Re: MS-GF+ added to SearchGUI and PeptideShaker

49 views
Skip to first unread message

Pratik Jagtap

unread,
Apr 11, 2014, 7:26:28 PM4/11/14
to Harald....@biomed.uib.no, Ira Cooke, Tim Griffin, Bart Gottschalk, Marc Vaudel, lennart.martens, Jim Johnson, Björn Grüning, Gerben Menschaert, Galaxy for Proteomics, John Chilton
Hello Harald,

Just confirming that the latest beta version of of PeptideShaker and SearchGUI is at the following location:


JJ was planning to have a preliminary look at the build and on what needs to be done for wrapping for Galaxy-P usage.

Bjoern, Ira, John and Gerben can offer inputs if need be.

Thanks and Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275


On Thu, Mar 20, 2014 at 4:28 PM, <Harald....@biomed.uib.no> wrote:

Hi all,

The bug in PeptideShaker with the parsing of PSMs with variable PTMs
has now been solved. Also fixed an issue with using relative paths in
the SearchGUI command line. New beta versions can be downloaded here:
https://www.dropbox.com/s/npb3gom82sjjgt8/PeptideShaker-0.26.3.1-beta.zip
https://www.dropbox.com/s/5jw67menx9vh4c3/SearchGUI-1.16.5.1-beta-windows.zip
https://www.dropbox.com/s/kyswwbsixegsjj8/SearchGUI-1.16.5.1-beta-mac_and_linux.tar.gz

And as always, if you find any bugs or have any comments don't hesitate
to let us know!


Best regards,
Harald




Quoting hba...@webmail.uib.no:


Hi again,

As mentioned in my previous e-mail: "there will probably be some bugs"...

Together with Sangtae I've now detected the first one: variable PTMs
in MS-GF+ mzid files are not handled correctly in PeptideShaker, resulting
in all PSMs from MS-GF+ with variable PTMs being kicked out before the
actual processing starts.

Already working on a fix. And this should not impact the development
of the wrappers which can go ahead as planned. Should just be a matter
of replacing the PeptideShaker jar file later on.

I'll let you know when a new version is available.

Best regards,
Harald





Quoting hba...@webmail.uib.no:


Hi all,

We've now completed the first beta versions of SearchGUI and
PeptideShaker with MS-GF+ support. You can download them here:
https://www.dropbox.com/s/f958r27tvgw7wc8/SearchGUI-1.16.5-beta-windows.zip
https://www.dropbox.com/s/1u0was6s4f8crrx/SearchGUI-1.16.5-beta-mac_and_linux.tar.gz
https://www.dropbox.com/s/te4jpm8yqaghqzx/PeptideShaker-0.26.3-beta.zip

We have not yet updated the command line wiki pages, so to see
the complete list of parameters available use:
java -cp SearchGUI-X.Y.Z.jar eu.isas.searchgui.cmd.SearchCLI
java -cp PeptideShaker-X.Y.Z.jar eu.isas.peptideshaker.cmd.PeptideShakerCLI
java -cp SearchGUI-X.Y.Z.jar eu.isas.searchgui.cmd.IdentificationParametersCLI
java -cp PeptideShaker-X.Y.Z.jar eu.isas.peptideshaker.cmd.IdentificationParametersCLI

Note that our MS-GF+ support is still in development and that
the results achieved will most likely improve in future versions.
We are in contact with Sangtae (the MS-GF+ developer), and hope
that he can help us further improve the results.

And being beta versions there will probably be some bugs we haven't
yet detected. So if you find any, please let us now.

But in any case, we should now have all we need to support SearchGUI
and PeptideShaker with MS-GF+ (and X!Tandem and OMSSA) in Galaxy right?

Finally, note that I added a new person in the CC: Gerben Menschaert.
He's a postdoc at Ghent University, and has showed an interest in
helping out with the implementation of the Galaxy wrappers for SearchGUI
and PeptideShaker. So instead of developing the same wrappers twice,
I thought I'd include him on our development. :)

Best regards,
Harald








Pratik Jagtap

unread,
Apr 11, 2014, 7:56:53 PM4/11/14
to Jim Johnson, Ira Cooke, Tim Griffin, Björn Grüning, Gerben Menschaert, John Chilton, Galaxy for Proteomics

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


On Fri, Apr 11, 2014 at 6:54 PM, Pratik Jagtap <pja...@umn.edu> wrote:
Thanks Harald !

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


On Fri, Apr 11, 2014 at 6:37 PM, <Harald....@biomed.uib.no> wrote:

Hi Pratik,

There is no need to use the beta versions in Dropbox anymore, as the proper releases are now available at Google Code.

