[Bioperl-l] SeqFeature::Generic 'frame' method

Jason Stajich jason@chg.mc.duke.edu
Mon, 26 Feb 2001 18:32:20 -0500 (EST)

That does seem safer.  I can easily transform from 1,2,3 to 0,1,2
so frame +1 in blast will be equivalent to frame 0 in GFF. But what about
cases where we don't have a frame, is Frame 0 technically correct there?

Strand information for a HSP currently comes from direction
of start/end points on the query seq.  Is it concievable that they
wouldn't match?  i.e. will a match in -3 frame always correspond to a
match on rev strand of query....  This might be a dilhema...  I already
checked in code which breaks because of invalid frames, I will fix so it
is at least transformed to valid number space of gff.  Strand will get
ignored for now.


On Mon, 26 Feb 2001, Hilmar Lapp wrote:

> Jason Stajich wrote:
> > 
> > Currently we only support frames of 0,1,2 as part of gff support I'm sure.
> > However, as I try and map a TBLASTN HSP to a SeqFeature, frames
> > can be -3,-2,-1,0,1,2,3.  Does anyone see problems with extending the
> > supported values in the frame method to handle those above.  Does this
> > break GFF standards somehow?  I can imagine someone parsing a blast
> > report and writing the HSPs out as GFF features so it is quite likely that
> > the two cases shall meet.
> > 
> Hmm. As long as there's a clear context of an HSP being the result of
> parsing a Blast report there's no problem with extending the set of
> allowed values. However, as soon as this context gets lost, like you
> said (and going through GFF is not the only way to lose that), there's
> a problem, because then frame 1 and frame 1 no longer mean the same
> thing. Bad.
> Can we map TBLASTN frame to the definitions we have set forth in
> SeqFeatureI? E.g. TBLASTN parsed frame -3 would become SeqFeatureI
> strand -1, frame 2. TBLASTN parsed frame 1 would be mapped to strand
> 1, frame 0, and so forth. This would keep both the definitions and the
> TBLASTN information.
> The only remaining problem is TBLASTN frame 0. What does it mean? If
> it's no frame, then it's equivalent to SeqFeatureI frame 0.
> What do you think?
> 	Hilmar
> -- 
> -------------------------------------------------------------
> Hilmar Lapp                            email: lapp@gnf.org
> GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
> -------------------------------------------------------------

Jason Stajich
Center for Human Genetics
Duke University Medical Center