Shiny application for sharing mass spectrometry data in Bryjalab

This repository contains instructions on how to run a shiny application, developed to facilitate the inspection of mass spectrometry (MS) datasets produced in Bryjalab.

The shiny application is composed of 3 modules:

• 'Explore dataset' module allows visual inspection of the data in an interactive table, visualization of selected contrasts in a volcano plot and optional gene ontology analysis on either upregulated or downregulated proteins
• 'Find protein' module allows searching for protein of interest across all available datasets and its associated statistics
• 'Description' module provides an information how the dataset was produced in our lab and on the proteomics core facility

The application was written in R, version 4.3.1, using Shiny library, version 1.7.5. The web version of application can be accessed only from the MUNI IP address range for the security reasons - for this reason, also data (containing statistical results of contrasts of interest) are stored within a private repository.

The application can be run also locally, using the model dataset, which is a part of already published manuscript.

Quick start (for Bryjalab)

Important Docker image needs to be built locally, it won't work with the virtual instance!! If any changes in app.R, data-preparation.R or updateGeneNames.R scripts are done, docker image needs to be built again.

Docker image can be built (locally using PowerShell) with: docker build -t kristinagomoryova/app:1.0.1 . (adjust the versioning!)

docker push kristinagomoryova/app:1.0.1

Then on the VM:

docker pull kristinagomoryova/app:1.0.1

If you change any of the scripts, don't forget to do also git pull to update the repository.

To add new dataset:

1. Copy the dataset in the VM \database folder (e.g. using mc commander)
2. Adjust the metadata.csv file in the \database folder (add new record, it's not needed to run metadata.R)
3. Run the updateGeneNames.R script using Rscript updateGeneNames.R. Gene names in all files will be automatically updated.
4. Run the data-preparation.R script using Rscript data-preparation.R
5. Run the docker run -p 3838:3838 -v /home/ubuntu/MS_shiny_bryjalab/database/:/app/database/ kristinagomoryova/app:1.0.1. Pay attention to backslashes and versioning of the docker image.

How to run the application

Running the model data:

Currently the repository contains two metadata.R scripts: metadata.R containing information about datasets produced in Bryjalab (which are private), and metadata_model.R, which is intended to be used to run the model data. In case own data would be added, change the metadata information accordingly.

Running the shiny app using RStudio

1. Clone this github repository locally
2. Copy the datasets to be displayed into the /database folder (be sure you added them also to the database/metadata.R). Model data are already in the /database folder.
3. Run the app.R script

Running the shiny app using Docker

1. Clone this github repository locally
2. Copy the datasets to be displayed into the /database folder (be sure you added them also to the database/metadata.R). Model data are already in the /database folder.
3. Mount the data to the docker image, using a similar command as e.g.: docker run -p 3838:3838 -v /home/ubuntu/MS_shiny_bryjalab/database/:/MS_shiny_bryjalab/database/ kristinagomoryova/app
4. Run the shiny application, either as http://localhost:3838, or http://public_IP:3838

Adding a new dataset

New dataset should be added to the /database folder. It is preferred, that upon addition of a new dataset, all datasets would get updated gene names in order to synchronize them. A model dataset, 2006_RNF43.csv, related to publication by Radaszkiewicz et al.(2021) is provided.

Gene names update can be done running the updateGeneNames.R script, which, however, requires installation of R, and specific packages, dplyr, here and HGNCHelper.

This can be done using commands: sudo apt install r-base-core sudo apt-get install r-cran-dplyr R install.packages("HGNCHelper")

The gene names update script can be run in the terminal using Rscript updateGeneNames.R

Don't forget to update the metadata.R and also metadata.csv produced by the R script to include new dataset!

The new dataset will be displayed automatically upon refreshing the application in the browser.