SearchGUI: http://searchgui.googlecode.com
PeptideShaker: http://peptide-shaker.googlecode.com

Direct links to the downloads:
http://genesis.ugent.be/maven2/eu/isas/searchgui/SearchGUI/1.18.0/SearchGUI-1.18.0-windows.zip
http://genesis.ugent.be/maven2/eu/isas/searchgui/SearchGUI/1.18.0/SearchGUI-1.18.0-mac_and_linux.tar.gz
http://genesis.ugent.be/maven2/eu/isas/peptideshaker/PeptideShaker/0.28.0/PeptideShaker-0.28.0.zip

Best regards,
Harald





Quoting Pratik Jagtap <pja...@umn.edu>:

--
You received this message because you are subscribed to the Google Groups "Galaxy for Proteomics" group.
To post to this group, send email to gal...@umn.edu.
Visit this group at http://groups.google.com/a/umn.edu/group/galaxyp/.
To view this discussion on the web visit https://groups.google.com/a/umn.edu/d/msgid/galaxyp/CAFMfZ43%2BvvcnjRyB8HPf7b16%2BC4rRFrdXO33et5Kf%2BpzAgdojg%40mail.gmail.com.

To unsubscribe from this group and stop receiving emails from it, send an email to galaxyp+u...@umn.edu.





Björn Grüning

unread,
Apr 12, 2014, 3:53:31 AM4/12/14
to Pratik Jagtap, Jim Johnson, Ira Cooke, Tim Griffin, Gerben Menschaert, John Chilton, Galaxy for Proteomics
Hi all,

I started here with a huge cleanup.

https://bitbucket.org/BjoernGruening/peptideshaker

It will hopefully improve the installation and readability. I also
shifted around the admin files and separated them from the wrappers and
I used new Galaxy features to get rid of files.

From that start I will try to migrate to newer versions. Anyone
interested please feel free to join.

Cheers,
Bjoern

Pratik Jagtap

unread,
Apr 12, 2014, 9:20:03 AM4/12/14
to Björn Grüning, Harald....@biomed.uib.no, Ira Cooke, Tim Griffin, Bart Gottschalk, Marc Vaudel, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Galaxy for Proteomics
> If anyone can provide me with some example data, that would be great.

Please see Fusion RAW file along with FASTA file and parameters used here:


You can also find mzml and MGF files on the usegalaxyp site:


Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275


On Sat, Apr 12, 2014 at 7:19 AM, Björn Grüning <bjoern....@gmail.com> wrote:
Hi,

latest wrappers with a dependency on the latest versions are here:

 http://testtoolshed.g2.bx.psu.edu/view/bgruening/upload_testing

Ira and JJ should have write access. If will dig through all the options tomorrow. If anyone can provide me with some example data, that would be great.

Cheers,
Bjoern



There is no need to use the beta versions in Dropbox anymore, as the
proper releases are now available at Google Code.

SearchGUI: http://searchgui.googlecode.com
PeptideShaker: http://peptide-shaker.googlecode.com

Direct links to the downloads:
Best regards,
Harald





Quoting Pratik Jagtap <pja...@umn.edu>:

Hello Harald,

Just confirming that the latest beta version of of PeptideShaker and
SearchGUI is at the following location:

https://www.dropbox.com/s/npb3gom82sjjgt8/PeptideShaker-0.26.3.1-beta.zip
JJ was planning to have a preliminary look at the build and on what needs
to be done for wrapping for Galaxy-P usage.

Bjoern, Ira, John and Gerben can offer inputs if need be.

Thanks and Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275


On Thu, Mar 20, 2014 at 4:28 PM, <Harald....@biomed.uib.no> wrote:


Hi all,

The bug in PeptideShaker with the parsing of PSMs with variable PTMs
has now been solved. Also fixed an issue with using relative paths in
the SearchGUI command line. New beta versions can be downloaded here:
https://www.dropbox.com/s/npb3gom82sjjgt8/PeptideShaker-0.26.3.1-beta.zip

And as always, if you find any bugs or have any comments don't hesitate
to let us know!


Best regards,
Harald




Quoting hba...@webmail.uib.no:


Hi again,

As mentioned in my previous e-mail: "there will probably be some
bugs"...

Together with Sangtae I've now detected the first one: variable PTMs
in MS-GF+ mzid files are not handled correctly in PeptideShaker,
resulting
in all PSMs from MS-GF+ with variable PTMs being kicked out before the
actual processing starts.

Already working on a fix. And this should not impact the development
of the wrappers which can go ahead as planned. Should just be a matter
of replacing the PeptideShaker jar file later on.

I'll let you know when a new version is available.

