Re: Getting gbrowse_syn to work with MySQL
Hi,
All I know is that in your previous email you posted an error that
included a Bio::DB::GFF adaptor backtrace, but your posted MySQL tables
are of the Bio::DB::SeqFeature::Store schema. Therefore, somewhere in your
configuration and/or pipeline you have mixed adaptors being used. Perhaps
a default value is coming in somewhere where you should instead be
specifically assigning the value?
Do you have defined database stanzas in your conf as described here
http://gmod.org/wiki/GBrowse_2.0_HOWTO#Database_Definitions ?
It should look something like this for a MySQL database
[cerevisiae_test:database]
db_adaptor = Bio::DB::SeqFeature::Store
db_args = -adaptor DBI::mysql
-dsn cerevisiae_test
-user nobody
-pass hello
Then in each track stanza (or in [GENERAL] if everything is the same) you
would put
database = cerevisiae_test
I think that should remove the problem with dueling adaptors. Unless there
is something special about gbrowse_syn; I don't have much experience with
that.
Tim
On 3/2/12 1:04 AM, "Julien Yann Dutheil" <jy.dutheil@...> wrote:
>Hi Tim,
>
>Thanks a lot for your reply. I am using Bio::DB:SeqFeature::Store
>adaptor everywhere, as far as I can see... in addition, I have tried
>to put everything in a single database per species (contigs + genes),
>and use the [GENERAL] configuration. And there it works! So it seems
>to me that there is sthg wrong with the split database settings, both
>for the user/password issue and for the display one, just like if
>GBrowse_syn did not recognize the database=sthg option in the conf
>file... could that be the case?
>
>Thanks again for your precious help.
>
>Julien.
>
>On Thu, Mar 1, 2012 at 6:24 PM, Timothy Parnell
><Timothy.Parnell@...> wrote:
>> Hi Julien,
>>
>> It looks like you are using different database schemas (adaptors and
>> loaders).
>>
>> The bp_seqfeature_load.pl script is used with the newer
>> Bio::DB::SeqFeature::Store adaptor. It's designed for GFF3 data.
>>
>> The bp_bulk_load_gff.pl and bp_fast_load_gff.pl scripts are used with
>>the
>> older Bio::DB::GFF adaptor. It's best used with GFF2 data.
>>
>> Each database schema has different SQL tables, and are not
>> cross-compatable.
>>
>> Hope that helps,
>> Tim
>>
>>
>>
>> On 3/1/12 2:15 AM, "Julien Yann Dutheil" <jy.dutheil@...> wrote:
>>
>>>Dear all,
>>>
>>>Concerning the issue I reported before, I have removed the password
>>>for www-data in MySQL. This allows me to go further (yet I still do
>>>not understand the previous problem). Now, however, I am facing a new
>>>error:
>>>
>>>------------- EXCEPTION -------------
>>>MSG: Couldn't execute query SELECT fref,
>>> IF(ISNULL(gclass),'Sequence',gclass),
>>> min(fstart),
>>> max(fstop),
>>> fstrand
>>> FROM fdata,fgroup
>>> WHERE fgroup.gname=?
>>> AND fgroup.gclass=?
>>> AND fdata.fref=?
>>> AND fgroup.gid=fdata.gid
>>> GROUP BY fref,fstrand
>>>:
>>> Table 'fungi_contigs.fdata' doesn't exist
>>>
>>>STACK Bio::DB::GFF::Adaptor::dbi::caching_handle::do_query
>>>/usr/local/share/perl/5.10.1/Bio/DB/GFF/Adaptor/dbi/caching_handle.pm:12
>>>8
>>>
>>>And indeed, looking at the content of the fungi_contigs database, I see:
>>>
>>>+---------------------------+
>>>| attribute |
>>>| attributelist |
>>>| feature |
>>>| interval_stats |
>>>| locationlist |
>>>| meta |
>>>| name |
>>>| parent2child |
>>>| sequence |
>>>| typelist |
>>>+---------------------------+
>>>
>>>I cannot understand what I am doing wrong (used bp_seqfeature_load.pl
>>>for generating the tables). Any help or tip would be highly welcome :s
>>>
>>>All the best,
>>>
>>>Julien.
>>>
>>>
>>>
>>>On Mon, Feb 27, 2012 at 4:27 PM, Julien Yann Dutheil
>>><jy.dutheil@...> wrote:
>>>> Dear all,
>>>>
>>>> I am trying to set up gbrowse_syn with 5 fungal genomes. Following the
>>>> tutorial and example files, I have managed to get gbrowse_syn to work
>>>> with my data on flat GFF files. As recommended, in order to improve
>>>> the performances, and also in order to enable gene names queries in
>>>> addition to coordinates queries, I have uploaded my GFF files into
>>>> MySQL tables, again following the tutorial. For my 5 species, I have 2
>>>> tables, [species]_contigs and [species]_genes, containing the contigs
>>>> data eventually mapped on chromosomes, and the gene annotations (mRNA
>>>> and CDS). There come the issues, as I could not manage to have
>>>> gbrowse_syn displaying the data as with the memory adaptor. After
>>>> several days of investigation, I could narrow the problem to this: the
>>>> 5 species configuration files are on this model:
>>>>
>>>> [GENERAL]
>>>> description = My fungi database
>>>> db_adaptor = Bio::DB::SeqFeature::Store
>>>> database = fungi_contigs
>>>>
>>>> [...]
