Neuron Morphology Community Portal Client

The Portal Client is the user-facing front end application.

Installation and Build

Building requires Node.js 22.18 or later (the GitHub workflow uses Node 24).

Neuroglancer is installed from a pinned commit of google/neuroglancer, recorded in package.json and package-lock.json. npm install downloads and builds that commit, so no separate Neuroglancer checkout is needed. The build installs Neuroglancer's own development dependencies, which include a Playwright Chromium download; set PLAYWRIGHT_SKIP_BROWSER_DOWNLOAD=1 for the install to skip it.

• Clone the nmcp-client repository
• In the repository directory npm install

Updating Neuroglancer

• Pick a commit on the upstream master branch and copy its full 40-character SHA. A shortened SHA is not enough: npm keeps the commit already in package-lock.json and reports the install as up to date.
• npm install --save-dev github:google/neuroglancer#<full-sha> updates package.json and package-lock.json together.
• Compare the overrides in Neuroglancer's package.json at that commit with the ones here and mirror any change. npm applies overrides only from the root project, so Neuroglancer's entries (today, keeping ikonate's unused build tooling out of the install) have no effect unless copied.
• Build, run the tests and check each Neuroglancer view.

Run Local

Depending on your setup, a number of env vars for the host/port of the database, API service, etc... will likely need to be set. See the nmcp-deploy project for details on system configuration. Once a local system is generally configured:

• tsc
• npm run debug

Build Production Distribution and Docker Images

Docker image building and deployment uses Task.

To build the dist package

• task build

To build the docker image

• task docker-build

To release the docker image (requires credentials to ghcr.io).

• task release

Authentication

NMCP_AUTHENTICATION_KEY must be set to match the value used for the API service.

Precomputed Skeleton Data Source

NMCP_PRECOMPUTED_OUTPUT must be set to define the precomputed skeleton location for Neuroglancer layers. This is just the location URL, e.g., s3://mybucket/skeletons/staging. The appropriate prefix will be automatically attached as a Neuroglancer data source (e.g., precomputed://)

Google Analytics

NMCP_GOOGLE_ANALYTICS_TAG optionally sets the Google Analytics measurement ID (e.g., G-XXXXXXXXXX) used by the gtag.js snippet injected into index.html at server startup. Set it to your own tag to distinguish deployments (e.g., production vs. test instances). If it is left unset, no analytics script is included in the page.

Development Status Notes

See here for additional details on the status of the code.