Best regards,
Harald





Quoting hba...@webmail.uib.no:


Hi all,

We've now completed the first beta versions of SearchGUI and
PeptideShaker with MS-GF+ support. You can download them here:
https://www.dropbox.com/s/te4jpm8yqaghqzx/PeptideShaker-0.26.3-beta.zip


We have not yet updated the command line wiki pages, so to see
the complete list of parameters available use:
java -cp SearchGUI-X.Y.Z.jar eu.isas.searchgui.cmd.SearchCLI
java -cp PeptideShaker-X.Y.Z.jar eu.isas.peptideshaker.cmd.
PeptideShakerCLI
java -cp SearchGUI-X.Y.Z.jar eu.isas.searchgui.cmd.
IdentificationParametersCLI
java -cp PeptideShaker-X.Y.Z.jar eu.isas.peptideshaker.cmd.
IdentificationParametersCLI

Note that our MS-GF+ support is still in development and that
the results achieved will most likely improve in future versions.
We are in contact with Sangtae (the MS-GF+ developer), and hope
that he can help us further improve the results.

And being beta versions there will probably be some bugs we haven't
yet detected. So if you find any, please let us now.

But in any case, we should now have all we need to support SearchGUI
and PeptideShaker with MS-GF+ (and X!Tandem and OMSSA) in Galaxy
right?

Finally, note that I added a new person in the CC: Gerben Menschaert.
He's a postdoc at Ghent University, and has showed an interest in
helping out with the implementation of the Galaxy wrappers for
SearchGUI
and PeptideShaker. So instead of developing the same wrappers twice,
I thought I'd include him on our development. :)

Best regards,
Harald



Pratik Jagtap

unread,
Apr 21, 2014, 10:17:11 AM4/21/14
to Marc Vaudel, Harald....@biomed.uib.no, Björn Grüning, Ira Cooke, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Galaxy for Proteomics
Hello Everyone,

Last week, I tried SearchGUI and PeptideShaker on one of the Fusion files and could get the results for X!tandem, OMSSA and MS-GF+ (all together and separately). The search times and outputs look very impressive ! There are many possibilities with the outputs and their integration into workflows.

This was done on the downloaded version though and not the wrapper version that Bjoern, Ira and JJ are developing.

Thanks and Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


On Mon, Apr 21, 2014 at 8:59 AM, Marc Vaudel <mva...@gmail.com> wrote:
Hi Björn,

As a complement of information to the email of Harald here is how the new versions are working:
You can set the path of every temporary file created by every tool using the command line PathSettingsCLI. It works the same as all the other command lines options of SearchGUI and PeptideShaker. As usual, if you run the command line without argument the possible options will be displayed. As soon as this will be officially released we will update the wiki on the webpages. You also have the option to write all files into a temporary folder using temp_folder as command line option instead of every path specifically. That should be the most convenient for you. This option can also be used in the standard command line by appending the temp_folder flag, but doing it only once is enough. Finally, you can also set the path to the folders manually by editing the path.txt file which is located in the resources/conf folder. If you need an example, this file is created after the first call of PathSettingsCLI.

Note that some temporary files like the fasta indexes are used by both SearchGUI and PeptideShaker. Creating these indexes can be time consuming so it is recommended to use the same directory for both tools and between subsequent runs :)

Best regards,

Marc




2014-04-19 11:10 GMT+02:00 <Harald....@biomed.uib.no>:

Quoting Björn Grüning <bjoern....@gmail.com>:

That remembers me on one thing we need to fix for full Galaxy support. All write operations in Galaxy should be done in one temporary folder.
So .home/.compomics/gene_mappings is not allowed. Also I saw that derby is writing to $HOME. Is there a way to avoid that? The reason is, that Galaxy jobs can run on cluster nodes, there you do not have write access to $HOME. Also you can have different jobs running on the same node, resulting in conflicting IO.

Do you see any way to prevent SearchGUI and PeptideShaker to write to $HOME?

We've now added an option to the command line parameters -temp_folder
that can be used to set the folder where all the files needed during
the analysis are stored.

The bug with the min and max charge being set to the same value has
also been fixed.

So please give these new versions a go and let us know how if you
are now able to run the complete pipeline in Galaxy.

The new beta versions can be found here:
https://www.dropbox.com/s/80v4gpymijf927t/SearchGUI-1.18.2-beta-windows.zip
https://www.dropbox.com/s/28e40pjx9yaaxce/SearchGUI-1.18.2-beta-mac_and_linux.tar.gz
https://www.dropbox.com/s/gfhui9jpollifm3/PeptideShaker-0.28.1-beta.zip

Best regards,
Harald


Pratik Jagtap

unread,
Apr 28, 2014, 7:20:27 AM4/28/14
to Marc Vaudel, Harald....@biomed.uib.no, Björn Grüning, Ira Cooke, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, Galaxy for Proteomics
Hello Ira, JJ and Bjoern,

Hope that you are making progress on this project. Please keep us updated as you see fit.

Thanks,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


Pratik Jagtap

unread,
Apr 30, 2014, 4:45:53 PM4/30/14
to Galaxy for Proteomics
Note of progress...

