[BioSQL-l] BioSQL and GBrowse
Genevieve DeClerck
gad14 at cornell.edu
Mon Mar 13 21:24:30 UTC 2006
Hi,
I am trying to get GBrowse running with a biosql database but i'm having
problems.
Here's what i'm working with:
GBrowse 1.64
Mysql database with fresh biosql-schema
bioperl 1.5.1 (CVS bioperl-live)
Perl 5.8.1
Mac OS X 10.3.9
At the moment, I can load in a web browser GBrowse db's that don't have
a Biosql schema just fine. When I try to load my BioSQL database in
Gbrowse I get the following in the web browser (printed with the nice
blue, red, and black flavored GBrowse font/format):
----
An internal error has occurred
Could not open database.
Can't locate object method "new" via package "Bio::DB::Das::BioSQL" at
/Library/Perl/5.8.1//darwin-thread-multi-2level/Bio/Graphics/Browser/Util.pm
line 185.
----
And the following in my apache error log:
----
GBROWSE VERSION MISMATCH:
GBrowse version 1.64 requires a compatible version of the Bio::Graphics
library.
You should either install BioPerl (the CVS live version) or reinstall
GBrowse,
which will patch Bio::Graphics to the latest version.
gbrowse: trying to reload config, cache must be stale at
/Library/Perl/5.8.1//darwin-thread-multi-2level/Bio/Graphics/Browser/Util.pm
line 172.
Can't locate object method "new" via package "Bio::DB::Das::BioSQL" at
/Library/Perl/5.8.1//darwin-thread-multi-2level/Bio/Graphics/Browser/Util.pm
line 185.
----
From what I've read in the docs, bioperl 1.5.1 should be the proper
version to run with GBrowse 1.64. I also installed biosql and bioperl-db
fresh. What does the error "gbrowse: trying to reload config, cache must
be stale at" mean in relation to Util.pm? I've looked at the code in
Util.pm -- the error seems related to the database adaptor...
My gbrowse .conf file is identical to the 06.biosql.conf file except for
the first few lines (pasted below).
I used the load_seqdatabase.pl to load an abreviated version of
NC_004578.gbk (p. syrinage dc3000 from ncbi refseq) into the biosql db.
---
my test_biosq.conf
[GENERAL]
description = test_biosql
db_adaptor = Bio::DB::Das::BioSQL
db_args = driver mysql
dbname test_biosql
namespace genbank
version 1
host localhost
user nobody
pass ""
port 3306
---
Any ideas or comments appreciated!
Thanks,
Genevieve
Michael Cipriano wrote:
> Ok, I installed the CVS version rather then the release version and
> everything seems to be working now, so if anyone else gets this problem,
> that is the simple solution.
>
> -Michael
>
> On Tue, 2006-03-07 at 09:39 -0800, Michael Cipriano wrote:
>
>>Hello,
>>
>>I'm not sure if this is a GBrowse problem, or a GBrowse on BioSQL
>>problem, so tell me if I should post this to the GBrowse list.
>>
>>I have a BioSQL database with bacterial genome sequences loaded from
>>genbank files. I can browse around the database just fine, but when I
>>try and use gbrowse_img, the image is messed up.
>>
>>Take a look at these two links, they should show basicaly the same
>>thing, but the gbrowse_img link shows the features all piled up at the
>>left side of the image, and they are barely visible.
>>
>>
>>http://alien.lbl.gov/cgi-bin/gbrowse/bacteria/?start=1277180;stop=1278571;ref=NC_000913;width=800;version=100;label=CDS-Genes;
>>http://alien.lbl.gov/cgi-bin/gbrowse_img/bacteria?name=NC_000913:1277180..1278571;type=CDS+Genes;width=800
>>
>>The weird thing is that I can sometimes get the images to work partialy,
>>but mostly on the most 5' end of the genome, in the low numbers. For
>>instance:
>>
>>http://alien.lbl.gov/cgi-bin/gbrowse/bacteria/?start=1000;stop=5000;ref=NC_000913;width=800;version=100;label=CDS-Genes;id=3ed54d66a36f50ddbe3c9cade69202e3
>>http://alien.lbl.gov/cgi-bin/gbrowse_img/bacteria?name=NC_000913:1000..5000;type=CDS+Genes;width=800;
>>
>>This shows more of the image in the gbrowse_img portion, but it is about
>>1kb off, so I think that the larger the position, the larger the stop
>>position is being displayed? This is purely conjecture, but it seems a
>>little weird.
>>
>>Here is part of my gbrowse conf file (below). I will note that I have
>>installed many configurations of gbrowse, but this is my first one using
>>BioSQL. This does seem to be a biosql only problem, as the included
>>in-memory GFF database shows the correct image.
>>
>>Any help would be appreciated.
>>
>>Thanks,
>>Michael Cipriano
>>
>>
>>
>>Versions from the top of the files:
>>
>>gbrowse_img:
>>$VERSION = 1.64;
>># $Id: gbrowse_img.PLS,v 1.27.4.8.2.4 2005/11/11 22:00:39 lstein Exp $
>>
>>gbrowse:
>>$VERSION = 1.64;
>>$BIOGRAPHICS_VERSION = 1.65;
>># $Id: gbrowse.PLS,v 1.119.4.57.2.10 2005/11/15 19:20:40 lstein Exp $
>>
>>
>>Here is the CDS stanza from gbrowse.
>>
>>[CDS]
>>feature = CDS
>>glyph = transcript2
>>bgcolor = turquoise
>>fgcolor = black
>>height = 10
>>connector = solid
>>label = sub {
>> my $feature = shift;
>> my @notes;
>> if ($feature->primary_tag eq "CDS")
>> {
>> foreach (qw(product gene))
>> {
>> next unless $feature->has_tag($_);
>> @notes = $feature->each_tag_value($_);
>> last;
>> }
>> return $notes[0];
>> } else
>> {
>> return $feature->display_name;
>> }
>> }
>>
>>description = sub {
>> my $feature = shift;
>> my @notes;
>> if ($feature->primary_tag eq "CDS")
>> {
>> foreach (qw(note))
>> {
>> next unless $feature->has_tag($_);
>> @notes = $feature->each_tag_value($_);
>> last;
>> }
>> return unless @notes;
>> substr($notes[0],30) = '...' if length $notes[0] > 30;
>> return $notes[0];
>> } else
>> {
>> my $description;
>> foreach ($feature->all_tags)
>> {
>> my @values = $feature->each_tag_value($_);
>> $description .= $_ eq 'note' ? "@values" :
>>"$_=@values; ";
>> }
>> $description =~ s/; $//; # get rid of last
>> return $description;
>> }
>> }
>>link = sub {
>> my $feature = shift;
>> my $pk = $feature->primary_key();
>> return
>>"/cgi-bin/regtransbase?page=geneinfo&seqfeature_id=$pk";
>> }
>>key = Predicted transcripts
>>
>>
>>
>>
>>_______________________________________________
>>BioSQL-l mailing list
>>BioSQL-l at lists.open-bio.org
>>http://lists.open-bio.org/mailman/listinfo/biosql-l
>
>
> _______________________________________________
> BioSQL-l mailing list
> BioSQL-l at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/biosql-l
>
More information about the BioSQL-l
mailing list