Skip to main content
Apps

What apps are

Concept

Apps are the tools you run inside a session. Here is what they are and how they stay contained.

In short

Apps are pre-approved tools that run inside a session. You do not install anything - you add the ones you need from the App Store and open them in your workspace.

What apps are​

An app is a software tool you use to work with data inside CHORUS - Jupyter, RStudio, VS Code, and tools specific to a research field. Apps run inside a session, on the platform's computers, on the data in your workspace.

You launch them from the App Store, the catalogue of what is available on your instance.

They run on the platform, not on you​

The app runs inside the workspace. What reaches your browser is its display, streamed to you - not the data. Nothing is installed on your machine, and closing your laptop does not stop the work.

This is why an app can work on a dataset far larger than your computer could hold, and why sensitive data stays contained while you use it.

What is actually available​

The catalogue spans the whole arc of a study - finding a cohort, preparing the data, analysing it, writing it up, publishing it.

What follows is a selection of what is commonly used, not the full list. The catalogue grows, and each institution decides what to enable on its own instance - so treat this as a sense of the range rather than an inventory. The App Store on your instance is the authoritative list.

Notebooks, code and writingWhere the analysis actually gets written - and written up.
JupyterLabNotebooks for interactive analysis in Python, R and more.
RStudioThe statistical computing environment, as you know it.
VS CodeA full code editor for scripts and pipelines.
SciterminalA terminal, for when you would rather type.
OnlyOfficeDocuments and spreadsheets, without leaving the environment.
Field-specific toolsWhatever a research domain needs - here, neuroimaging and electrophysiology.
BrainstormAnalysis of MEG, EEG and intracerebral signals.
FreeSurferCortical surface reconstruction and segmentation.
FSLThe neuroimaging analysis toolbox.
ITK-SNAPSegmentation and navigation of 3D medical images.
HiBoP3D visualisation of intracerebral data.
BTVReplayViewer for intracerebral EEG recordings.
LocalizerLocalisation of implanted electrodes.
Preparing and standardising dataGetting a dataset into shape - and stripped of identifiers - before analysis.
BidsificatorOrganise a dataset into the BIDS standard.
TRCAnonymizerAnonymise Micromed recordings.
ARXAnonymisation of structured data.
Privacy ToolboxDe-identification of sensitive datasets.
File ManagerMove and organise files inside your workspace.
Finding and publishingLocating the cohort at the start, and sharing the result at the end.
HORUS ExplorerExplore patient cohorts and check feasibility.
HORUS RestitutionRetrieve the data an approved request covers.
ZenodoPublish and archive datasets with a citable identifier.
Chorus AssistantHelp with using the platform itself.

What this changes​

Read that list again as a working day rather than a catalogue. Cohort exploration, de-identification, BIDS conversion, imaging, statistics, writing, publishing - normally spread across a departmental server, a personal laptop, a licensed workstation down the corridor, and whatever the collaborating hospital uses.

Here they sit in one environment, on the same data, in the same browser tab. Which removes a set of obstacles most researchers have stopped noticing:

  • No installing, and no admin rights to request. The tool is already prepared and approved.
  • No moving data between tools. They all see the same workspace storage, so there is no export-import shuffle between one step and the next.
  • No specification limits. The analysis runs on the platform's compute, not on your laptop.
  • No waiting for a licence or a machine. Whatever is in the catalogue is available now.
  • No copies to keep track of. One dataset, in one place, with one set of rules.

The point is not that any individual tool is unusual - most of them you already know. It is that the chain between them is gone.

Versions live side by side​

The same tool is usually available in more than one version, and you choose which to launch.

That matters more than it first appears. An analysis run with one version of a statistical package may not reproduce identically under the next. Being able to launch the version you originally used - months or years later - is what makes the work defensible.

Apps run with restricted privileges​

Every app runs as an ordinary user, never as an administrator, with system-level privileges removed.

In practice: you can install packages into your own working area as usual, but you cannot change the underlying system. If a tool insists on administrator rights to work, it needs to be packaged differently rather than granted them - which is a conversation with the platform team.

Can any tool be added?

In principle yes. There is no restriction on the language, framework or runtime - anything that can be packaged as a container image is a candidate, which covers essentially the whole scientific software landscape.

What decides it in practice is not technical but practical: someone has to package the tool, it has to run without administrator privileges, and its licence has to permit use on the platform. Commercial tools requiring a licence server are the usual sticking point.

If you need something that is not listed, ask - see The App Store.

Where apps come from​

Available apps live in the App Store, and everything in it has been prepared and approved to run inside the secure environment. See The App Store.

<- Previous
Resources & lifecycle
Next ->
Launch an app