Mangul Lab joins urgent efforts to track COVID-19 by zipcode
Lana Martin April 08, 2020
Mangul Lab joins urgent efforts to track COVID-19 by zipcode
Members of Mangul Lab have joined efforts with UCLA researchers to curate data that will help hospitals—already burdened by the COVID-19 pandemic—predict patient load in coming weeks.
For the past two weeks, Jaque (Mangul Lab postdoc), Jeremy (software engineer), and Niko (undergraduate researcher) have been working with Rob Brown (UCLA Computational Medicine postdoc) to find, format, and post datasets that promise to help researchers develop a better understanding of how COVID-19 spreads.
The goal of their project to predict infection growth rates at the country and zip code level. Predicting the growth rate of COVID-19 cases will help scientists understand how the disease spreads and to forecast accurate predictions. With resources stretched thin, counties and hospitals can leverage accurate prediction to better allocate resources and schedule necessary non-COVID related procedures.
Jaque plays a central role in creating and maintaining the project’s github repo as a part of the BigBio group. In order to ensure the repo is intuitive to understand, Jaque is developing formatting guidelines so that researchers may quickly and efficiently use curated data.
Jeremy and Niko are finding, formatting, and uploading new data sources that may build robust predictions of infection growth rates. Data speaking to routine human behavior can be used to build predictive epidemiological models; they have collected data on how people travel to work, population density, household density, and standard demographic statistics for every census tract, county, and state in the United States.
Air Quality Index scores and distance to grocery markets are two covariates that may impact individuals’ susceptibility to contracting COVID-19 and local leaderships’ ability to manage citizen mobility; Jeremy and Niko are now focused on obtaining these datasets.
“For a simple example,” Jeremy notes, “a high population density would likely allow a virus to spread at a faster rate, making an impact on predicting how many patients a hospital might expect in a week.”
The platform and data our team are curating is now helping UCLA researchers to better fit SIR epidemic models with time-varying social distancing metrics and will be used in conjunction with data collected from UCLA’s new survey that aims to define city-specific trends: https://stopcovid19together.org.
COVID-19 by Country and Zip Code GitHub https://github.com/Big-Bio/COVID19byCountryAndZip
UCLA Newsroom: UCLA web app will enlist public’s help in slowing the spread of COVID-19 https://newsroom.ucla.edu/releases/ucla-web-app-slowing-spread-covid-19