This repository houses code developed for the analysis presented in the scientific publication "Seasonal influenza: Modelling approaches to capture immunity propagation" by Edward M. Hill, Stavros Petrou, Simon de Lusignan, Ivelina Yonova, and Matt J. Keeling.
Publication details: Hill et al. (2019) Seasonal influenza: Modelling approaches to capture immunity propagation. PLOS Computational Biology. 15(10): e1007096. doi: 10.1371/journal.pcbi.1007096.
Please find below an explainer of the directory structure within this repository.
Population mortality data (from ONS)
Using RCGP data, compute cumulative ILI cases (sum of weekly ILI rate per 100,000) for 2009/10 - 2017/2018 seasons.
Non-age & age-structured model variants.
Values are scaled using weekly influenza positivity data.
Construct arrays storing weekly influenza positivity values.
Monitored through the RCGP sentinel swabbing scheme in England.
Population-level vaccine efficacy estimates
Daily vacc. uptake proportions for at-risk, low risk, entire population groups
Seasonal influenza vaccine uptake by season, for 2008/09 onwards
Pandemic influenza vaccine uptake for the 2008/09 influenza season
Data on circulating influenza virus composition from World Health Oragnisation FluNet database
Scripts to produce bar plots, violin plots and scatter plots.
ModelSimnData directory contains a file per fit performed, each containing outputs from 1,000 model simulation replicates.
Data files for the two simulated vaccine efficacy scenarios and Fig. 7 plot code.
Community-sourced files that are used to generate the figures.
Contains scripts to produce end-of-generation threshold value plots and posterior histograms.
Outputs from the ABC inference scheme reside in the following directories, comprising retained parameter sets and end-of-generation threshold values.
Output files generated fitting to the empirical data covering 2012/13-2016/17
Output files generated fitting to the empirical data covering 2012/13-2015/16
Output files generated fitting to the empirical data covering 2012/13-2017/18
Output files generated fitting to the synthetic data
Perform parameter inference on data using Approximate Bayesian Computation with an Adpative Population Monte Carlo scheme.
Directory housing files containing the ABC adaptive population Monte Carlo algorithm
Run simulations of the model using parameter sets representing samples from the posterior distribution.
Functions to numerically solve ODEs and update exposure history array