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


On Wed, Apr 30, 2014 at 3:36 PM, <Harald....@biomed.uib.no> wrote:
Quoting Björn Grüning <bjoern....@gmail.com>:

Hi Harald,

can you point me to the latest releases? I would like to spend a few hours in the next days on updating the wrappers.

Thanks!
Bjoern


The latest versions (released yesterday) are available here:
http://genesis.ugent.be/maven2/eu/isas/searchgui/SearchGUI/1.18.2/SearchGUI-1.18.2-windows.zip
http://genesis.ugent.be/maven2/eu/isas/searchgui/SearchGUI/1.18.2/SearchGUI-1.18.2-mac_and_linux.tar.gz
http://genesis.ugent.be/maven2/eu/isas/peptideshaker/PeptideShaker/0.28.1/PeptideShaker-0.28.1.zip

Let me know if you come across any problems.

BTW, note that I added a new guy to the CC, Elvis Ndah. He's working
with Gerben in Ghent, and also has an interest in the wrappers. So
please include him in future communication too.

Best regards,
Harald

Pratik Jagtap

unread,
May 2, 2014, 8:00:26 AM5/2/14
to Galaxy for Proteomics
Updates on SearchGUI and PeptideShaker development.

Discussion about makeblastdb 32-bit version and that it is used only for OMSSA.



Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


On Fri, May 2, 2014 at 6:45 AM, <Harald....@biomed.uib.no> wrote:
Quoting Björn Grüning <bjoern....@gmail.com>:

I also found out that the name for the created decoy database from FastaCLI changed from input_database_decoy.fasta to input_database_concatenated_target_decoy.fasta.
Such changes will break the wrappers. Can we add a parameter the FastaCLI to provide a specific name?

Sure. We'll add a parameter for this to FastaCLI. The suffix will
not change unless changed by the user. Which at the moment is only
available in the GUI. So unless you add this option to the Galaxy
wrapper the suffix should always be "_concatenated_target_decoy.fasta".



We should also think about a way to communicate such changes (and new parameters, deprecated paramters etc) early, so that we can adjust the wrappers.

We use the PeptideShaker Google Group for this:
http://groups.google.com/group/peptide-shaker

Harald

Pratik Jagtap

unread,
May 4, 2014, 12:08:23 PM5/4/14
to Harald....@biomed.uib.no, Marc Vaudel, Ira Cooke, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Björn Grüning, Galaxy for Proteomics
Thanks Harald,

> The result is that the SearchGUI > PeptideShaker pipeline now works in
> Galaxy!

That is great news !

Bjoern and Ira - let us know if the plans are to place the wrapper in the toolshed. JJ mentioned about having some plans for having a development site so that the initial version(s) can be tested by a selected developers and users.

Regards,

Pratik



Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


On Sun, May 4, 2014 at 10:16 AM, <Harald....@biomed.uib.no> wrote:

Hi all,

Just a quick update on the status of makeblastdb on Linux 64 bit.

Björn and I have been working on this over the weekend and have
together come up with a solution that now seems to work. The problem
had two parts: 1) the makeblastdb Linux version included with SearchGUI
was supported on Linux 32 bit only (unless additional libraries were
installed), and 2) the version of makeblastdb in SearchGUI was outdated.

So we fixed this by adding a new option to the SearchGUI command line
(-makeblastdb_folder) that allows the user to change the folder where
makeblastdb is installed. And by adding new version of makeblastdb
for Linux 32 and 64 bit, and updating the makeblastdb command line to
support both the old and the new versions of makeblastdb.

The new included makeblastdb versions should work for most Linux
distributions, but if it does not the users can now simply refer to
their own makeblastdb installation.

We've also fixed some other issues and extended the command line support
for setting the decoy suffix tag when adding decoys using FastaCLI.

The result is that the SearchGUI > PeptideShaker pipeline now works in
Galaxy!

Best regards,
Harald

Björn Grüning

unread,
May 4, 2014, 12:17:33 PM5/4/14
to Pratik Jagtap, Harald....@biomed.uib.no, Marc Vaudel, Ira Cooke, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Hi Pratik,

>> The result is that the SearchGUI > PeptideShaker pipeline now works in
>> Galaxy!
>
> That is great news !
>
> Bjoern and Ira - let us know if the plans are to place the wrapper in the
> toolshed. JJ mentioned about having some plans for having a development
> site so that the initial version(s) can be tested by a selected developers
> and users.

After some testing, the plan is to move it to the toolshed, yes. I will
wait for Ira to review my changes.

What about a GalaxyP Docker Image? Easy to test for everyone:

https://github.com/bgruening/docker-recipes

Cheers,
Bjoern

Pratik Jagtap

unread,
May 4, 2014, 12:30:01 PM5/4/14
to Björn Grüning, Harald....@biomed.uib.no, Marc Vaudel, Ira Cooke, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Thanks Bjoern,

This sounds great.


