[Bioperl-l] bp_genbank2gff problems

[Lincoln Stein]

I've just confirmed that you do *not* need to split the GFF3 files 
into the annotation and DNA parts in order to use the current 
generation of Bio::DB::GFF loaders.  Have fun!

Lincoln

On Monday 15 December 2003 12:29 pm, Neil Saunders wrote:
> I'm having a frustating time with the bp_genbank2gff.pl script.
>
> I have 2 systems:
>
> (1) Debian sid, perl 5.8.2, latest CVS bioperl-live, bioperl-run
> and Gbrowse.
> (2) Debian woody, perl 5.6.1, CVS versions as above.
>
> I have written a script that takes a set of GenBank files and pipes
> them through various processes to generate GFF, Fasta and conf
> files for use with Gbrowse.  On system (1) above, I use:
>
> bp_genbank2gff.pl -file <file> -stdout
>
> No problems, GFF3 file output appears.
>
> On system (2) above, the same command gives errors of the type:
>
> ------------- EXCEPTION  -------------
> MSG: Can't connect to database: Access denied for user:
> '@localhost' to database 'test'
> STACK Bio::DB::GFF::Adaptor::dbi::caching_handle::new
> /usr/local/share/perl/5.6.1/Bio/DB/GFF/Adaptor/dbi/caching_handle.p
>m:89 STACK Bio::DB::GFF::Adaptor::dbi::new
> /usr/local/share/perl/5.6.1/Bio/DB/GFF/Adaptor/dbi.pm:93
> STACK Bio::DB::GFF::Adaptor::dbi::mysql::new
> /usr/local/share/perl/5.6.1/Bio/DB/GFF/Adaptor/dbi/mysql.pm:270
> STACK Bio::DB::GFF::Adaptor::biofetch::new
> /usr/local/share/perl/5.6.1/Bio/DB/GFF/Adaptor/biofetch.pm:95
> STACK Bio::DB::GFF::new
> /usr/local/share/perl/5.6.1/Bio/DB/GFF.pm:599 STACK toplevel
> /home/neil/gbrowse/genomes/scripts/bp_genbank2gff.pl:218
>
> --------------------------------------
> DBI->connect(test) failed: Access denied for user: '@localhost' to
> database 'test' at
> /usr/local/share/perl/5.6.1/Bio/DB/GFF/Adaptor/dbi/caching_handle.p
>m line 139
>
>
> Clearly the bp_genbank2gff script is trying to access a database
> 'test' on 'localhost' with no user.  I guess my question is: why? 
> I have told it to send to stdout.  As I have the same bioperl
> version on both machines, I'm pretty confused.  The only thing I
> noticed from 'make test' on the woody machine was this failure:
>
> t/GFF.t       255 65280    32   24  75.00%  21-32
>
> Relevant?
>
> thanks for any pointers,
>
> Neil