[BioSQL-l] Relationship between Gene CDS and mRNA Seqfeatures

[Peter]

On Wed, Nov 3, 2010 at 3:49 PM, Nicholas Thrower <throwern at msu.edu> wrote:
>
> Hello all,
>
> I¹m new to this list so I hope this question hasn¹t been asked before. I
> couldn¹t find an archive to search.
>
> I have a biosql schema in oracle loaded with several genbank entries
> including TAIR9. I would like some help determining the relationship between
> mrna and cds features that belong to a single Gene entry.
>
> My seqfeature table has the following entries.
>
> OID         RANK    DISPLAY_NAME    ENT_OID     TYPE_TRM_OID
> SOURCE_TRM_OID
> 666269  22747   Gene                       643516        626447
> 626431
> 666270  22748   Mrna                       643516        626451
> 626431
> 666271  22749   Mrna                       643516        626451
> 626431
> 666272  22750   Cds                          643516        626455
> 626431
> 666273  22751   Cds                          643516        626455
> 626431
>
> I¹m trying to correlate CDS to mRNA so that I can display them. All of the
> seqfeatures have the same exact locus_tag qualifier and I can¹t find any
> other qualifiers noting this relationship. Do I need to make an assumption
> based on seqfeature id, or rank in order to determine the relationship
> between mRNA and CDS.
>
> Here is an example from TAIR of what this data could look like.
> http://arabidopsis.org/servlets/TairObject?id=134549&type=locus
>
> How do I know which CDS and mRNA are paired together?
>
> -Nick

In a GenBank file this relationship between CDS and mRNA or gene
features is implicit from the order in the file. I think the rank is used in
BioSQL to record the feature order, so you'd have to look at that.
(This is assuming the Bio* wrapper you used to load the GenBank
file into BioSQL isn't making any inferences like this for you).

P.S. Which Bio* wrapper are you using? BioPerl?

Peter