>>>>
>>>> and further away:
>>>>
>>>> #################################
>>>> # database definitions
>>>> #################################
>>>>
>>>> [fungi_contigs:database]
>>>> db_adaptor = Bio::DB::SeqFeature::Store
>>>> db_args = -dsn
>>>>
>>>>dbi:mysql:database=fungi_contigs;host=localhost;user=gbrowse_syn;passwo
>>>>rd
>>>>=mypassword
>>>>
>>>> #################################
>>>> # database definitions
>>>> #################################
>>>>
>>>> [fungi_genes:database]
>>>> db_adaptor = Bio::DB::SeqFeature::Store
>>>> db_args = -dsn
>>>>
>>>>dbi:mysql:database=fungi_genes;host=localhost;user=gbrowse_syn;password
>>>>=m
>>>>ypassword
>>>>
>>>> The corresponding tables fungi_contigs and fungi_genes have been
>>>> granted SELECT access in mysql for the user gbrowse_syn. The error I
>>>> get with these settings is: "Cannot connect to database for
>>>> 'www-data' <at> 'localhost' with password=NO".
>>>>
>>>> The strange thing is that, if I use these settings instead:
>>>>
>>>> [GENERAL]
>>>> description = My fungi database
>>>> db_adaptor = Bio::DB::SeqFeature::Store
>>>> db_args = -dsn
>>>>
>>>>dbi:mysql:database=fungi_contigs;host=localhost;user=gbrowse_syn;passwo
>>>>rd
>>>>=Sposci2012
>>>>
>>>> Then it works again, without any access complaint, yet not displaying
>>>> the gene information, only the contigs present in fungi_contigs.
>>>>
>>>> This is puzzling me a bit... I am sure I am doing something wrong, but
>>>> I cannot figure it out, and any help / tip would be very much
>>>> appreciated. I can of course provide more information on my
>>>> configuration files if that would be needed. I have to specify that I
>>>> have configured gbrowse_syn in a freshly installed system (debian6 in
>>>> a virtual machine), using the installation from CPAN.
>>>>
>>>> Many thanks in advance for your help,
>>>>
>>>> Julien.
>>>>
>>>>
>>>> --
>>>> Julien Y. Dutheil, Ph-D
>>>> 0 (+49) 6421 178 530
>>>>
>>>> § Max Planck Institute for Terrestrial Microbiology
>>>> Department of Organismic Interactions
>>>> Marburg -- GERMANY
>>>>
>>>> § Intitute of Evolutionary Sciences - Montpellier
>>>> University of Montpellier 2 -- FRANCE
>>>
>>>
>>>
>>>--
>>>Julien Y. Dutheil, Ph-D
>>>0 (+49) 6421 178 530
>>>
>>>§ Max Planck Institute for Terrestrial Microbiology
>>>Department of Organismic Interactions
>>>Marburg -- GERMANY
>>>
>>>§ Intitute of Evolutionary Sciences - Montpellier
>>>University of Montpellier 2 -- FRANCE
>>>
>>>------------------------------------------------------------------------
>>>--
>>>----
>>>Virtualization & Cloud Management Using Capacity Planning
>>>Cloud computing makes use of virtualization - but cloud computing
>>>also focuses on allowing computing to be delivered as a service.
>>>http://www.accelacomm.com/jaw/sfnl/114/51521223/
>>>_______________________________________________
>>>Gmod-gbrowse mailing list
>>>Gmod-gbrowse@...
>>>https://lists.sourceforge.net/lists/listinfo/gmod-gbrowse
>>
>
>
>
>--
>Julien Y. Dutheil, Ph-D
>0 (+49) 6421 178 530
>
>§ Max Planck Institute for Terrestrial Microbiology
>Department of Organismic Interactions
>Marburg -- GERMANY
>
>§ Intitute of Evolutionary Sciences - Montpellier
>University of Montpellier 2 -- FRANCE
------------------------------------------------------------------------------
Virtualization & Cloud Management Using Capacity Planning
Cloud computing makes use of virtualization - but cloud computing
also focuses on allowing computing to be delivered as a service.
http://www.accelacomm.com/jaw/sfnl/114/51521223/
No need for a feature request, that will be fine, I was just wondering it is was possible. Thanks
RSS Feed