> What about a GalaxyP Docker Image? Easy to test for everyone: https://github.com/bgruening/docker-> recipes

Bart - can you and your team look at what is the best way to move forward (for storing developed wrappers and for testing)?

Thanks and Regards,

Pratik



Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


Pratik Jagtap

unread,
May 6, 2014, 3:08:16 PM5/6/14
to Björn Grüning, Harald....@biomed.uib.no, Marc Vaudel, Ira Cooke, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Hello Bjoern and Ira,

I just spoke with Bart and his team.

It will be great if you or Ira can place the wrapper in Galaxy toolshed. The developers at MSI plan to install on http://galaxyp-dev.msi.umn.edu:8080/ for testing.

Please let us know if you have any questions.

Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275


Ira Cooke

unread,
May 6, 2014, 3:47:34 PM5/6/14
to Pratik Jagtap, Björn Grüning, Harald....@biomed.uib.no, Marc Vaudel, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Hi Pratik,

The wrapper is on the test toolshed ... it should include all the latest changes up to yesterday.


Cheers
Ira

Björn Grüning

unread,
May 6, 2014, 4:11:45 PM5/6/14
to Ira Cooke, Pratik Jagtap, Harald....@biomed.uib.no, Marc Vaudel, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
> Hi Pratik,

Is test toolshed ok? If the testing goes fine, we will migrate it to the
tool shed. Ira also integrated some unit tests and are waiting for the
automatic tool tests to run.

Cheers,
Bjoern

Ira Cooke

unread,
May 6, 2014, 4:15:40 PM5/6/14
to Björn Grüning, Pratik Jagtap, Harald....@biomed.uib.no, Marc Vaudel, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
> Ira also integrated some unit tests and are waiting for the automatic tool tests to run.

Well .. first run of the automated tests failed .. but don't let that put you off ;).

Pratik Jagtap

unread,
May 8, 2014, 11:51:56 AM5/8/14
to Ira Cooke, Björn Grüning, Harald....@biomed.uib.no, Marc Vaudel, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Hello  Bjoern and Ira,

> Is test toolshed ok? If the testing goes fine, we will migrate it to the tool shed. Ira also integrated 
> some unit tests and are waiting for the automatic tool tests to run.

Yes - I confirmed with JJ that test toolshed should work fine.

> Well .. first run of the automated tests failed .. but don't let that put you off ;).

:-) Let us know when the tool is tested and available.

Thanks again for all the work from everyone to come to this point !

Regards,

Pratik


Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


Björn Grüning

unread,
May 9, 2014, 6:36:06 AM5/9/14
to Pratik Jagtap, Ira Cooke, Harald....@biomed.uib.no, Marc Vaudel, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Hi all!

I put together a docker GalaxyP test image:

Install docker and and run:

docker run -d -p 8080:80 bgruening/galaxyp

You will find a complete Galaxy instance with peptide-shaker! under
http://localhost:8080

For more information please see:
https://github.com/bgruening/docker-recipes/tree/master/galaxyp

Happy testing!
Bjoern

P.S. JJ lets add all the other tools step by step :)

Marc Vaudel

unread,
May 12, 2014, 3:13:57 AM5/12/14
to Ira Cooke, Pratik Jagtap, Björn Grüning, Harald....@biomed.uib.no, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Hi Ira,

This is very peculiar, can it be an IO related issue like the one you reported earlier? It seems that the connection to our index of the protein database is somehow lost. I will improve the catching of this exception so that we get a clearer picture. We are busy with a course here so it might take a few days to get this sorted out :)

Best regards,

Marc


2014-05-12 8:42 GMT+02:00 Ira Cooke <I.C...@latrobe.edu.au>:
Dear Harald and Marc (sorry I’m not quite sure who this should go to). 

I have been trying to develop a test case that works on a minimal dataset … the dataset I’ve used is available in the test-data directory of the peptideshaker tool on the test toolshed (and also checked in to Bjoern’s peptideshaker bitbucket repository). 


Unfortunately I’m running into a non-deterministic kind of error … (I’ve attached a full transcript from executing the various stages).  On my system this fails around 80% of the time ( I ran it 10 times and got 2 successes).  I realise this sounds weird but when I repeatedly run it with the exact same inputs I sometimes get an error and sometimes success.  I realise that such things can be tricky to debug … I hope you can reproduce it on your system.  I’ve attached the full command that was run to generate this output as well … but of course you’d need to adjust it for your system.  This is the command that the galaxy tool currently runs.

Anyway … perhaps this isn’t such a great example dataset.  I created it by manually picking spectra and matching proteins from an existing dataset I have .. it has the advantage of being tiny .. and if we can get it to work it would be good as a toolshed test as it executes in just a few seconds.

