Jalview Version 2.8 now available from new look www.jalview.org

Dear all.

We are very pleased to announce that Jalview 2.8 is now available from our brand new website at www.jalview.org.

What's new ?

···

========

Jalview 2.8 includes a number of enhancements and new features that have been in development since July 2010. It is also the first Jalview release to incorporate RNA visualization features developed by Lauren Lui and Jan Engelhart during their Google Summer of Code (http://code.google.com/soc/) projects.

Highlights in Jalview Version 2.8

* JABAWS support (http://www.compbio.dundee.ac.uk/jabaws/)
- improved client and support for new JABAWS 2.0 Services
   o AACon alignment conservation
   o Protein disorder - DisEMBL, RONN, GlobPlot and IUPred
   o Clustal Omega for creating huge protein alignments

* Support for RNA
   o Import sequence and alignment associated WUSS or VIENNA dot-bracket notation from files and the RFAM database
   o Interactive editing of RNA secondary structure annotation
   o Colour scheme for purine/pyrimidine and to highlight RNA helices
   o RNA canonical base pair consensus score and sequence logo
   o VARNA (http://varna.lri.fr/) RNA secondary structure viewer in the Jalview Desktop

Issues fixed in this release include compatibility of the desktop with OSX Mountain lion, Windows 8, numerous minor glitches and patches to the JalviewLite Javascript API. For the comprehensive list of new features and bug fixes, take a look at the Jalview 2.8 Release Notes (http://www.jalview.org/development/release-history/Version-28).

The New Look www.jalview.org site

In the last year, we've been busy developing a logo and a new website for Jalview.

The new look website includes blogs for training, community and Jalview development news, and we'll be working to add more functionality over the coming months. If you have any comments or
suggestions, then we'd love to hear from you. You can find all the ways to get in contact with us over at http://www.jalview.org/community

Happy jalviewing!
From all the Jalview Team
http://www.jalview.org/about/credits

Hi there!

This is really a good news! Congratulation to the great job.

Also the new web site looks pretty nice !

Best ,
Paolo

···

On Nov 12, 2012, at 2:40 PM, Jim Procter <jprocter@compbio.dundee.ac.uk> wrote:

Dear all.

We are very pleased to announce that Jalview 2.8 is now available from
our brand new website at www.jalview.org.

What's new ?

Jalview 2.8 includes a number of enhancements and new features that have
been in development since July 2010. It is also the first Jalview
release to incorporate RNA visualization features developed by Lauren
Lui and Jan Engelhart during their Google Summer of Code
(http://code.google.com/soc/) projects.

Highlights in Jalview Version 2.8

* JABAWS support (http://www.compbio.dundee.ac.uk/jabaws/)
- improved client and support for new JABAWS 2.0 Services
  o AACon alignment conservation
  o Protein disorder - DisEMBL, RONN, GlobPlot and IUPred
  o Clustal Omega for creating huge protein alignments

* Support for RNA
  o Import sequence and alignment associated WUSS or VIENNA dot-bracket
notation from files and the RFAM database
  o Interactive editing of RNA secondary structure annotation
  o Colour scheme for purine/pyrimidine and to highlight RNA helices
  o RNA canonical base pair consensus score and sequence logo
  o VARNA (http://varna.lri.fr/) RNA secondary structure viewer in the
Jalview Desktop

Issues fixed in this release include compatibility of the desktop with
OSX Mountain lion, Windows 8, numerous minor glitches and patches to the
JalviewLite Javascript API. For the comprehensive list of new features
and bug fixes, take a look at the Jalview 2.8 Release Notes
(http://www.jalview.org/development/release-history/Version-28).

The New Look www.jalview.org site

In the last year, we've been busy developing a logo and a new website
for Jalview.

The new look website includes blogs for training, community and Jalview
development news, and we'll be working to add more functionality over
the coming months. If you have any comments or
suggestions, then we'd love to hear from you. You can find all the ways
to get in contact with us over at http://www.jalview.org/community

Happy jalviewing!
From all the Jalview Team
http://www.jalview.org/about/credits

_______________________________________________
Jalview-dev mailing list
Jalview-dev@jalview.org
http://www.compbio.dundee.ac.uk/mailman/listinfo/jalview-dev

Hi Paolo - thanks for the nice words ... I'd like to thank you again for your contributions, too!

It'd be great to be able to run M-COFFEE from Jalview directly, now we can visualize the reliability score. That's a bit tricky, but if you're interested in helping out then we should talk more :slight_smile:

Jim.

