Skip to content

Latest commit

 

History

History

LCMSDataVisualisation

README.md

This module lets you setup a visualization server that allows comparative display of UntargetedMetabolomics analysis.

Requirements

R version > 3.1

Instructions

Get a fresh version of the reachables project

  • If needed, clone a new version of the act repository
mkdir lcms-viz 
git clone [email protected]:20n/act.git lcms-viz/
  • Sync your lib directory as instructed here reachables/lib and install the Chemaxon license
cd lcms-viz/reachables/ 
rsync -azP LOC_LIB lib/ 
license CHEMAXON_LICENSE.cxl

At this point, you should be able to fully compile your SBT project. Try performing a clean compilation and running tests on the project: sbt clean compile test.

Link to outside resources

The app needs to access outside resources, including the 20n logo, but more importantly some methods in the reachables project, such as the MS1 class for m/z values computation. By creating symlinks in the app's directory, we provide the path to these resources.

  • Create a fat JAR with SBT
sbt assembly
  • Create symlinks in the app directory:
cd src/main/r/LCMSDataVisualisation/
ln -s lcms-viz/reachables/target/scala-2.10/reachables-assembly-0.1.jar reachables-assembly-0.1.jar
ln -s lcms-viz/reachables/src/main/resources/20n.png 20nlogo
  • Finally, make sure that the directory data/mol-structure-cache/ exists. This is where molecule structure images will be stored. TODO: add a symlink for this too!

Get R dependencies and install Scala through R

  • Install libraries netCDF and BOOST:
sudo apt-get install libnetcdf-dev
sudo apt-get install libboost-all-dev

You might have to upgrade to the latest version of R (e.g., in case "mzR" below fails). Search for "r install ubuntu" and follow instructions.

  • Start R and install the required packages
sudo R

In the R console:

R.Version() # Should be > 3.1
install.packages(c("shiny", "rscala", "dplyr", "plot3D", "classInt", "jsonlite", "logging", "digest"))
library(rscala) # loads the "rscala" library
rscala::scalaInstall() # downloads and installs Scala
# Package "mzR" needs to be installed through the bioconductor package
source("https://bioconductor.org/biocLite.R") # try replacing http:// with https:// if it gives you an error
biocLite("mzR")

Start the server

In the R console:

library(shiny)
runApp(port = 9090, host = "0.0.0.0", launch.browser = FALSE)

The app should be now accessible from http://hostname:9090

Test using demo UntargetMetabolomics file

Try uploading the lcms-demo.json file for sample visualization.