Looking fwd to hearing your thoughts

Best wishes
Ira

Pratik Jagtap

unread,
May 12, 2014, 7:22:23 AM5/12/14
to Ira Cooke, Marc Vaudel, Björn Grüning, Harald....@biomed.uib.no, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Thanks Ira and Marc,

>  In the interim I think we will just move on with the galaxy tool … I think we can continue without a 
> working functional test for now anyway.

Ira and Bjoern - please let MSI developers (Bart and his team) know whether and where has the latest wrapper version of the SearchGUI and PeptideShaker tool been stored on galaxy toolshed. They can install it on an appropriate galaxyp instance for us to test.

Thanks again for all the hard work.

Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275


On Mon, May 12, 2014 at 2:50 AM, Ira Cooke <I.C...@latrobe.edu.au> wrote:
Hi Marc, 

Thanks for your quick response. If I discover anything else related  I’ll let you know.  As you say it sounds like an IO related issue … maybe the operating system doing some funny caching?  I don’t know to be honest.  In the interim I think we will just move on with the galaxy tool … I think we can continue without a working functional test for now anyway.

Cheers
Ira


--
You received this message because you are subscribed to the Google Groups "Galaxy for Proteomics" group.
To post to this group, send email to gal...@umn.edu.
Visit this group at http://groups.google.com/a/umn.edu/group/galaxyp/.

Björn Grüning

unread,
May 12, 2014, 7:58:03 AM5/12/14
to Pratik Jagtap, Ira Cooke, Marc Vaudel, Harald....@biomed.uib.no, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Hi all,

Am 12.05.2014 13:22, schrieb Pratik Jagtap:
> Thanks Ira and Marc,
>
>> In the interim I think we will just move on with the galaxy tool … I
> think we can continue without a
>> working functional test for now anyway.
>
> Ira and Bjoern - please let MSI developers (Bart and his team) know whether
> and where has the latest wrapper version of the SearchGUI and PeptideShaker
> tool been stored on galaxy toolshed. They can install it on an appropriate
> galaxyp instance for us to test.

The latest wrappers are here:
http://testtoolshed.g2.bx.psu.edu/view/bgruening/peptideshaker
https://bitbucket.org/BjoernGruening/peptideshaker

Ciao,
Bjoern
>> https://groups.google.com/a/umn.edu/d/msgid/galaxyp/CAE1e1dsb0jtLUZSuoweb4%3Dk%2BEwfexUJTEVM3V2_zQBWeAVvrqw%40mail.gmail.com<https://groups.google.com/a/umn.edu/d/msgid/galaxyp/CAE1e1dsb0jtLUZSuoweb4%3Dk%2BEwfexUJTEVM3V2_zQBWeAVvrqw%40mail.gmail.com?utm_medium=email&utm_source=footer>
>> .

Pratik Jagtap

unread,
May 12, 2014, 7:59:46 AM5/12/14
to Björn Grüning, Ira Cooke, Marc Vaudel, Harald....@biomed.uib.no, Tim Griffin, Bart Gottschalk, lennart.martens, Jim Johnson, Gerben Menschaert, John Chilton, Elvis Ndah, Galaxy for Proteomics
Thanks Bjoern !

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


Pratik Jagtap

unread,
May 20, 2014, 2:37:45 PM5/20/14
to Harald....@biomed.uib.no, Ira Cooke, Björn Grüning, Tim Griffin, Bart Gottschalk, Jim Johnson, Galaxy for Proteomics
Hello Ira and Bart,

> BTW, we used to refer to https://bitbucket.org/galaxyp/peptideshaker
> for Galaxy wrappers for PeptideShaker, but I guess this refers to the
> old wrappers? Is there a new page we should now refer to instead? We
> need this for the PeptideShaker manuscript which we are about to 
> resubmit.

Would this be the test toolshed link now?

Thanks,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


On Tue, May 20, 2014 at 8:32 AM, <Harald....@biomed.uib.no> wrote:

Hi Ira,

We've now had a look at this and tried to reduce the chance of the
exception being thrown. It is basically due to the IO taking too
long, so we've made it a bit more flexible and patient. For quicker
setups there should be no change. Please let us know if this new
version fixes the problem.

Also, we've updated the version of MS-GF+ as the old version did
not respond well to being given large amounts of memory. So we
got a new MS-GF+ version from Sangtae that solved this.

Note that setting the memory limit for MS-GF+ is usually done via
the GUI in either SearchGUI or PeptideShaker (Edit > Java Settings),
and we have yet to replicate this on the command line. The plan
is to add this as an option to the command line or simply reuse
the memory settings given to SearchGUI.

However, until this is implemented you have to go via our old way
of setting the memory limits, which is the file called JavaOptions
located in SearchGUI-1.18.6\resources\conf. Simply change the line
"-Xmx****M" to the wanted setting and run SearchGUI. This limit will
then be used for both SearchGUI, PeptideShaker and MS-GF+. Note
that the value has to be given in MB, i.e., 2048M instead of 2GB.

