Hi, I would like to add some tools of the EMBOSS suite (6.6.0) to my Galaxy instance. It seems that the only wrapper available for Galaxy is for the 5.0.0 version ( http://dev.list.galaxyproject.org/Galaxy-with-EMBOSS-5-or-EMBOSS-6-td4658043...).
I would like to know if someone has already tried (successfully or not) to wrap this tools base on the old wrappers to estimate the workload that awaits me (for prettyseq and water mostly). Could there be an invalid datatype format for example ?
I have another question regarding the NCBI BLAST+ suite : is there a way to edit the XML file to enable the selection of multiple databases as proposed by the command lines tool ?
Thanks in advance, F.T
On Wed, Jan 20, 2016 at 9:19 AM, Floreline TOUCHARD floreline.touchard@gmail.com wrote:
Hi,
I would like to add some tools of the EMBOSS suite (6.6.0) to my Galaxy instance. It seems that the only wrapper available for Galaxy is for the 5.0.0 version (http://dev.list.galaxyproject.org/Galaxy-with-EMBOSS-5-or-EMBOSS-6-td4658043...).
I would like to know if someone has already tried (successfully or not) to wrap this tools base on the old wrappers to estimate the workload that awaits me (for prettyseq and water mostly).
There is work in progress here, see: https://github.com/galaxyproject/tools-iuc/issues/65 https://github.com/galaxyproject/tools-iuc/pull/131
Have you compared the API options to see if anything changed here in EMBOSS? I don't recall any major changes to water, but I don't use prettyseq at all.
Could there be an invalid datatype format for example ?
Many datatypes were already defined for the EMBOSS 5.0.0 wrappers, so that is less likely to be a problem for EMBOSS 6.6.0
I have another question regarding the NCBI BLAST+ suite : is there a way to edit the XML file to enable the selection of multiple databases as proposed by the command lines tool ?
Possible by editing the wrappers, yes. However it will likely break if any of the paths have spaces in them due to the NCBI wanting a space separated list of database paths - a potential problem already in the makeblastdb wrapper.
Are you wanting to use multiple databases from the *.loc files at once, multiple databases from the current history, or both? Please log this as an enhancement request here:
https://github.com/peterjc/galaxy_blast/issues
Thanks,
Peter
Thanks for the answer !
I compared the two versions of tools. It does not seem to be any major differences indeed. I added these tools to my instance successfully. Additional options in version 6 are not to be integrate to my Galaxy instance immediately. I will update my XML file later if need arises. The old wrappers are perfect for the moment.
For the NCBI BLAST+ suite I would like to select multiple databases only from the *.loc files. There is no spaces in the paths in my *.loc files (however there is spaces in the names of these databases, but I guess this is not a problem). I thought I just have to add a tag "multiple=true" in the XML file but it seems a bit more complex. Am I missing something ?
F.T
2016-01-20 10:33 GMT+01:00 Peter Cock p.j.a.cock@googlemail.com:
On Wed, Jan 20, 2016 at 9:19 AM, Floreline TOUCHARD floreline.touchard@gmail.com wrote:
Hi,
I would like to add some tools of the EMBOSS suite (6.6.0) to my Galaxy instance. It seems that the only wrapper available for Galaxy is for the 5.0.0 version (
http://dev.list.galaxyproject.org/Galaxy-with-EMBOSS-5-or-EMBOSS-6-td4658043... ).
I would like to know if someone has already tried (successfully or not)
to
wrap this tools base on the old wrappers to estimate the workload that awaits me (for prettyseq and water mostly).
There is work in progress here, see: https://github.com/galaxyproject/tools-iuc/issues/65 https://github.com/galaxyproject/tools-iuc/pull/131
Have you compared the API options to see if anything changed here in EMBOSS? I don't recall any major changes to water, but I don't use prettyseq at all.
Could there be an invalid datatype format for example ?
Many datatypes were already defined for the EMBOSS 5.0.0 wrappers, so that is less likely to be a problem for EMBOSS 6.6.0
I have another question regarding the NCBI BLAST+ suite : is there a way
to
edit the XML file to enable the selection of multiple databases as
proposed
by the command lines tool ?
Possible by editing the wrappers, yes. However it will likely break if any of the paths have spaces in them due to the NCBI wanting a space separated list of database paths - a potential problem already in the makeblastdb wrapper.
Are you wanting to use multiple databases from the *.loc files at once, multiple databases from the current history, or both? Please log this as an enhancement request here:
https://github.com/peterjc/galaxy_blast/issues
Thanks,
Peter
On Wed, Jan 20, 2016 at 2:22 PM, Floreline TOUCHARD floreline.touchard@gmail.com wrote:
Thanks for the answer !
I compared the two versions of tools. It does not seem to be any major differences indeed. I added these tools to my instance successfully.
Great :)
Additional options in version 6 are not to be integrate to my Galaxy instance immediately. I will update my XML file later if need arises. The old wrappers are perfect for the moment.
If you have any feedback on the IUC work, please comment on the GitHub issue.
For the NCBI BLAST+ suite I would like to select multiple databases only from the *.loc files. There is no spaces in the paths in my *.loc files (however there is spaces in the names of these databases, but I guess this is not a problem). I thought I just have to add a tag "multiple=true" in the XML file but it seems a bit more complex. Am I missing something ?
As far as I know, the multiple="true" tag only works on input datasets from your history. But that would make sense... I don't think I have ever tried this.