Hi Jim,

It’s a think we could discuss. What would be the trickiest part to make Jalview able to run M-Coffee? Should it based on a web-service interaction?

One more thing, I was trying to integrate the latest release in our web server to show the T-Coffee alignment score, but I’m unable to get it work.

It shows the alignment as usual, but not the coloured scores. This is the applet code I’m using:

<applet id="applet" code="jalview.bin.JalviewLite" width="800" height="600" archive="public/jalviewApplet.jar">
<param name="file" value="http://localhost:9000/data/7ce04b2b/result.fasta_aln">
<param name="scoreFile" value="http://localhost:9000/data/7ce04b2b/result.score_ascii" >
<param name="embedded" value="true">
<param name="showFeatureSettings" value="false">
<param name="widthScale" value="1.4">
<param name="heightScale" value="1.4">
<param name="nojmol" value="true" >

Any idea?

Cheers,
Paolo

···

On Nov 14, 2012, at 5:31 PM, Jim Procter <jprocter@compbio.dundee.ac.uk> wrote:

Hi Paolo - thanks for the nice words … I’d like to thank you again for
your contributions, too!

It’d be great to be able to run M-COFFEE from Jalview directly, now we
can visualize the reliability score. That’s a bit tricky, but if you’re
interested in helping out then we should talk more :slight_smile:

Jim.


Jalview-dev mailing list
Jalview-dev@jalview.org
http://www.compbio.dundee.ac.uk/mailman/listinfo/jalview-dev

It’s a think we could discuss. What would be the trickiest part to make Jalview able to run M-Coffee? Should it based on a web-service interaction?

Ideally, Jalview should be able to run M-Coffee from the Web Service->Alignment submenu. This is nearly possible now, except that the alignment webservices don’t include support to send back any annotation, only the alignment data. That means that a new type of service needs to be created that returns annotation on the alignment.
One way might be to add a ‘supplementary information’ service to Jabaws that Jalview could call to discover any additional information provided by a service in addition to the standard data returned… so for an M-COFFEE alignment, jalview gets the alignment as normal, but then queries the jabaws server for the score file. Such a service might also allow Jalview to recover the guide trees and any other data generated that aren’t currently returned to the user.

Apart from supporting this additional nuance, getting M-Coffee into JABAWS is also a little tricky. The way that M-COFFEE downloads all the additional tools is really amazing, but might alarm some sysadmins - who probably don’t expect a web service to start downloading and compiling code after they’ve installed it. Ideally, the installation of the T-COFFEE suite shipped with JABAWS should be preconfigured to run M-COFFEE using the four other alignment methods available, and come with instructions for telling M-COFFEE about any other alignment tools available on the system. Is there a standard way of doing that when compiling the suite ?

One more thing, I was trying to integrate the latest release in our web server to show the T-Coffee alignment score, but I’m unable to get it work.

It shows the alignment as usual, but not the coloured scores. This is the applet code I’m using:

<applet id="applet" code="jalview.bin.JalviewLite" width="800" height="600" archive="public/jalviewApplet.jar">
<param name="file" value="http://localhost:9000/data/7ce04b2b/result.fasta_aln">
<param name="scoreFile" value="http://localhost:9000/data/7ce04b2b/result.score_ascii" >
<param name="embedded" value="true">
<param name="showFeatureSettings" value="false">
<param name="widthScale" value="1.4">
<param name="heightScale" value="1.4">
<param name="nojmol" value="true" >

hmm. not sure - is there anything shown on the java console ? The code is definitely in there still :slight_smile:

Add to the applet tag to make the applet more verbose. There should be some kind of message explaining why it didn’t add the score file.

Jim.

···

Hi Paolo.

On 14/11/2012 20:21, Paolo Di Tommaso wrote:

Hi Jim,

It turns out that the T-Coffee score file format has an extended syntax including the residue numbers at the begin and the end of each line.

This was not supported by my original parser. I’ve just submitted a patch to handle it correctly.

I’m going to reply to the remaining part of your message in a separate email.

Cheers,
Paolo

···

Hi Paolo.

On 14/11/2012 20:21, Paolo Di Tommaso wrote:

Hi Jim,

About the M-Coffee web service integration in Jalview, I see two alternatives.

The first is the one you’ve described, modify the web service so that it will able to send back the T-Coffee score file (as well as other annotations if required).

