[Annotation] interpro scan question

[Harold Drabkin]

Doug howe wrote:
> I've got two protein sequences (found below) which are 99% identical.  
> Both belong to zebrafish synaptosomal protein snap25a, but one of them 
> is a fragment.  The fragment turns up positive for the interpro domain 
> IPR002197 (helix-turn-helix, Fis-type), while the full length protein 
> does not.  The fragment is the only protein on the gene that is 
> positive for this interpro domain.   It turns out that IPR002197 maps 
> through interpro2go to 'transcription factor activity' and 'regulation 
> of transcription'.  Is it possible that the protein fragment is 
> getting a false positive hit for the domain IPR002197?  If so, why 
> doesn't it also make a false positive hit on the full length proteins?
>
> Any help welcomed...
>
> O93578
> LGKFCGLCSCPCNKMKSGASKAWGNNQDGVVASQPARVVDEREQMAISGGFIRRVTDDAR
> ENEMDENLEQVGGIIGNLRHMALDMGNEIDTQNRQIDRIMEKADSNKTRIDEANQRATKM
> LGSG
>
> Q693L7
> MAEDSDMRNELADMQQRADQLADESLESTRRMLQLVEESKDAGIRTLVMLDEQGEQLERI
> EEGMDQINKDMKDAEKNLNDLGKFCGLCSCPCNKMKSGASKAWGNNQDGVVASQPARVVD
> EREQMAISGGFIRRVTDDARENEMDENLEQVGGIIGNLRHMALDMGNEIDTQNRQIDRIM
> EKADSNKTRFDEANQRATKMLGSG
>
>
why can I not find either of these ids in UniProtKB?

Could the fragment be a Trembl record as thus has not been fully 
annotated for domains (this is why we block display of domains 
associated with TreEMBL records in our db and only display and use the 
ones from a fully curated UniProt record.