iMotifs doesn't open output from NestedMica

[Niall]

Hi,

I'm trying to view the output from NestedMica and the iMotif program
is the only one I can find. However, it doesn't seem to open the file
linked to below. Actually I've run about 20 sets of proteins through
and can't get any of them open. I hope someone is still listening to
this mailing list.

http://sgenomics.org/~haslam/TrgErKdel.xms

Thanks,

Niall.

[Matias Piipari]

Hi Niall!

iMotifs doesn't unfortunately currently support visualising protein motifs. I'll be sure to put this feature in the next release! Could you maybe point me at a colour coding of the different amino acids that you would like? 

Best wishes

Matias

- - - - - - - - - - -
Matias Piipari
twitter: http://twitter.com/mz2

[Niall Haslam]

Hi Matias,

Thanks for getting back to me. I think the link below shows fairly standard ones:

http://www.bioinformatics.nl/~berndb/aacolour.html

But its worth checking clustalX for other ones as well. I think having a few different ones and letting the user select is a good option.

By the way - I just realised we've met before. I came to work with Andreas a few years ago for a week or two. Small world!

Thanks,

Niall.

[Matias Piipari]

Hi Niall

I've just added preliminary support for protein motifs to iMotifs. I'm
not sure if the motifs you sent have very low information content or
if my rendering routine might scale everything down a bit too much
(have a look at the attached screenshot) -- I have a feeling there's
an assumption somewhere of scaling columns assuming max 2 bits of
info, which doesn't apply to aminoacid alphabet of course. Will
investigate...

M

[Niall Haslam]

On Thu, Nov 26, 2009 at 10:26 AM, Matias Piipari <matias....@gmail.com> wrote:

Hi Niall

I've just added preliminary support for protein motifs to iMotifs. I'm not sure if the motifs you sent have very low information content or if my rendering routine might scale everything down a bit too much  (have a look at the attached screenshot) -- I have a feeling there's an assumption somewhere of scaling columns assuming max 2 bits of info, which doesn't apply to aminoacid alphabet of course. Will investigate...

Hi Matias,

They may have low information content - but I would still want to render them. I'm trying to replicate results from a recent paper using Nmica and can't seem to find a decent viewer that allows me to investigate the results for protein motifs.

The picture you sent is already an improvement on MotifExplorer which is the other tool I tried to use to view protein motifs.

Thanks,

Niall.

M





M

On 25 Nov 2009, at 18:43, Niall wrote:

[Matias Piipari]

Hi Niall

Here's an experimental v0.2.5 beta build that includes the protein motif XMS file reading capability that I showed in the earlier message:

http://cloud.github.com/downloads/mz2/imotifs/iMotifs-v0.2.5-EXPERIMENTAL.zip

I won't push this build to the automated update feed before figuring out how to best handle the protein motif column scaling issue, and disabling nucleotide motif specific functionality from sequence set documents containing protein motifs (there's some analysis tools that only make sense for DNA motifs). 

You can see the motifs a bit better by repeatedly pressing Cmd + Up Arrow to scale up the motif column height. Cmd + Right Arrow increases the column width.

There's more help material here: http://wiki.github.com/mz2/imotifs . Would be very happy to get any feedback, ideas and feature requests. 

Also, could you perhaps provide me with an example protein motif XMS file that I could include in the installation package as an example for new users to try the app with?

Best wishes

Matias

- - - - - - - - - - -

Matias Piipari

twitter: http://twitter.com/mz2

[Niall Haslam]

Hi Matias,

This looks really interesting. Thanks very much for the speed of response - its very impressive.

I'll get back to you about a sample xms file. I guess it would have to one that looks good in the viewer by default. I'll have a think about it. Once the protein motif stuff is a bit more stable I will direct people to this software and encourage them to use it. I'd be interested in creating some kind of standard format for motifs. There has been some discussion of this already - though it hasn't yet mentioned profile based methods.

Niall.

[Matias Piipari]

On 26 Nov 2009, at 18:24, Niall Haslam wrote:

Hi Matias,

This looks really interesting. Thanks very much for the speed of response - its very impressive.

Thanks!

I'll get back to you about a sample xms file. I guess it would have to one that looks good in the viewer by default. I'll have a think about it. Once the protein motif stuff is a bit more stable I will direct people to this software and encourage them to use it. I'd be interested in creating some kind of standard format for motifs. There has been some discussion of this already - though it hasn't yet mentioned profile based methods.

We've recently submitted a paper about iMotifs and the XMS file format to a certain bioinformatics journal :-) In addition to the viewer application itself we're trying to encourage the use of a standard motif file format for position weight matrix data (XMS). The source repository at http://github.com/mz2/imotifs actually contains IO APIs for reading and writing XMS formatted files in Perl, Ruby and R, and BioJava already supports for XMS. So that should be a good basis for standardising. More documentation about those APIs is going to appear in near future, and we should of course make ready made packages available for each of them in addition to the source code repository.  

WRT supporting other types of motifs, the XMS format could be extended to contain other types of motifs to just weight matrices and in fact the sequence motif renderer in iMotifs is also pretty pluggable. Happy to talk more about this.

Matias

[Niall Haslam]

Hi Matias,

Any idea when the paper will be coming out? I and others would be interested in this.

Niall.

[Matias Piipari]

Once we've addressed the reviewer comments :-) 

The 0.2.5 release with protein motif support and a number of other smaller bug fixes is coming up very soon, will announce it here on the list too.

Matias

[Matias Piipari]

Hi Niall

The iMotifs paper has now been accepted to Bioinformatics. More info will follow soon. The release 0.2.5 that's currently the newest public one, contains initial protein motif support. Would be happy to hear feedback regardings this! 

I'm pushing a new release 0.2.6 with a few more error situations handled that I noticed. Again, very much appreciate all bug and especially crash reports!

Matias

[Niall Haslam]

Hi,

I'm actually teaching a course partly on motifs. Unfortunately, I've finished that section already. But I run it on an anual basis so I'll be sure to mention iMotifs for the next version. Now that the offical version has support for protein motifs I'll be sure to mention the program to the others that I work with. Hopefully that will increase the number of potential bug finders.

Thanks for your help, and I'll get back to you once the paper comes out.

Niall