We'll simplify all of this as soon as we get the time. But for now
it should give you what you need to run MS-GF+ with higher memory
setting.

The new versions are available here:
http://genesis.ugent.be/maven2/eu/isas/searchgui/SearchGUI/1.18.6/SearchGUI-1.18.6-windows.zip
http://genesis.ugent.be/maven2/eu/isas/searchgui/SearchGUI/1.18.6/SearchGUI-1.18.6-mac_and_linux.tar.gz
http://genesis.ugent.be/maven2/eu/isas/peptideshaker/PeptideShaker/0.29.0/PeptideShaker-0.29.0.zip

BTW, we used to refer to https://bitbucket.org/galaxyp/peptideshaker
for Galaxy wrappers for PeptideShaker, but I guess this refers to the
old wrappers? Is there a new page we should now refer to instead? We
need this for the PeptideShaker manuscript which we are about to resubmit.

Best regards,
Harald





Quoting Ira Cooke <I.C...@latrobe.edu.au>:

Dear Harald and Marc (sorry I?m not quite sure who this should go to).

I have been trying to develop a test case that works on a minimal dataset ? the dataset I?ve used is available in the test-data directory of the peptideshaker tool on the test toolshed (and also checked in to Bjoern?s peptideshaker bitbucket repository).

https://bitbucket.org/iracooke/peptideshaker/src/486ef1d3f2c6f66f5cb4d678b666de667bc967e1/test-data/?at=default

Unfortunately I?m running into a non-deterministic kind of error ? (I?ve attached a full transcript from executing the various stages).  On my system this fails around 80% of the time ( I ran it 10 times and got 2 successes).  I realise this sounds weird but when I repeatedly run it with the exact same inputs I sometimes get an error and sometimes success.  I realise that such things can be tricky to debug ? I hope you can reproduce it on your system.  I?ve attached the full command that was run to generate this output as well ? but of course you?d need to adjust it for your system.  This is the command that the galaxy tool currently runs.

Anyway ? perhaps this isn?t such a great example dataset.  I created it by manually picking spectra and matching proteins from an existing dataset I have .. it has the advantage of being tiny .. and if we can get it to work it would be good as a toolshed test as it executes in just a few seconds.


Looking fwd to hearing your thoughts

Best wishes
Ira



On 9 May 2014, at 1:51 am, Pratik Jagtap <pja...@umn.edu<mailto:pjagtap...@umn.edu>> wrote:

Hello  Bjoern and Ira,

Is test toolshed ok? If the testing goes fine, we will migrate it to the tool shed. Ira also integrated
some unit tests and are waiting for the automatic tool tests to run.

Yes - I confirmed with JJ that test toolshed should work fine.

Well .. first run of the automated tests failed .. but don't let that put you off ;).

:-) Let us know when the tool is tested and available.

Thanks again for all the work from everyone to come to this point !

Regards,

Pratik


Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275


On Tue, May 6, 2014 at 3:15 PM, Ira Cooke <irac...@gmail.com<mailto:iraco...@gmail.com>> wrote:
 Ira also integrated some unit tests and are waiting for the automatic tool tests to run.

Well .. first run of the automated tests failed .. but don't let that put you off ;).

On 7 May 2014, at 6:11 am, Björn Grüning <bjoern....@gmail.com<mailto:bjoern.gruening@gmail.com>> wrote:

Hi Pratik,

Is test toolshed ok? If the testing goes fine, we will migrate it to the tool shed. Ira also integrated some unit tests and are waiting for the automatic tool tests to run.

Cheers,
Bjoern

The wrapper is on the test toolshed ... it should include all the latest changes up to yesterday.

http://testtoolshed.g2.bx.psu.edu/view/bgruening/peptideshaker

Cheers
Ira


On 7 May 2014, at 5:08 am, Pratik Jagtap <pja...@umn.edu<mailto:pjagtap...@umn.edu>> wrote:

Hello Bjoern and Ira,

I just spoke with Bart and his team.

It will be great if you or Ira can place the wrapper in Galaxy toolshed. The developers at MSI plan to install on http://galaxyp-dev.msi.umn.edu:8080/ for testing.

Please let us know if you have any questions.

Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275<tel:612-624-9275>



On Sun, May 4, 2014 at 11:30 AM, Pratik Jagtap <pja...@umn.edu<mailto:pjagtap...@umn.edu>> wrote:
Thanks Bjoern,

This sounds great.


What about a GalaxyP Docker Image? Easy to test for everyone: https://github.com/bgruening/docker-> recipes

Bart - can you and your team look at what is the best way to move forward (for storing developed wrappers and for testing)?

