
      <VERSION id="swissprot/45" description="Some SProt version" />
   </SOURCE>

   <SOURCE id="pdb" description="PDB"
           source_type="structure">
      <VERSION id="pdb/2005q3" description="Fall 2005 release of all PDB 
structures" />
   </SOURCE>

</SOURCES>

That is, the SOURCES request returns information about genomic,
protein sequence and structure databases.

If this occurs then there will need to be a few changes to the spec.
For example, 'taxon' is probably only properly part of the genomic
sources and not in the others, so perhaps move the taxon information
into a subelement of those SOURCE elements with 'source_type' == 
'genome'.


   <SOURCE id="volvox" description="Volvox Example Database"
           source_type="genome">
      <TAXON>http://www.ncbi.nlm.nih.gov/taxon-browser?id=29118</TAXON>

      <VERSION id="volvox/1" description="Build 1, October 2002">
        <ASSEMBLY id="http://www.ensembl.org/das/genome/vv116" />
      </VERSION>
   </SOURCE>
    ...

					Andrew
					dalke at dalkescientific.com