Regarding the problem of T-Coffee that automatically download and install the missing packages, this is true only if you use the source code based distribution package.

But we are also distributing a pre-compiled installer which contains all the dependencies required by T-Coffee. Once you have installed this, T-Coffee has everything it needs to work and no magic installations happen :wink:

You can find it at the following link http://www.tcoffee.org/Packages/Stable/Latest/linux/

In any case, as far as I’ve understood having Jalview to support M-Coffee in this way, is more a problem to make JABAWS able to handle the extra annotation produced by T-Coffee, more than modify Jalview to support it.

An alternative approach could be to have Jalview interacting with our T-Coffee server (http://tcoffee.crg.cat) which provides as REST based interface. It is a very easy API based on a GET request which return an XML response. It has already support for all the result files produced by T-Coffee (alignment, scores, etc).

In this case the problem would be to make Jalview able to interact with it.

I will try to join the Jalview hackathon to collaborate on on with this.

Cheers,
Paolo

···

Hi Paolo.

On 14/11/2012 20:21, Paolo Di Tommaso wrote:

Morning Paulo,

The first is the one you've described, modify the web service so that
it will able to send back the T-Coffee score file (as well as other
annotations if required).

I think this would be preferable in the long run - but it is a not insignificant amount of work to create new service archetypes in JABAWS.

Regarding the problem of T-Coffee that automatically download and
install the missing packages, this is true only if you use the source
code based distribution package.

But we are also distributing a pre-compiled installer which contains
all the dependencies required by T-Coffee. Once you have installed
this, T-Coffee has everything it needs to work and no magic
installations happen :wink:

ah. ok. Then all that would be needed is to get JABAWS to use that version, which is a matter of changing the config file.

You can find it at the following link
http://www.tcoffee.org/Packages/Stable/Latest/linux/

ok. Thanks for that - am cc-ing this to the jabaws discussion list, so it's on the record over there.

In any case, as far as I've understood having Jalview to support
M-Coffee in this way, is more a problem to make JABAWS able to handle
the extra annotation produced by T-Coffee, more than modify Jalview to
support it.

exactly.

An alternative approach could be to have Jalview interacting with our
T-Coffee server (http://tcoffee.crg.cat) which provides as REST based
interface. It is a very easy API based on a GET request which return
an XML response. It has already support for all the result files
produced by T-Coffee (alignment, scores, etc).

In this case the problem would be to make Jalview able to interact
with it.

A native T-COFFEE client would be a great addition.

I will try to join the Jalview hackathon to collaborate on on with this.

I was going to suggest that ! I'll add it to the list of topics.

Jim.

···

On Fri Nov 16 10:44:56 2012, Paolo Di Tommaso wrote:

Hi Paolo.

···

On 16/11/2012 11:50, Jim Procter wrote:

An alternative approach could be to have Jalview interacting with our
T-Coffee server (http://tcoffee.crg.cat) which provides as REST based
interface. It is a very easy API based on a GET request which return
an XML response. It has already support for all the result files
produced by T-Coffee (alignment, scores, etc).

In this case the problem would be to make Jalview able to interact
with it.

I was just experimenting with POSTing to the http://tcoffee.crg.cat/apps/tcoffee/do:core URL with an alignment - but didn't have much luck. Do you have any documentation for the API ?

Jim.

Hi Jim,

Sorry if I didn’t join the Jalview hackathon past month, but at the end I had to participate to an overlapping event.

About the T-Coffee server API unfortunately there is any documentation. But the good news is that a full featured Java client is available. I think it could be easily integrated with Jalview to make it interacts with the T-Coffee web server.

The project is available on Github, here

https://github.com/paoloditommaso/tcoffee-client

The main client class is this

https://github.com/paoloditommaso/tcoffee-client/blob/master/src/org/tcoffee/client/TCoffeeClient.java

You may find an example how to use it, here

https://github.com/paoloditommaso/tcoffee-client/blob/master/src/org/tcoffee/client/CloudCoffee.java

Let me know if it helps.

Cheers,
Paolo

···

On Feb 16, 2013, at 3:19 PM, Jim Procter <jprocter@compbio.dundee.ac.uk> wrote:

Hi Paolo.

On 16/11/2012 11:50, Jim Procter wrote:

An alternative approach could be to have Jalview interacting with our
T-Coffee server (http://tcoffee.crg.cat) which provides as REST based
interface. It is a very easy API based on a GET request which return
an XML response. It has already support for all the result files
produced by T-Coffee (alignment, scores, etc).

