Skip to content


Repository files navigation


Build Status codecov DOI

This is the repository for the neuroscout server.

Requirements: Docker and docker-compose.


First, set up the main environment variables in .env (see: .env.example). Set DATASET_DIR, KEY_DIR, and FILE_DATA to folders on the host machine.

Optionally, set up pliers API keys for feature extraction in .pliersenv (see: .pliersenv.example). More information on pliers API keys

Next, set up the Flask server's environment variables by modifying neuroscout/config/ and saving as neuroscout/config/

Finally, set up the frontend's env variables by modifying neuroscout/frontend/src/config.ts.example and saving as neuroscout/frontend/src/config.ts.

For single sign on using Google, a sign-in project is needed.

Initalizing backend

Build the containers and start services using the development configuration:

docker-compose -f docker-compose.yml -f build
docker-compose -f docker-compose.yml -f up -d

The server should now be running at http://localhost/

Next, initialize, migrate and upgrade the database migrations. If you have a database file, load it using pg_restore. Otherwise, delete the migrations folder, initalize the database and add a test user.

docker-compose exec neuroscout bash
rm -rf /migrations/migrations
python db init
python db migrate
python db upgrade
python add_user useremail password

Staging & production server

For the staging server, you can trigger a manual build as follows:

docker-compose -f docker-compose.yml -f build
docker-compose -f docker-compose.yml -f up -d

For the staging or production server, you can trigger pull a pre-built image from GHCR: First set the variable IMAGE_TAG to the apprioriate image tag

docker-compose -f docker-compose.yml -f docker-compose.image.yml build
docker-compose -f docker-compose.yml -f docker-compose.image.yml up -d

Setting up front end

The frontend dependencies are managed using yarn

Enter the neuroscout image, and install all the necessary libraries like so:

docker-compose exec neuroscout bash
cd frontend

You can then start a development server:

yarn start

Or make a production build:

yarn build

Ingesting datasets and extracting features

You can use commands to ingest data into the database. Run the following commands inside docker: docker-compose exec neuroscout bash

To add BIDS datasets

python add_task bids_directory_path task_name

For example for dataset ds009

python add_task /datasets/ds009 emotionalregulation

Finally, once having added a dataset to the database, you can extract features using pliers into the database as follows:

python extract_features bids_directory_path task_name graph_json

For example:

python extract_features /datasets/ds009 emotionalregulation graph.json

Even easier, is to use a preconfigured dataset config file, such as:

docker-compose exec neuroscout python ingest_from_json /neuroscout/config/ds009.json

Maintaining docker image and db

If you make a change to /neuroscout, you should be able to simply restart the server.

docker-compose restart neuroscout

If you need to upgrade the db after changing any models:

docker-compose exec neuroscout python db migrate
docker-compose exec neuroscout python db upgrade

To inspect the database using psql:

docker-compose run postgres psql -U postgres -h postgres


Once the server is up and running, you can access the API however you'd like.

The API is document using Swagger UI at:



To authorize API requests, we use JSON Web Tokens using Flask-JWT. Simply navigate to localhost:5000/auth and post the following

    "username": "",
    "password": "string"

You will receive an authorization token in return, such as:

    "access_token": "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpZGVudGl0eSI6MSwiaWF0IjoxNDQ0OTE3NjQwLCJuYmYiOjE0NDQ5MTc2NDAsImV4cCI6MTQ0NDkxNzk0MH0.KPmI6WSjRjlpzecPvs3q_T3cJQvAgJvaQAPtk1abC_E"

You can then insert this token into the header to authorize API requests:

GET /protected HTTP/1.1
Authorization: JWT eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpZGVudGl0eSI6MSwiaWF0IjoxNDQ0OTE3NjQwLCJuYmYiOjE0NDQ5MTc2NDAsImV4cCI6MTQ0NDkxNzk0MH0.KPmI6WSjRjlpzecPvs3q_T3cJQvAgJvaQAPtk1abC_E

Note that in order to use any protected routes, you must confirm the email on your account. Confusingly, you can get a valid token without confirming your account, but protected routes will not function until confirmation.

Running backend tests

To run tests, after starting services, create a test database:

docker-compose exec postgres psql -h postgres -U postgres -c "create database scout_test"

and execute:

docker-compose run -e "" --rm -w /neuroscout neuroscout python -m pytest neuroscout/tests

or run them interactively: docker.compose exec neuroscout bash python -m pytest neuroscout/tests/ --pdb

To run frontend tests run:

docker-compose run --rm -w /neuroscout/neuroscout/frontend neuroscout npm test

Running frontened tests

To run frontend tests, have Cypress 6.0 or greater installed locally. First, ensure neurscout is running:

docker-compose -f docker-compose.yml -f up -d

Next, set up the test environment:

docker-compose exec neuroscout bash

In a separate window, you can run cypress:

cd neuroscout/frontend cypress open

Once done, kill the first command, and run the following to tear down the test db

docker-compose exec -e neuroscout python teardown_test_db