Overview

The sparrow.shiny package provides an interactive shiny applications that enables users to explore the results of a gene set enrichment analysis performed using the sparrow

Although this is a standalone package, it’s really only used as an “enhancement” to the sparrow package itself, as well as providing shiny modules to the FacileAnalysis package, too.

All that of that is to say: there isn’t much end-user stuff here to play with.

Usage

More thorough documentation of the shiny application will be provided in the near future in the form of a vignette, or more likely a screen cast.

In the meantime, this will just have to get you started:

vm <- sparrow::exampleExpressionSet(dataset = 'tumor-vs-normal', do.voom = TRUE)
gdb <- sparrow::exampleGeneSetDb()
sr <- sparrow::seas(gdb, vm, vm$design, "tumor", methods = c("camera", "fry"))
sparrow.shiny::explore(sr)

The explore function will launch the application and load it with the SparrowResult object produced by the call to the sparrow::seas() function. You can then explore the results of the “camera” or “fry” analysis through there.

Users can serialize SparrowResult objects to *.rds files on their filesystem, which can also be loaded individually once the application is launched.

Application Deployment

Analysts can simply launch the sparrow.shiny::explore() application from their workstation, however these applications can also be deployed to a shiny server.

Docker

The inst/docker folder provides examples of how to containerize and deploy this application in different contexts.

The Dockerfile-base creates a docker image that, when run, launches the shiny app on http://container.ip/sparrow (ie. http://localhost/sparrow).

ShinyProxy

The Dockerfile-shinyproxy creates an image that can be deployed via a ShinyProxy server.

Notes on setting up a ShinyProxy server on AWS are provided in the aws-ubuntu-deployment.md file.

Installation

The sparrow suite of package will soon be submitted to bioconductor and installable via the recommended BiocManager mechanism. In the meantime, these packages can be installed like so:

# install.packages("remotes")
remotes::install_github("lianos/sparrow.shiny")