The PVClinical project aims to build a web-based platform, built to support Active Pharmacovigilance processes via the integration of several heterogeneous data sources.
- Redis Server (used for caching)
- Postgresql Database Server (used as the Django app's database)
- MongoDB Database Server (used for the database containing openFDA data)
- Virtuoso Database Server (used for the Knowledge Graph, concerning Drugbank and MedDRA graphs)
- Shiny Server (deploying the Twitter workspace and openFDA applications)
- OpenFDA screenshots services (screenshots API for openFDA workspace)
- Docker Server (deploying the Broadsea Docker containers, for OHDSI tools and Atlas, used by OHDSI workspace)
- Apache Server (hosting and serving the produced reports and analytics/results screenshots from various workspaces)
- Twitter Account and Credentials ( to utilize the historical data Twitter APIv2, through Twitter workspace )
- Mendeley and Pubmed Accounts and Credentials ( to utilize Mendeley and PubMed APIs through PubMed workspace )
- [Optional] Mail Server (used by the auth procedures of the Django app)
Twitter Analytics, are shiny apps that utilize Twitter API v2 endpoints, to offer analytics to users, using global, real-time and historical data that are provided by Twitter, for academic research.
More info about Twitter Analytics shiny apps
OpenFDA shiny apps are based on openfdashinyapps developed by Jonathan G Levine ( https://github.com/jonathanglevine/openfdashinyapps ).
The initial apps have been adapted to use locally stored data, that have been downloaded from openFDA official repositories ( https://open.fda.gov/data/downloads/ ).
Further modifications have been made to the apps, regarding both view aspects and functionality, in order to fulfill PVClinical needs.
More info about the original apps
OpenFDA screenshots services is a special API, offering the capability to get screenshots of various openFDA-apps' views.
More info about openFDA screenshots services
Broadsea deploys the full OHDSI technology stack (R methods library & web tools), using cross-platform Docker container technology.
Broadsea Docker containers can be deployed, according to the instructions given at: https://github.com/OHDSI/Broadsea
To integrate the Broadsea docker containers into OHDSI workspace of PVClinical platform, apart from the proper variables ( the ones with OHDSI_ prefix ) that need to be set in the settings.py file of the project, before starting the containers, someone has to also substitute the default atlas directory in brodsea-webtools container, with the modified atlas directory provided on our repository, by mapping it to the respective directory in the container.
This can be done by adding a simple line to volumes subsection of broadsea-webtools section, in docker-compose.yml file.
The line should be something similar to: - ./atlas:/usr/local/tomcat/webapps/atlas:ro
[Optional] Set up a virtualenv and activate
git clone https://github.com/inab-certh/PVClinical.git
cd PVClinical
pip install -r requirements.txt
Finally, in order to run PVClinical platform, someone has to set the proper values for the variables found in gentelella/gentelella/settings.py file (a sample settings.example.py file is provided on our repository, rename it to settings.py). corresponding to the particular system configuration and setup.
cd gentelella
python manage.py runserver
Go to http://localhost:8000/
On our repository we provide a sample WSGI file ( gentelella/gentelella/wsgi.py ), to deploy PVClinical platform on a web server (e.g. Apache). More information on WSGI and ASGI interfaces, and configurations, can be found in the following link:
@bdimitriadis @Dimstella @ckakalou
Licensed under the Apache License, Version 2.0 (the "License") modified with Commons Clause Restriction -- see the LICENSE file for details