GENIAL: Gene ontology ENrIchment AnaLysis

Authors: Kevin Durimel & Victor Mataigne

Affiliation: University of Rouen (France)

Web application now availaible (still a work in progress) : https://genial.shinyapps.io/GenialGSEA/

Path organization

.
│
│--bin : windows executable files
│
│--dependencies : R package dependencies list (JSON)
│
│--log : log files
│
│--example_data : experiment specific raw_data (currently --> Workflow input) 
│
│--results : results examples (one path generated per result, Day_Month_HH_MM-SS-DD_YY format)
│
│--src : R scripts

The project genesis

From a input dataset of our choice from the literature, we had 20 hours to implement R packages to realize all three GO,metabolic pathways, and protein functional domains enrichment annotations, implementing robust statistical procedures. Outputs will be both charts (pie, bar plot, graphs, tracks) and texts results (tables). This workfow must be interfaced with Shiny, a R package developed by RStudio. This script was developed under the context of a light bioinformatic project and will be not long term supported!

Shiny interface overview

Divided in 5 main sections :

• Load and explore raw data

• Vulcano plot

• GO enrichment

    GO classification (barplot)
    Ability to set a threshold for go annotations level / ontologies
    GO three browsing
  

• KEGG enrichment

    KEGG enrichment results overview
    KEGG enrichment results browsing by typing target pathway IDs on a web form
  

• Protein domains

    Hypergeometric test
    Ability to add metatada on the results :
    	Panther_Family_ID Panther_Family_Description ; EnsemblTranscriptID ; EnsemblPeptideID ; PfamID ; PfamSTART ; PfamEND
  

Shiny web application overiew

Same as the Shiny interface. Available here : https://genial.shinyapps.io/GenialGSEA/

Standalone application

This is the GUI interfaced R script (i.e Non-Shiny), containing more features that the interfaced one.

Usage (Linux/Mac OS X)

From scratch :

Rscript GENIAL-NOShiny.R

OR, set up the scripts permissions inside the src path :

chmod +x src/*

Then, add the scripts to your bashrc (/home/username/.bashrc) :

export PATH=$PATH:src/
alias Genial='Rscript GENIAL-NOShiny.R'

Then, you can run it as a shell command :

GENIAL-NOShiny.R

Dependencies

GENIAL has been developped with R 3.3.2

External packages

• shiny - Version 1.0.1
• shinythemes - Version 1.1.1
• ggplot2 - Version 2.2.1
• clusterProfiler - Version 3.2.11
• GO.db - Version 3.2.0
• org.Hs.eg.db - Version 3.4.0
• topGO - Version 2.26.0
• DT wrapper - Version 0.2
• biomaRt - Version 2.30.0
• MASS - Version 7.3-45

Be cautious!

• WARNING 1: This workflow needs some containerization (Conda, Packrat... ) to work properly / ensure repdroducibility, which was not done under the context of this very short project. You need to install exactly the same R and package versions as cited in the "external dependencies" sections. Due to this, the Web application may stop working properly at any time. This project is not long term supported and any future bugs will be not fixed.

• WARNING 2: Due to Shiny technical limitations, tasks launched in different tabs are not parallelised or processed asynchronously.

• WARNING 3: This application works only with data formated like "human_retine.txt" example file (columns names and content).

• WARNING 4: A lot of APIs are used --> The better your internet connection is, the faster the application will run.

Example dataset obtained from :

Sara R. Savage, Colin A. Bretz, and John S. Penn ; RNA-Seq Reveals a Role for NFAT-Signaling in Human 
Retinal Microvascular Endothelial Cells Treated with TNFα ; PLoS One. 2015

doi:  10.1371/journal.pone.0116941

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4305319/

---

Kevin Durimel, Victor Mataigne - M2.1 BIM - University of Rouen, 2017.

Standalone script - Web app : Kevin Durimel
Shiny interfacing : Victor Mataigne