Thanks and Regards,

Pratik



Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275<tel:612-624-9275>



On Sun, May 4, 2014 at 11:17 AM, Björn Grüning <bjoern....@gmail.com<mailto:bjoern.gruening@gmail.com>> wrote:
Hi Pratik,


The result is that the SearchGUI > PeptideShaker pipeline now works in
Galaxy!

That is great news !

Bjoern and Ira - let us know if the plans are to place the wrapper in the
toolshed. JJ mentioned about having some plans for having a development
site so that the initial version(s) can be tested by a selected developers
and users.

After some testing, the plan is to move it to the toolshed, yes. I will wait for Ira to review my changes.

What about a GalaxyP Docker Image? Easy to test for everyone:

https://github.com/bgruening/docker-recipes

Cheers,
Bjoern


Regards,

Pratik



Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108
Phone: 612-624-9275<tel:612-624-9275>

Bart Gottschalk

unread,
May 20, 2014, 2:43:52 PM5/20/14
to Pratik Jagtap, Harald....@biomed.uib.no, Ira Cooke, Björn Grüning, Tim Griffin, Jim Johnson, Galaxy for Proteomics
It would make sense to me to use  http://testtoolshed.g2.bx.psu.edu/view/bgruening/peptideshaker

JJ, do you have any comments on this? 

Bart


Bart Gottschalk
Manager of Application Development Services and User Support
Minnesota Supercomputing Institute
University of Minnesota

Björn Grüning

unread,
May 20, 2014, 2:43:57 PM5/20/14
to Pratik Jagtap, Harald....@biomed.uib.no, Ira Cooke, Tim Griffin, Bart Gottschalk, Jim Johnson, Galaxy for Proteomics
Hi,

the new wrappers are located under:

https://bitbucket.org/BjoernGruening/peptideshaker

If everything is ok, I will create a PR for Bart and it can be
reintegrated into the original one.

@Bart to you intend to maintain the repository?
If you wish I can also migrate it to my galaxytools repository (I'm more
comfortable with one mega-repository than with X small ones).

@Pratik, are there any news from the testing unit :) ?

Cheers,
Bjoern

Am 20.05.2014 20:37, schrieb Pratik Jagtap:

Björn Grüning

unread,
May 20, 2014, 2:53:53 PM5/20/14
to Bart Gottschalk, Pratik Jagtap, Harald....@biomed.uib.no, Ira Cooke, Tim Griffin, Jim Johnson, Galaxy for Proteomics
Oh, that was also only a test repository, I would recommend the original
galaxyp account or the IUC account.

JJ?

Am 20.05.2014 20:43, schrieb Bart Gottschalk:
> It would make sense to me to use *
> http://testtoolshed.g2.bx.psu.edu/view/bgruening/peptideshaker
> <http://testtoolshed.g2.bx.psu.edu/view/bgruening/peptideshaker>*
>>> http://genesis.ugent.be/maven2/eu/isas/searchgui/
>>> SearchGUI/1.18.6/SearchGUI-1.18.6-windows.zip
>>> http://genesis.ugent.be/maven2/eu/isas/searchgui/
>>> SearchGUI/1.18.6/SearchGUI-1.18.6-mac_and_linux.tar.gz
>>> http://genesis.ugent.be/maven2/eu/isas/peptideshaker/
>>> PeptideShaker/0.29.0/PeptideShaker-0.29.0.zip
>>>
>>>> mailto:bjoern....@gmail.com>> wrote:
>>>>
>>>> Hi Pratik,
>>>>>>
>>>>>
>>>>> Is test toolshed ok? If the testing goes fine, we will migrate it to
>>>>> the tool shed. Ira also integrated some unit tests and are waiting for the
>>>>> automatic tool tests to run.
>>>>>
>>>>> Cheers,
>>>>> Bjoern
>>>>>
>>>>> The wrapper is on the test toolshed ... it should include all the
>>>>>> latest changes up to yesterday.
>>>>>>
>>>>>> http://testtoolshed.g2.bx.psu.edu/view/bgruening/peptideshaker
>>>>>>
>>>>>> Cheers
>>>>>> Ira
>>>>>>
>>>>>>
>>>>>> On 7 May 2014, at 5:08 am, Pratik Jagtap <pja...@umn.edu<mailto:

Pratik Jagtap

unread,
May 20, 2014, 3:02:56 PM5/20/14
to Björn Grüning, Bart Gottschalk, Harald....@biomed.uib.no, Ira Cooke, Tim Griffin, Jim Johnson, Galaxy for Proteomics
Hello Bjoern,

> @Pratik, are there any news from the testing unit :) ?

I will test it as soon as the test toolshed wrapper is installed on the galaxyp dev site.  I will let you know.

Regards,

Pratik

Pratik Jagtap,
Managing Director,
Center for Mass Spectrometry and Proteomics,
43 Gortner Laboratory
1479 Gortner Avenue
St. Paul, MN 55108


Reply all
Reply to author
Forward
0 new messages