MVSE, Mosquito-borne Viral Suitability Estimator, provides methods for the estimation of an index of climate-based transmission potential for mosquito-borne viruses. It can be parameterized for any combination of host, virus and mosquito-species of interest with local climate data.
The MVSE package is currently not hosted on CRAN (Comprehensive R Archive Network). We thus need to install the package directly from the GithHub repo where it is hosted.
install.packages("devtools")
install.packages("tidyverse")
require(tidyverse)
require(devtools)
install_github("TaishiNakase/MVSE")
Let’s go through a short example of how one might use MVSE to estimate Index P for some mosquito-borne virus.
First, we will install some packages.
library(MVSE)
We first need to create a mvse_model
object, which requires time
series data for climatic variables (i.e. temperature, humidity and
rainfall) as well as epi-entomological prior distributions for the
virus/vector/host system of interest.
# climate data
data("climateFSA")
# user-defined model
priors <- list(mosq_life_exp=list(pars=c(mean=12, sd=2), dist="normal"),
mosq_inc_per=list(pars=c(mean=7, sd=2), dist="normal"),
mosq_biting_freq=list(pars=c(mean=0.25, sd=0.01), dist="normal"),
human_life_exp=list(pars=c(mean=71.1, sd=2), dist="normal"),
human_inc_per=list(pars=c(mean=5.8, sd=1), dist="normal"),
human_inf_per=list(pars=c(mean=5.9, sd=1), dist="normal"))
example_mvse_model <- mvse_model(model_name="user_test", climate_data=climateFSA, priors=priors)
Let’s take a look at the input data to be used in the sampling procedure.
example_mvse_model
#> Class: mvsemodel
#> Model name: user_test
#> Model category: user-defined
#> Climate data (limited to first 10 rows):
#> date T H
#> 1 2015-01-01 27.54 65.25
#> 2 2015-01-02 26.06 72.75
#> 3 2015-01-03 25.58 71.00
#> 4 2015-01-04 24.62 73.50
#> 5 2015-01-05 26.26 75.75
#> 6 2015-01-06 26.50 77.50
#> 7 2015-01-07 25.38 81.75
#> 8 2015-01-08 24.22 95.25
#> 9 2015-01-09 24.50 85.25
#> 10 2015-01-10 25.10 80.25
#> Priors:
#> Mosquito life expectancy (days) : normal(mean=12, sd=2)
#> Mosquito incubation period (days) : normal(mean=7, sd=2)
#> Mosquito biting frequency (bites/female/day) : normal(mean=0.25, sd=0.01)
#> Human life expectancy (years) : normal(mean=71.1, sd=2)
#> Human incubation period (days) : normal(mean=5.8, sd=1)
#> Human infectious period (days) : normal(mean=5.9, sd=1)
Next, let’s perform the MCMC (Markov chain Monte Carlo) sampling procedure to estimate Index P.
example_mvse_fit <- sampling(example_mvse_model, verbose=FALSE)
Finally, let’s take a look at the estimated distribution of the time series of Index P.
indexP_plot <- mcmc_index_dist(example_mvse_fit, index="indexP")
indexP_plot