This tool is supposed to plot somatic mutations on their respective proteins, something that is easier said than done (at least with a nice graphical output and without twiggling too much with code. Can anybody help me on that?
unfortunately I don't have perl programming skills, so it would definitely require something more than 4 hours x me. Any suggestion as an alternative which is not "Take a perl course?" ;-)
I was able to successfully produce MuSiC lolliplots by downloading the source code from github (git clone https://github.com/genome/gms-core.git) and hacking into MutationDiagram.pm. This way I was able to decouple the MutationDiagram module from its required input data structures and plug in my own data (InterPro domains, genomic variants). This worked well for me without too much hassle (~4 hours of work), but it definitely requires good Perl skills and some understanding about the function of this module.
Glad to hear that you got it working, Christian. If you have a stand-alone version that you'd like to contribute back to the community, we'd be happy to put it up on our site. (with proper credit given to you, naturally). One of our motivations for open-sourcing all of our code is to enable people to use our tools - even the ones we haven't had time to package up neatly yet! Even rough code might give someone else a cleaner place to start.
I would have gladly provided my code if it is in a useful state. But as it is now, I only substituted one internal data structure (MuSiC) for another (my own), which is hardly more useful than the current implementation of MutationDiagram.pm.
What is needed is some kind of standardized interface to this module that allows users to run it off their own input text files without the need to touch the code (e.g., by reading genomic variants in VCF format, gene models from GFF/BED format, and domains in HMMer format). Maybe I will invest this extra work and let you know.
As the answer linked above suggests, the lollipot code is heavily dependent on our internal databases. The code is all open-sourced on github, but getting it to run is currently non-trivial, as it hasn't been fully packaged up for stand-alone use yet. The TVAP team hopes to do so soon, which is why a placeholder page is sitting on that site. Sorry for the confusion!
This might help - Music Proximity Analysis Mutation Diagrams
unfortunately I don't have perl programming skills, so it would definitely require something more than 4 hours x me. Any suggestion as an alternative which is not "Take a perl course?" ;-)