[Gusdev-gusdev] GUS schema requirements

[Arnaud K. <ax...@sa...>]

Hi

While loading our whole set of data into the Feature-Land/Central Dogma 
part of GUS, we've got some issues with the items below. I would like to 
know what other groups using GUS think about them.

* Ontologies:
   => parsing the OBO format. As the OBO format has become the standard 
for ontologies, is there any plan to update the GO loading plugin to be 
able to parse this format ?
   => GO and sequence ontology are in different tables, Could we 
generalize the schema for handling ontologies, by designing a unique set 
of tables that would be used for loading any ontology. In future we will 
need extra ontologies, I'm thinking of the Evidence Code Ontology for 
example but that should apply to any ontology part of the OBO project 
(http://obo.sourceforge.net).

* Gene Naming nomenclature:

There was a thread a while ago about redesigning the naming nomenclature 
of genes in GUS, I don't think anything has been implemented yet. This 
is something important for us and it is not fully covered yet by GUS.

This is our requirements:

    * /systematic_id - final systematic name for when chromosome is
      finished or stable sequence is submitted, will be title for gene
      page in absence of a primary_name.
    * /temporary_systematic_id - for temporary systematic name used
      during projects where sequence is unfinished, i.e temporary name
      for the shotgun sequences.
    * /previous_systematic_id - for systematic names *no longer in use*.
      Can be seen on geneDB pages and queried. When systematic_id is
      assigned, the temporary_systematic_id can be assigned to this
      qualifier.
    * /synonym - used for other gene names still in use and to be
      displayed on the gene page
    * /obsolete_name - redundant gene names that can be queried but *are
      not visible on gene page* eg. errors
    * /primary_name - for published or agreed unique *user friendly gene
      name*, following the convention set out for kinetoplastids, will
      be the title for gene page. NB. This is *equivalent to* the
      EMBL-qualifier *standard_name* so it should be used "to give full
      gene name, but use /gene to give gene symbol". 
    * /reserved_name - pre-publication names that will, presumably,
      become the primary_name.


* Curation and comments:
We can't at the moment store in GUS our curation and note qualifiers. 
There is a DoTS.Comments table but how can we attach comments to feature 
objects ?

* Sres.BibliographicReference:
How can they be associated with sequence and feature objects ?

* Core.Algorithm table:
Is this table only for registering plugins ?
Can we use it as a controlled vocabulary table for storing software 
entries or it is not meant for this ?
We'd like to store software entries, for example a GeneFeature predicted 
by geneid v1.2 or a SignalPeptide Feature generated by SignalP v3.0

* Similarity data:
Can it be used for loading any similarity data, e.g. Fasta, BLAST, 
exonerate, Blat etc. I think it is possible but how can we store which 
software has been used (that refers to the former item).

* Curation of homology data:
Is it possible to specify that a orthologous or paralogous group has 
been manually curated (by adding a review_status_id attribute ?).

cheers
Arnaud