Skip to content
This repository has been archived by the owner on Sep 4, 2024. It is now read-only.

Supporting material for the paper "The impact of past vaccination coverage and immunity on pertussis resurgence", M. Domenech de Cellès, F. M. G. Magpantay, A. A. King, and P. Rohani, Science Translational Medicine 10(434): eaaj1748, 2018. https://doi.org/10.1126/scitranslmed.aaj1748

License

Notifications You must be signed in to change notification settings

kingaa/massachusetts-honeymoon

Repository files navigation

massachusetts-honeymoon

Supporting material for the 2018 Science Translational Medicine paper, "The impact of past vaccination coverage and immunity on pertussis resurgence", by M. Domenech de Cellès, F. M. G. Magpantay, A. A. King, and P. Rohani.

This repository contains saved the R objects (.rds) and R scripts (.R) needed to run the estimations for the deterministic and stochastic models. Download and unpack all the files first. Next, run the script install_packages.R; it installs the "checkpoint" package (if needed) and the other required packages at their version on 1 October 2015. (In particular, this ensures that the pomp package version is 1.2.1.1) Please note that the estimations are based on synthetic data, simulated from the waning stochastic model. This is because we were not allowed by the Massachusetts Department of Public Health (MDPH) to make their data public. The real data are available from the MDPH upon request to the Office of Integrated Surveillance and Informatic Services help desk (isishelp@state.ma.us).

The main scripts are then:

  1. run_traj_match.R: run trajectory matching for the deterministic models. This script example runs a single estimation from one parameter set starting value, with maximum execution time of 1 min.
  2. run_mifs.R: run the maximum iterated filtering algorithm to estimate the parameters of the stochastic models. This script example runs a single estimation (with 2 MIF iterations and 2e2 particles), followed by 2 evaluations of the likelihood using particle filtering (with 2e2 particles).

Other items:

  1. R objects:
    • covars.rds: data frame containing the covariates used in the model (birth and age-specific migration rates)
    • simulated_data.rds: data frame that contains the simulated data, i.e., monthly age-specific case reports during 1990–2005
    • mle_waning_model_deterministic.rds: named vector containing the model parameters (fixed+estimated MLE) of the deterministic model
    • mle_waning_model_stochastic.rds: named vector containing the model parameters (fixed+estimated MLE) of the stochastic model
  2. Other files:
    • model_equations.c: C code implementing the observation and the process (deterministic and stochastic variants) models
    • create_pomp_model.R: R script implementing a function that creates the pomp object.

About

Supporting material for the paper "The impact of past vaccination coverage and immunity on pertussis resurgence", M. Domenech de Cellès, F. M. G. Magpantay, A. A. King, and P. Rohani, Science Translational Medicine 10(434): eaaj1748, 2018. https://doi.org/10.1126/scitranslmed.aaj1748

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published