Data and Code to Reproduce Results in "Effects of assortative mixing and sex-traits on male-bias in tuberculosis: A modelling study"

Authors:

• Paige B. Miller
• Chris C. Whalen
• John M. Drake

Summary:

Globally, Tuberculosis disease (TB) is more common among males than females. Recent research proposes that differences in preferential social mixing by sex, or sex-assortativity, can alter infection patterns in TB. We conducted a simulation study to see whether sex-assorted mixing patterns can explain the global ratio of male:female TB cases and what factors might cause sex-disparities in infectious diseases to be sensitive to assortative mixing. Simulations showed sex-assortativity alone cannot cause sex-bias in TB. However, we find an effect of interaction between assortativity and sex-traits that suggests a role for behaviour to influence sex- specific epidemiology of infectious diseases. In our study, the role of sex-assortativity was especially apparent for slower spreading infectious diseases, like TB. We also examined how assortativity and sex-traits affect the final outbreak size and other epidemic dynamics. These results are important for understanding when sex- assortativity, a common feature across human populations, can change epidemiological patterns.

Contents:

• main.pdf: Current draft of manuscript
• supp.pdf: Current draft of supplementary material and figures
• random_modular_generator_variable_modules.py: code from Sah et al. 2014 [1] used to model assorted networks in main text
• rewire_nets.R: Rewiring algorithm used to generate assorted networks, results shown in supplementary materials
• run_simulations.py: main script used to simulate disease spread on Sah networks & rewired networks, extract simulation information, save results as csv's
• SLIRS_tau.py: script used to model outbreak size as a function of varying transmission rates (tau)
• analysis/analysis.Rmd: uses data from run_simulations.py and SLIRS_tau.py to generate figures in /male-bias-figs folder
• analysis/rewired-networks: simulated rewired networks used in simulations
• analysis/sah-networks: simulated sah networks used in simulations
• analysis/SLIRS-res: files used to generate figures in main and supplementary materials

Instructions to reproduce results:

• To rerun full set of network simulations in main text, run rewire-nets.R and run_simulations.py. Results were obtained with R Version 4.0.0 and python version 2 on a linux machine with 32 cores, 188G of memory, and a ATI Mobility Radeon HD 5430 graphics card. Users may wish to decrease the number of simulations and number of cores to produce a subset of results.
• Figures can be reproduced without rerunning network simulations with data stored in SLIRS-res and code in analysis.Rmd. Required packages and versions used to produce results are listed at the top of .Rmd file.

[1] Sah P, Singh LO, Clauset A, Bansal S. Exploring community structure in biological networks with random graphs. BMC Bioinformatics. BioMed Central; 2014;15(220).