- Simple - Analyze variants in a simple to use web interface.
- Aggregation - Combine results from multiple analyses and VCFs into a centralized database.
- Collaboration - Write comments and share cases between users and institutes.
This README only gives a brief overview of Scout, for a more complete reference, please check out our docs: www.clinicalgenomics.se/scout.
git clone https://github.com/Clinical-Genomics/scout
cd scout
pip install --requirement requirements.txt --editable .
Scout PDF reports are created using Flask-WeasyPrint. This library requires external dependencies which need be installed separately (namely Cairo and Pango). See platform-specific instructions for Linux, macOS and Windows available on the WeasyPrint installation pages.
You also need to have an instance of MongoDB running. I've found that it's easiest to do using the official Docker image:
docker run --name mongo -p 27017:27017 mongo
Once installed, you can setup Scout by running a few commands using the included command line interface. Given you have a MongoDB server listening on the default port (27017), this is how you would setup a fully working Scout demo:
scout setup demo
This will setup an instance of scout with a database called scout-demo
. Now run
scout --demo serve
And play around with the interface. A user has been created with email [email protected] so use that adress to get access
To initialize a working instance with all genes, diseases etc run
scout setup database
for more info, run scout --help
If you intent to use authentication, make sure you are using a Google email!
The previous command setup the database with a curated collection of gene definitions with links to OMIM along with HPO phenotype terms. Now we will load some example data. Scout expects the analysis to be accomplished using various gene panels so let's load one and then our first analysis case:
scout load panel scout/demo/panel_1.txt
scout load case scout/demo/643594.config.yaml
Scout may be configured to visualize coverage reports produced by Chanjo. Instructions on how to enable this feature can be found in the document chanjo_coverage_integration.
Scout needs a server config to know which databases to connect to etc. Depending on which information you provide you activate different parts of the interface automatically, including user authentication, coverage, and local observations.
This is an example of the config file:
# scoutconfig.py
# list of email addresses to send errors to in production
ADMINS = ['[email protected]']
MONGO_HOST = 'localhost'
MONGO_PORT = 27017
MONGO_DBNAME = 'scoutTest'
MONGO_USERNAME = 'testUser'
MONGO_PASSWORD = 'testPass'
# enable user authentication using Google OAuth
GOOGLE = dict(
consumer_key='CLIENT_ID',
consumer_secret='CLIENT_SECRET',
base_url='https://www.googleapis.com/oauth2/v1/',
authorize_url='https://accounts.google.com/o/oauth2/auth',
request_token_url=None,
request_token_params={
'scope': ("https://www.googleapis.com/auth/userinfo.profile "
"https://www.googleapis.com/auth/userinfo.email"),
},
access_token_url='https://accounts.google.com/o/oauth2/token',
access_token_method='POST'
)
# enable Phenomizer gene predictions from phenotype terms
PHENOMIZER_USERNAME = '???'
PHENOMIZER_PASSWORD = '???'
# enable Chanjo coverage integration
SQLALCHEMY_DATABASE_URI = '???'
REPORT_LANGUAGE = 'en' # or 'sv'
# other interesting settings
SQLALCHEMY_TRACK_MODIFICATIONS = False # this is essential in production
TEMPLATES_AUTO_RELOAD = False # consider turning off in production
SECRET_KEY = 'secret key' # override in production!
Starting the server in now really easy, for the demo and local development we will use the CLI:
scout serve --config ./config.py
When running the server in production you will likely want to use a proper Python server solution such as Gunicorn. This is also how we can multiprocess the server and use encrypted HTTPS connections.
SCOUT_CONFIG=./config.py gunicorn --workers 4 --bind 0.0.0.0:8080 --access-logfile - --error-logfile - --keyfile /tmp/myserver.key --certfile /tmp/server.crt wsgi_gunicorn:app
The
wsgi_gunicorn.py
file is included in the repo and configures Flask to work with Gunicorn.
Starting from release 4.4, Scout offers integration for patient data sharing via MatchMaker Exchange. General info about MatchMaker and patient matching could be found in this paper.
For a technical guideline of our implementation of MatchMaker Exchange at Clinical Genomics and its integration with Scout follow this link.
A user-oriented guide describing how to share case and variant data to MatchMaker using Scout can be found here.
TODO.
If you want to contribute and make Scout better, you help is very appreciated! Bug reports or feature requests are really helpful and can be submitted via github issues. Feel free to open a pull request to add a new functionality or fixing a bug, we welcome any help, regardless of the amount of code provided or your skills as a programmer. More info on how to contribute to the project and a description of the Scout branching workflow can be found here.