19andMe

Introduction

19andMe is an interactive dashboard for users to calculate their individualized risk score of contracting COVID-19 and related adverse health outcomes. https://19andme.covid19.mathematica.org/

How 19andMe Works?

19andMe takes user input on where you live, who you are, and what you do in the pandemic, uses the best available science to provide a ballpark estimation on the how likely someone with similar characteristics like you may contract COVID-19, and if infected, how likely is the outcome going to be severe. Please see the "Methods" tab in the app for more details on our methodology.

Contents

• app/
  • R Scripts for different modules of the app
  • Google Analytics plug-in
  • CSS
  • www/
    • Mathematica logo
    • How 19andMe works

Authors

• Cindy Hu - product owner
• Emma Pendl-Robinson - full-stack developer
• Erin Lipman - back-end developer and data engineer
• Jennifer Starling - validation lead
• Margaret Luo - API support and technical consultant
• Max Dulieu - API tech lead

Reviewers and Support

• George Luo - front-end developer
• Arpan Bhattacharya, George Gallo - AWS architects
• Aaron White - technical consultant
• Sean Kirk - DevOps engineer
• Kelsey Skvoretz, Fei Xing - QA reviewers
• Matt Salganik - expert reviewer

Acknowledgements

We appreciate additional support towards this project from Alex Bohl, Dave Peters, and Matt Gillingham.

Contact info

For more information, please contact Cindy Hu at CHu@mathematica-mpr.com or at covid.risk.score@gmail.com.

DockerShinyApp

Installing

This project is built to use Docker and docker-compose to make development easy across all machines and remove host machine configuration as a potential issue. Use docker-compose to get started quickly. You will need to install docker and docker-compose.

• Docker
• Docker Compose

Use docker-compose to build the image:

docker-compose build

Remember that docker images are immutable once built. Only changes to files in /home/rstudio will persist after restarts.

Development

To create an RStudio environment preloaded with all dependencies

1. Create a .env file that defines the desired RStudio password, e.g.

   PASSWORD=mystrongpassword

   There is an example sample.env you can also use. Copy it and rename it to .env.

2. Email the Mathematica Communications team to get an x-api-key. We will use this to make POST requests for the covid-risk-score-api. Add X_API_KEY to your .Renviron file.

3. Start the environment docker-compose up

4. Visit http://localhost:8787 and start hacking.

5. Keep shiny app code in app.R. Launch it for development with shiny::runApp('app.R').

Deployment

Utilize the attached Dockerfile for a simple but efficient deployment setup.