In this case the problem would be to make Jalview able to interact
with it.

I was just experimenting with POSTing to the
http://tcoffee.crg.cat/apps/tcoffee/do:core URL with an alignment - but
didn’t have much luck. Do you have any documentation for the API ?

Jim.


Jalview-dev mailing list
Jalview-dev@jalview.org
http://www.compbio.dundee.ac.uk/mailman/listinfo/jalview-dev

The main client class is this

https://github.com/paoloditommaso/tcoffee-client/blob/master/src/org/tcoffee/client/TCoffeeClient.java

I see your using Xstream (http://xstream.codehaus.org/) - I’d not heard of it before, it looks really useful!

From what I can see here, the client takes native tcoffee suite command lines. Is there some kind of machine readable command line argument description file in the source ?

Jim.

···

Hi Paolo - thanks for this, it helps muchly !

On 16/02/2013 16:27, Paolo Di Tommaso wrote:

Hi Jim,

https://github.com/paoloditommaso/tcoffee-client/blob/master/src/org/tcoffee/client/TCoffeeClient.java

I see your using Xstream (http://xstream.codehaus.org/) - I’d not heard of it before, it looks really useful!

Yes, pretty useful!

From what I can see here, the client takes native tcoffee suite command lines. Is there some kind of machine readable command line argument description file in the source ?

What do you mean by “machine readable command line argument description”. The idea is exactly as you said, the client takes any native T-Coffee command line options, and execute it on the server side. When it terminates, collects the result files and send them back to the client.

The only requirement is to prefix input file arguments with the “file:” string, doing that the client takes care to upload them in the scratch folder where T-Coffe will be execute.

It is tricky, isn’t it ? :wink:

Cheers,
Paolo

Hi.

What do you mean by "machine readable command line argument
description". The idea is exactly as you said, the client takes any
native T-Coffee command line options, and execute it on the server
side. When it terminates, collects the result files and send them back
to the client.

what i mean is, can I get an xml description of all the command line options ?

The only requirement is to prefix input file arguments with the
"file:" string, doing that the client takes care to upload them in the
scratch folder where T-Coffe will be execute.

ok. That's good to know - although with Jalview I try to avoid creating files on the client side if possible.

Incorporating the client would be a good summer project for a student - want to propose it as a google summer of code project ?

Jim.

···

On Tue Feb 19 11:54:47 2013, Paolo Di Tommaso wrote:

Hi Jim,

what i mean is, can I get an xml description of all the command line
options ?

No, unfortunately not. How do you think this could be useful ?

The only requirement is to prefix input file arguments with the
"file:" string, doing that the client takes care to upload them in the
scratch folder where T-Coffe will be execute.

ok. That's good to know - although with Jalview I try to avoid creating
files on the client side if possible

I see, but it wouldn't be so difficult to have the client to handle a byte buffer instead of a File.

Incorporating the client would be a good summer project for a student -
want to propose it as a google summer of code project ?

Yes, it could fit as a google summer project, but quite likely we are already going to apply for another project; and also frankly, I think this is more a work on Jalview side, so it could have more sense if you apply for it.

Cheers,
Paolo

Hello Paolo.

what i mean is, can I get an xml description of all the command line
options ?

No, unfortunately not. How do you think this could be useful ?

It would allow the client to autogenerate command lines. In Jalview's situation, a parameter dialog could be shown with all the different options available for a particular command. This is exactly why we created JABAWS (which has a command line generator and XML representation for describing a command line).

The only requirement is to prefix input file arguments with the
"file:" string, doing that the client takes care to upload them in the
scratch folder where T-Coffe will be execute.

ok. That's good to know - although with Jalview I try to avoid creating
files on the client side if possible

I see, but it wouldn't be so difficult to have the client to handle a byte buffer instead of a File.

it shouldn't be - and would be much more desirable. Maybe you might even consider having an API with that option.

Incorporating the client would be a good summer project for a student -
want to propose it as a google summer of code project ?

Yes, it could fit as a google summer project, but quite likely we are already going to apply for another project; and also frankly, I think this is more a work on Jalview side, so it could have more sense if you apply for it.

I'd certainly act as mentor.. but I was thinking you might consider acting as consultant for the project, should a student apply to the NESCent org with a proposal like this.

So - is there going to be a CRG google summer of code organisation this year ? That would be great !

Jim.

···

On Wed Feb 20 13:42:32 2013, Paolo Di Tommaso wrote: