ClinicalTrialsDataAnalyzer

Bachelor’s Thesis - Web-App

Summary

ClinicalTrialsDataAnalyzer is a webapp which I developed for my bachelor thesis. The aim was to extract, consolidate and aggregate data from ClinicalTrials.gov in order to provide better information to clinicians or clinical trial designers about which parameters are best to choose for each clinical trial.

Folder structure

'.\backend' for flasks-server and NER-processing '.\frontend' for nuxt.js files '.\training' for training data, model and the ipython-notebook for spacy training

Installation

1. Download and extract the package provided on the repository, tagged 1.0.
2. Install Python/Anaconda distribution
3. Install revelant python dependancies via the console: 'pip install flask flask-cors spacy spacy-annotator'
4. Download relevant model for spacy (for further custom training) with 'python -m spacy download en_core_web_sm'
5. Install Node.js
6. Navigate to /frontend/ and install all needed dependancies via 'npm install'
7. After the installation run 'npm audit fix' if vulnerabilities are shown

Launching the App in Dev-Mode

1. Navigate to /backend/ and start the server with 'flask run'
2. Navigate to /frontend/ and start the node-server with 'npm run dev'
3. Open localhost:3000 on your browser and use the app

Hosting for production

1. For the flask - server follow the following instructions from the official-doc
2. For Nuxt buid the frontend with 'npm run build' and then start with 'npm run start' to deply the server-side version of the frontend

Disclamer

The data used in this software is publicly available and has been provided by the following providers:

1. ClinicalTrials.gov via their API
2. Clinical Tirals Transformation Initiative(CTTI) for the AACT-database