WGPA combine the most recent scores developed to asses the intolerance of a gene to mutations (RVIS, EvoTol, Gene Constraints, ...) whith a set of widely used analysis (GSEA, fatmhh, gene networks, ...) in order to distinguish the genes containing disease-relevant mutations from genes containing irrelevant mutations..

The website is completely developed using Mojolicious so you can check their site for more information.

It's necessary to change the variable $rootFolder in lib/WGPA/Utils/Paths.pm to the path of the folder where you are deploying the website to.

Also it's necesary to update the database credentials on 3 files:

• lib/WGPA/DB.pm
• lib/fatmhh/fathmm.py (line 237)
• lib/GSEA/config.properties

A script is included to build a production ready version of the website:

  script/build.sh

It will create a folder dist that will contain your website excluding the tests and other unnecesary files and including the compiled GSEA runner (It's necessary to have java sdk installed for this).

If the script is not executable you can use chmod u+x script/build.sh and re-try.

To learn more about mojolicious I would recomend going to their website but just to get started mojolicious has two built in server.

Morbo is singlethereaded and good to get started and test locally. It will serve the website on port 3000.

  morbo script/wgpa

Hypnotoad is pre-forked and starts the processes independent of the terminal and is more suitable for production. It will serve the website on port 3000.

  hypnotoad script/wgpa

hypnotoad accept the followign flags:

• Stop gracefully the server: -s / --stop
• Run hypnotoad dependent to the terminal: -f / --foreground

Due to the performance issues that will produce to run the GSEA analysis on real time, the website just queue them and there is a java process that run on the background and process all the analysis.

The GSEA Runner has two dependencies that need to be added to the project: the GSEA jar file and the JDBC jar file. In our production website I use slightly modified version of the GSEA file but the normal distribution available at their website works to. Both jars should be place on lib/GSEA/libs.

For convenience there is a script to manage the GSEA Runner porcess

  script/gsearunner.sh -start
  
  script/gsearunner.sh -stop
  
  script/gsearunner.sh -restart

If the script is not executable you can use chmod u+x script/gsearunner.sh and re-try.

• Gene Set Enrichment Analysis (GSEA)
• Functional Analysis through Hidden Markov Models (fathmm)
• PolyPhen-2

• WGPA (available upon request)
• fathmm (Available on fathmm GitHub)
• PolyPhen-2 (PolyPhen-2 depends on several databases but we actually use their servers to run the analysis)
• STRING (Available on STRIGN web) (Actually we don't use the STRING database but a table extracted from it that's contained in the WGPA Database)