repost - Re: [Bioperl-l] Huh? Bioperl Seq objects and strands

[Hilmar Lapp]

Mark Wilkinson wrote:
> 
> Just to re-post this question as it seems to be either (a) a bug, or (b) an
> unusual desire by the person who wrote the parser...
> 
> David Block wrote:
> 
> > Hi experienced bioperlers,
> >
> > We're trying to map a Bio::SeqI compliant object onto a Tk Map.  We're
> > very close (code within days for this part), but we're fighting with the
> > Seq object's definitions of gene, exon, and CDS.
> >
> > We loaded up a Bio::Seq object from GenBank using SeqIO.  Top-level
> > features for genes on the -1 strand have the tags
> > gene  strand 1
> > gene_span strand 0
> > intron strand -1
> > exon strand -1
> > etc...
> >
> > How can we tell which top-level features with strand 1 are actually strand
> > -1 without looking at sub-features?  We would like to be able to decide so
> > we can map the gene on the correct side of the axis.
> >
> > Also, why are introns and exons top-level features of a sequence, when
> > they are also (obviously) sub-features of a gene?
> >
> > Shouldn't it be (top-level features only):
> > gene strand -1
> > gene_span strand 0
> >
> > and that gene's subfeatures:
> > intron strand -1
> > exon strand -1
> >

Sorry for the delayed response, but it hasn't been forgotten (at least I
haven't).

Could you submit the sequence causing the mentioned misannotation. I have
to add that I will probably not have time to look at this in more detail
before the weekend. Maybe someone else wants to dig into it.

I should mention that the feature table is parsed on a line-by-line
basis, the features ending up as top-level features of the sequence
together with their tags. There is no back-log kept, so if the reverse
strand is not indicated immediately, there may be features for which the
strand is not switched accordingly. If this is the case, it's certainly a
bug.

Without further information, however, your problem is difficult to trace.


	Hilmar

-- 
-----------------------------------------------------------------
Hilmar Lapp                                email: hlapp@gmx.net
NFI Vienna, IFD/Bioinformatics             phone: +43 1 86634 631
A-1235 Vienna                                fax: +43 1 86634 727
-----------------------------------------------------------------