Code for: "Infections are not alike: the effects of covariation between individual susceptibility and transmissibility on epidemic dynamics"

Jeremy D Harris, Esther Gallmeier, Jonathan Dushoff, Stephen J. Beckett, Joshua S. Weitz (2024)

This repository contains all the MATLAB codes for both running & plotting results from SIR-like models incorporating both susceptibility and transmissibility variation. Variation is included in susceptibility ($\varepsilon$) and transmissibility ($\delta$) and compared to models with variation in susceptilibility alone and the standard SIR model, i.e., homogeneous susceptibility and transmissibility. We implement the following initial joint distributions in $\varepsilon$ and $\delta$:

1. Uncorrelated gamma distributions;
2. Low- & high-variance truncated Gaussian distributions.

A preprint of the manuscript can be found on medRxiv. Please cite our work as:

Harris J.D., Gallmeier E., Dushoff J., Beckett S.J., Weitz J.S. 2024. Infections are not alike: the effects of covariation between individual susceptibility and transmissibility on epidemic dynamics. medRxiv doi: 10.1101/2024.10.11.24315334

Instructions:
MATLAB 2023b and 2024a was used to run model simulations and plot figures. We also use the Statistics and Machine Learning Toolbox for setting bivariate distributions.

First, downloaded the Github repository. To plot the figures found in the manuscript, navigate to the subdirectory 'code_plt.' The name of the MATLAB script file (*.m file) can be found in the tabulation file: 'tabulatefigures.md' in the main directory. Open the appropriate script file in MATLAB. To reproduce the simulation data, navigate to the subdirectory 'code_sims' and run the appropriate function. See again, the tabulation file: 'tabulatefigures.md' in the main directory.

See below for subfolder descriptions.

Folder descriptions:

• code_plt: All code to plot model simulation results.
  

  Read in the data from 'data/' and plot the figures in the manuscript. At the top of these files, there is an option to save the figure or not. If 'save_ans'= 1, the figures will save to the folder 'figures,' a subdirectory of the main directory. These are the figures included in the manuscript.

• code_sims: All code to simulate models.
  

  At the top of these files, there is an option to save the simulation data or not. If 'save_ans'= 1, the data will save to the folder 'data,' a subdirectory of the main directory. To run and save additional results, set each of the options equal to 1.
  

  To simulate (nonzero) correlations, first set the options, 'save_distributions' = 1 and 'readin_init_joint = 0.' This will produce a transient at the beginning of the epidemic dynamics and save the distributions at time point 'index_day_distribution' = 40. Next, run the simulation again, but now setting the options, 'save_distributions' = 0 and 'readin_init_joint = 1.' This will remove the transient, because the distributions have already converged to the eigendistributions of the epidemic dynamics.

• data: Simulation data is saved here and used by plotting codes.
  

• figures: Figure files are saved to this folder.