https://www.biorxiv.org/content/early/2019/01/06/512152
This application has been tested on R 3.4.4, OSX 10.14 / Windows 7/8/10, but should be supported on any platform that is supported by R and Shiny.
R can be downloaded from the main R Project page or downloaded with the RStudio Application
This application uses some libraries, but is set up to download them on the first run.
For a complete list of the dependencies of this application, view global.R.
The easiest way to install and run is: shiny::runGitHub('DO-MS', 'SlavovLab')
Or, install this application by downloading it from the GitHub page as a .zip archive, or from a bundle on the release page.
Alternatively, if you are comfortable from the command line and have git installed:
git clone https://github.com/SlavovLab/DO-MS
You will also need to install the shiny library via. R. In R, run:
install.packages('shiny')
We are in the process of submitting this application to the CRAN for easier installation. Stay tuned!
The easiest way to run the app is directly through RStudio, by opening the DO-MS.Rproj Rproject file
and clicking the "Run App" button at the top of the application.
You can also run the application by running the script start_server.sh (OSX/Linux), or start_server.bat (Windows). This option is less desirable as some of the dependencies bundled by RStudio have to be found or loaded manually (see pandoc requirements)
For OSX/Linux, Rscript must be available on the path, and for Windows, you will need to edit the start_server.bat script to point to the specific Rscript.exe executable.
If you are having trouble loading shiny, please confirm that you have installed the package in the "Installation" step. If so, then you may need to define a new environment variable, R_LIBS_USER, that points to the library path of your R installation. See this StackOverflow answer for more details on this issue.
Also, some users have reported issues with the package opening within RStudio's viewer pane, rather than within a separate browser window. Please make sure that you have the appropriate option checked in the dropdown menu of the 'Run App' button, if using RStudio.
You can automatically generate PDF/HTML reports without having to launch the server by running the do-ms_cmd.R script, like so:
$ Rscript do-ms_cmd.R config_file.yaml
This requires a configuration file, and you can find an example one here.
The UI for this application is generated dynamically from the contents of the modules folder. Users can add their own modules to this folder, following the template provided in modules/__template.R. Modules can be disabled by appending two underscores to the filename, e.g., disable_module.R --> __disable_module.R.
Currently modules are limited to plotting and providing data for download. More features, such as custom inputs are forthcoming.
As this application requires a large amount of computational power (CPU), on-hand memory (RAM), and possibly the storage of large amounts of mass-spec data (Storage), we do not recommend running this on a standalone server, as it is not cost-efficient.
If you wish to host a server for internal usage, i.e., within an organizational intranet, you can change the host IP to "0.0.0.0" instead of "127.0.0.1", which exposes the server outside of the machine itself.
This application is currently maintained for MaxQuant >= 1.6.0.16. Adapting this application to other search engines is straightfowards but does require some code editing. Please see this document on implementing other search engines for more detailed instructions.
For any bugs, questions, or feature requests, please use the GitHub issue system to contact the developers.