[DAS] DASSequence.java behaves somehow strange

Thorsten Jansen thorsten.jansen@lionbioscience.com
Thu, 28 Nov 2002 16:30:02 +0100


Tried it with the build from today and it seems to work.
The dasserver version is: DazzleServer/0.97 (20020924; BioJava pre-1.3)

Thorsten

-----Original Message-----
From: Thomas Down [mailto:td2@sanger.ac.uk]
Sent: Thursday, November 28, 2002 3:00 PM
To: Thomas Down
Cc: Thorsten Jansen; das@biodas.org; Greg Cox
Subject: Re: [DAS] DASSequence.java behaves somehow strange


On Thu, Nov 28, 2002 at 01:03:42PM +0000, Thomas Down wrote:
> On Thu, Nov 28, 2002 at 12:08:06PM +0100, Thorsten Jansen wrote:
> > Hi,
> > testing my dasclient classes I recognized that the actual build from
biojava
> > and the build from 18.10.2002 show different behaviour in loading the
> > sequence string using the DASSequence class (used method:
> > dasSeq.subList(startPos,stopPos).seqString()). Using the ensembl (homo
> > sapiens 8.30) data as reference server the following happens:
> >
> > Using new build: Loading the sequence for chromosome 1 from position
30001
> > to 40000 results in a "normal" sequence string from 30001 to 33556. From
> > 33557 to 40000 only "N"s are listed. This is somehow related to the
shown
> > features "AC114498 (1 to 33556)" and "AL669831 (33557 to 212222)".
>
> Okay, this is definitely wrong.  The reason why the first
> clone worth of sequence is okay but the second fails is that
> AL669831 is longer than a particular threshold (100000) which
> causes its seqeunce to be fetched in tiles rather than in a single
> request.  Something must have broken in the code which stitches
> these tiles back together.  Looking now...

Actually, I can't reproduce this with the current CVS code.  I'm
not sure what's changed since October, but it doesn't hurt to try
the current version.

Let me know if you're still seeing the problem.  It would also
be helpful to know about the version of the server which you're
using.

     Thomas.