Even if that worked, the <command> tag would also need modifying to account for the multiple values from db_opts.database.
Another potential option is to extend makeblastdb which can in principle combine multiple FASTA and pre-existing databases...
Peter
Hi,
A long time ago, in a far far svn repo, we did this "multiple=true" on Peter wrappers.
I will try to propose it quickly :)
Gildas
----------------------------------------------------------------- Gildas Le Corguillé - Bioinformatician/Bioanalyste Plateforme ABiMS (Analyses and Bioinformatics for Marine Science)
Station Biologique de Roscoff - UPMC/CNRS - FR2424 Place Georges Teissier 29680 Roscoff FRANCE tel: +33 2 98 29 23 81 http://abims.sb-roscoff.fr ------------------------------------------------------------------
Le 20 janv. 2016 à 15:37, Peter Cock p.j.a.cock@googlemail.com a écrit :
On Wed, Jan 20, 2016 at 2:22 PM, Floreline TOUCHARD floreline.touchard@gmail.com wrote:
Thanks for the answer !
I compared the two versions of tools. It does not seem to be any major differences indeed. I added these tools to my instance successfully.
Great :)
Additional options in version 6 are not to be integrate to my Galaxy instance immediately. I will update my XML file later if need arises. The old wrappers are perfect for the moment.
If you have any feedback on the IUC work, please comment on the GitHub issue.
For the NCBI BLAST+ suite I would like to select multiple databases only from the *.loc files. There is no spaces in the paths in my *.loc files (however there is spaces in the names of these databases, but I guess this is not a problem). I thought I just have to add a tag "multiple=true" in the XML file but it seems a bit more complex. Am I missing something ?
As far as I know, the multiple="true" tag only works on input datasets from your history. But that would make sense... I don't think I have ever tried this.
Even if that worked, the <command> tag would also need modifying to account for the multiple values from db_opts.database.
Another potential option is to extend makeblastdb which can in principle combine multiple FASTA and pre-existing databases...
Peter ___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: https://lists.galaxyproject.org/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
On Wed, Jan 20, 2016 at 2:43 PM, Gildas Le Corguillé lecorguille@sb-roscoff.fr wrote:
Hi,
A long time ago, in a far far svn repo, we did this "multiple=true" on Peter wrappers.
I will try to propose it quickly :)
Gildas
Great - a pull request against https://github.com/peterjc/galaxy_blast but if your SVN repo is still online, or you can make a patch/diff file, we could look at that and try to port the changes. My guess is a lot of macro work was done in the meantime so it may not be trivial to repeat your changes :(
Thanks,
Peter
Hi,
sorry for the late response ! I tested the modifications proposed in the "Allow to select multiple database" pull request and it works !!!
Thanks a lot ! Floreline.
P.S : I'm not familiar with all github functionnalities yet, but I will work on it ;)
2016-01-20 16:07 GMT+01:00 Peter Cock p.j.a.cock@googlemail.com:
On Wed, Jan 20, 2016 at 2:43 PM, Gildas Le Corguillé lecorguille@sb-roscoff.fr wrote:
Hi,
A long time ago, in a far far svn repo, we did this "multiple=true" on
Peter
wrappers.
I will try to propose it quickly :)
Gildas
Great - a pull request against https://github.com/peterjc/galaxy_blast but if your SVN repo is still online, or you can make a patch/diff file, we could look at that and try to port the changes. My guess is a lot of macro work was done in the meantime so it may not be trivial to repeat your changes :(
Thanks,
Peter
Good to know :)
https://github.com/peterjc/galaxy_blast/pull/77
The pull request looks good - sadly distracted by TravisCI failures from Galaxy/planemo/dependency changes.
Thanks,
Peter
On Tue, Feb 9, 2016 at 3:28 PM, Floreline TOUCHARD floreline.touchard@gmail.com wrote:
Hi,
sorry for the late response ! I tested the modifications proposed in the "Allow to select multiple database" pull request and it works !!!
Thanks a lot ! Floreline.
P.S : I'm not familiar with all github functionnalities yet, but I will work on it ;)
Hi Floreline,
just to add to this discussion that we were at some point in contact with Peter Rice of EMBOSS and we discussed possibility to generate more easily Galaxy wrappers out of the ACD files.
I still hope this will be integrated in EMBOSS directly, so we can reuse it easily and have EMBOSS Galaxy releases synchronised with EMBOSS releases.
Cheers, Bjoern
Am 20.01.2016 um 10:19 schrieb Floreline TOUCHARD:
Hi, I would like to add some tools of the EMBOSS suite (6.6.0) to my Galaxy instance. It seems that the only wrapper available for Galaxy is for the 5.0.0 version ( http://dev.list.galaxyproject.org/Galaxy-with-EMBOSS-5-or-EMBOSS-6-td4658043...).
I would like to know if someone has already tried (successfully or not) to wrap this tools base on the old wrappers to estimate the workload that awaits me (for prettyseq and water mostly). Could there be an invalid datatype format for example ?
I have another question regarding the NCBI BLAST+ suite : is there a way to edit the XML file to enable the selection of multiple databases as proposed by the command lines tool ?
Thanks in advance, F.T
Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: https://lists.galaxyproject.org/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
galaxy-dev@lists.galaxyproject.org
-
Björn Grüning -
Floreline TOUCHARD -
Gildas Le Corguillé -
Peter Cock