Written by Sarah Cobey. I did a bit of cleanup and debugging, but this is mostly Sarah's work.

Sarah wrote this simulator so that I could confirm that Migrate could successfully infer migration patterns from sequence data taken from an evolving virus population. The results of that analysis are detailed in the "Global migration dynamics" paper.

This simulation implements a multi-strain multi-deme epidemiological model. Each strain has equivalent "fitness", i.e. cross-immunity between strains is complete. Evolution of the virus population is approximated by having previously immune hosts return to the susceptible class over time. This represents an SIRS model. Contact between hosts in different demes is determined by a matrix of per-capita contact rates. Thus, viral migration patterns emerge dynamically from host-contact patterns. Transmission rates β can be seasonally forced according to a sinusoidal function. For further methodological details, please consult the paper.

Simulation source code is provided as well as the Mathematica notebook I used to analyze the simulation output and Migrate results. If you find this useful in your work, please cite the above "Global migration dynamics" paper.

Current version: 1.15
Released 3/8/2011

Download:       source      

Join the discussion