[Biojava-l] Ensembl read problems

Thomas Down td2@sanger.ac.uk
Mon, 25 Nov 2002 23:32:25 +0000


On Mon, Nov 25, 2002 at 12:10:17PM +0100, saerts wrote:
> 
> Hello,
> I'm still experiencing problems when reading the EMBL formatted output of the
> "export data" of Ensembl. Both for human and mouse, although the parsing errors
> are different in both cases. I'm using a BioJava CVS-build of last week
> 
> In one case, no features can be parsed (human).
> In the other case, some of the locations are out of the source location (mouse).
> 
> I'm using the code and the sequences below, and the errors of both cases are
> pasted in this mail.
> 
> Does anyone have an idea why this goes wrong?

One of the Ensembl developers has been working today to eliminate
`out of range' features and other issues with the EMBL dumps  He's
been in contact with me, and I've seen some test files which do now
parse fine with BioJava.  Hopefully, the new dumping code will make
it to the main Ensembl site soon.

     Thomas.