Genetic analysis of the Zika epidemic in Colombia

Genomic epidemiology supports multiple introductions and cryptic transmission of Zika virus in Colombia.

Allison Black1,2, Louise H. Moncla2, Katherine Laiton-Donato3, Lissethe Pardo3, Catalina Tovar4, Diana P. Rojas5, Ira M. Longini5, M. Elizabeth Halloran1,2, Dioselina Peláez-Carvajal3, Juan D. Ramirez4, Marcela Mercado-Reyes3, Trevor Bedford1,2.

* These authors contributed equally to the work.

1 University of Washington, Seattle, Washington, United States. 2 Fred Hutchinson Cancer Research Center, Seattle, Washington, United States. 3 Instituto Nacional de Salud, Bogota, Colombia. 4 Universidad del Rosario, Bogota, Colombia. 5 University of Florida, Gainesville, Florida, United States.


Colombia was the second-most affected country during the American Zika virus (ZIKV) epidemic, with over 100,000 reported cases. Despite the scale of the outbreak, limited genomic sequence data were available from Colombia. We sequenced ZIKV genomes from Colombian clinical diagnostic samples and infected Aedes aegypti samples across the temporal and geographic breadth of the Colombian epidemic. Phylogeographic analysis of these genomes, along with other publicly-available ZIKV genomes from the Americas, indicates two separate introductions of ZIKV to Colombia, one of which was previously unrecognized. We estimate the timing of each introduction to Colombia, finding that ZIKV was introduced and circulated cryptically for 4 to 6 months prior to initial case detection in September 2015. These findings underscore the utility of genomic epidemiological studies for understanding epidemiologic dynamics, especially when many infections may be asymptomatic.

Instructions for reproducing the Zika-Colombia analysis

Clone this repo by running the following command in your terminal.

git clone

This analysis uses the nextstrain command line interface (CLI). Instructions for how to install and run nextstrain-cli can be found here.

After installing nextstrain-cli, change back to your zika-colombia directory.

Run the analysis with nextstrain build .

Then, visualize the analysis with nextstrain view auspice

side note In some of my figures I have a map inset that is from nextstrain, but lacks the bezier curves that normally show transmission. To make these figures I did the following.

  1. clone the auspice repo.
  2. move the files that get put into the zika-colombia/auspice into auspice/data.
  3. Comment out these lines in the mapHelpers.js script that are responsible for plotting the Bezier curves.
  4. Check out your build by running npm start from the auspice directory.
  5. Go check out the build at http://localhost:4000/local/zika.

If you want to customize your own analysis, you can do so by editing the Snakefile. The cli provides a dockerized wrapper for Nextstrain Augur and Nextstrain Auspice, so any commands that you could normally give to augur or auspice can be integrated into the Snakefile by editing shell commands and adding params. File extenstions and locations can also be edited by changing paths in the Snakefile.

Note, if you are running the analysis on a Mac, you may have to change the amount of memory that you allocate to Docker. The default can sometimes be too low, which causes mafft to fail during the augur align step. To get around this, I have raised Docker’s memory allocation from 2GB to 8GB. Docker memory allocation can be changed by going to Docker > Preferences > Advanced and toggling up the memory, then applying and restarting.

If you are wondering what commands you can give the different augur executables, you can look at augur help statements. To do so, run nextstrain shell to bring up the augur environment. Then explore with augur --help or augur <program> --help.