Some suggestions for jalview and a bug

Jalview Discussion
#1

Hi,

I've been using Jalview some time and I really like the way the program works. It is simple but provides me with everything I need to do. I have a few suggestions for the program, some small and some considerably larger. These suggestions would probably make my dream sequence viewer program.

1. Allow local sequence alignments. I don't always want to have to connect to a server to do a sequence alignment and our computers in lab are powerful enough for local alignments.

2. Global option to not display borders around groups. When I make groups, the border often infringes on the letters and makes it difficult to read. If there were an option somewhere to disable all borders, it would help a lot

3. Improve custom color. One specific feature would be to assign multiple amino acids a single color at the same time. For example, I made a custom color for labeling certain beta strands called "yellow." Basically everything is yellow except for histidines. TO make this, I had to click on each amino acid and then click on the yellow color which was somewhat tedious. It would have been nice to select everything and define it as yellow and then histidines as blue.

4. Improve print options and add print preview. It would be nice to have a feature for printing where Jalview would decide how small to make the print to get everything to fit. This could be a vertical fit or a horizontal fit option

5. Allow editting multiple names at once in a sequence. I have a list of GI numbers and their aligned sequences in Jalview and to some gorups i would like to add "1-" or "2-" to all of the seuqneces selected. This could be great for keeping track of which group of sequences are in a group. Ideally the Muscle scoring should keep the groups together but I would prefer to just look at the number.

6. Add and align sequences without destroying groups. I label my groups, then decide I want to add one more sequence. I wish there was a way to add a sequence and align it without ruining all my hard 10 minutes of work.

7. This suggestion is kind of out there, but I use jalview mainly for looking at groups in cytoscape. If there was any way to tie-in to the cytoscape plug in blast 2 similarity graph, that would be amazing. Of course this woul dbe more of a project than a small suggestion but it would be incredibly useful.

Also I have come across one very annoying bug.

Normaly my projects are saved with 8 groups each with the custom color I call "yellow." When I restore the project, only 1 group is actualy colored with yellow. The rest are still groups and I can see the border and color them again, but they certainly are not restored.

Thanks for your time. If any of the suggestions are already features, please let me know!
-Daniel Hitchcock

#2

Thanks for this email, Daniel - lots of very useful suggestions!

1. Allow local sequence alignments. I don't always want to have to connect to a server to do a sequence alignment and our computers in lab are powerful enough for local alignments.
   

We've got a new version of the jalview alignment service engine in the pipeline that will let you do local alignment. I hope to make this available for testing by the end of April.

2. Global option to not display borders around groups. When I make groups, the border often infringes on the letters and makes it difficult to read. If there were an option somewhere to disable all borders, it would help a lot
   

Noted.

3. Improve custom color. One specific feature would be to assign multiple amino acids a single color at the same time. For example, I made a custom color for labeling certain beta strands called "yellow." Basically everything is yellow except for histidines. TO make this, I had to click on each amino acid and then click on the yellow color which was somewhat tedious. It would have been nice to select everything and define it as yellow and then histidines as blue.
   

Again, noted. The rendering style controls are all due for an overhaul, but I may not get around to this until September.

4. Improve print options and add print preview. It would be nice to have a feature for printing where Jalview would decide how small to make the print to get everything to fit. This could be a vertical fit or a horizontal fit option
   

This is definitely a +1 for improving the print facilities - I made some similar suggestions myself in a posting back in February (http://www.compbio.dundee.ac.uk/pipermail/jalview-discuss/2010-February/000291.html).

5. Allow editting multiple names at once in a sequence. I have a list of GI numbers and their aligned sequences in Jalview and to some gorups i would like to add "1-" or "2-" to all of the seuqneces selected. This could be great for keeping track of which group of sequences are in a group. Ideally the Muscle scoring should keep the groups together but I would prefer to just look at the number.
   

I may come back to you on this. There's a slight difference between editing sequence names (which I agree calls for some kind of search/replace box with regex capability), and jalview adding additional annotation to the sequence label (like it's position within a sequence group).

Furthermore, a work around for keeping track of groupings is to load a tree onto the alignment - jalview will order sequences according to a tree, and you can create groups based on the tree very easily. Until its easy to cut and paste other types of properties (e.g. sequence ordering, groups, etc) between different alignments, this is the only way to do this, I'm afraid.

6. Add and align sequences without destroying groups. I label my groups, then decide I want to add one more sequence. I wish there was a way to add a sequence and align it without ruining all my hard 10 minutes of work.
   

Noted. Lots of ways that this can be done though - so again, I may send you an email in a month or two about the specific way that you are working with Jalview here.

7. This suggestion is kind of out there, but I use jalview mainly for looking at groups in cytoscape. If there was any way to tie-in to the cytoscape plug in blast 2 similarity graph, that would be amazing. Of course this woul dbe more of a project than a small suggestion but it would be incredibly useful.
   

The major feature in the jalview road map is to adopt the same java plugin architecture that is used by tools like Cytoscape and Taverna. I hope to make significant steps towards this by Summer 2010 - with the result that it should be much easier to use Jalview and cytoscape components together. If you know of anyone interested in this, then please do put them in contact - since I'm not really a cytoscape user.

Also I have come across one very annoying bug.
   

:frowning: oh well.

Normaly my projects are saved with 8 groups each with the custom color I call "yellow." When I restore the project, only 1 group is actualy colored with yellow. The rest are still groups and I can see the border and color them again, but they certainly are not restored.
   

Thanks very much for spotting this. The bug is also reproducible in the development version - so I've filed it in our bug tracker.

Thanks for your time. If any of the suggestions are already features, please let me know!
   

As I've said above, I'll get back to you about some of these. I'm trying to get a release out at the moment, but after that, some of these features may make it in to the next package of development work.

Jim.

···

On 22/03/2010 21:12, Hitchcock, Daniel Stephen wrote:

--
-------------------------------------------------------------------
J. B. Procter (JALVIEW/ENFIN) Barton Bioinformatics Research Group
Phone/Fax:+44(0)1382 388734/345764 http://www.compbio.dundee.ac.uk
The University of Dundee is a Scottish Registered Charity, No. SC015096.