[Gusdev-gusdev] Re: why do we need GeneInstanceCategory and friends?

[Steve F. <sfi...@pc...>]

ok, i've had a better look at the Wiki in which the bootstrap data is 
described
http://www.gusdb.org/wiki/index.php/Bootstrap%20data%20II

It seems to me that the design outlined there is project/application 
specific.

I also looked at the schema, and saw that gene_instance_category_id is 
not nullable.

to me that explains why we are attempting to include the categories as 
bootstrap data.

Instead, i think it makes sense:
  - to make that field nullable
  - for any plugin that wants to fill that stuff in to enforce itself 
the values it expects there

Comments?

steve

Steve Fischer wrote:

> folks-
>
> mike and i are moving ahead putting together the new GUS 3.5 install 
> procedure.  It will include pipelines to introduce bootstrap data.
>
> Some of the data that GUS seems to need are values for:
>  DoTS.GeneInstanceCategory
>  DoTS.RNAInstanceCategory
>  DoTS.ProteinInstanceCatetory
>
> I am wondering why we need these categories.  It seems to me that if 
> the category would be implicit in the feature and/or sequence that the 
> instance is linking to.
>
> here are the suspicious values we have in CBIL's GUS for these tables:
>
> SQL> select name from Dots.RNAINSTANCEcategory;
> unknown
> mRNA
> assembly
>
> SQL> select name from Dots.proteininstancecategory;
> inferred from Assembly
> GenBank transcript
>
> SQL> select name from Dots.geneinstancecategory;
> gDG
> unknown
>
>