MGDrivE 2: A simulation framework for gene drive systems incorporating seasonality and epidemiological dynamics

Abstract

Interest in gene drive technology has continued to grow as promising new drive systems have been developed in the lab and discussions are moving towards implementing field trials. The prospect of field trials requires models that incorporate a significant degree of ecological detail, including parameters that change over time in response to environmental data such as temperature and rainfall, leading to seasonal patterns in mosquito population density. Epidemiological outcomes are also of growing importance, as: i) the suitability of a gene drive construct for release will depend on its expected impact on disease transmission, and ii) initial field trials are expected to have a measured entomological outcome and a modeled epidemiological outcome. We present MGDrivE 2 (Mosquito Gene Drive Explorer 2): an extension of and development from the MGDrivE 1 simulation framework that investigates the population dynamics of a variety of gene drive architectures and their spread through spatially-explicit mosquito populations. Key strengths and improvements of the MGDrivE 2 framework are: i) the ability of parameters to vary with time and induce seasonal population dynamics, ii) an epidemiological module accommodating reciprocal pathogen transmission between humans and mosquitoes, and iii) an implementation framework based on stochastic Petri nets that enables efficient model formulation and flexible implementation. Example MGDrivE 2 simulations are presented to demonstrate the application of the framework to a CRISPR-based homing gene drive system intended to drive a disease-refractory gene into a population, incorporating time-varying temperature and rainfall data, and predict impact on human disease incidence and prevalence. Further documentation and use examples are provided in vignettes at the project's CRAN repository. MGDrivE 2 is an open-source R package freely available on CRAN. We intend the package to provide a flexible tool capable of modeling gene drive constructs as they move closer to field application and to infer their expected impact on disease transmission.

Eliminating Aedes aegypti from its southern margin in Australia: insights from genomic data and simulation modeling

“…To model the expected performance of tank-sealing and IIT releases at suppressing and eliminating local Ae. aegypti populations in Murgon and Wondai, we used the MGDrivE 2 framework (17). This framework models egg, larval, pupal and adult mosquito life stages with overlapping generations, and larval mortality increasing with larval density.…”

“…aegypti populations. We employ a recently developed simulation modeling framework MGDrivE 2 (17), which enables simulations of seasonally-variable mosquito population dynamics and is capable of modeling different control strategies in spatially-structured mosquito populations (17). To explore how vulnerable each WBBR town would be to mosquito re-invasion (post elimination), we ascertained historical and contemporary population connectivity through the analyses of population genomic data from Ae.…”

A dynamically-structured matrix population model based on renewal processes for accumulative development under variable environmental conditions

“…It is important for the predictive models of insect population dynamics to incorporate realistic development time distributions, which behave reliably under variable environmental conditions. Numerous modelling frameworks have been implemented specifically to tackle the problem of predicting the impact of environmental variation on insect populations, including the degree-day approach (24–26), matrix population models (with environmental perturbations (3, 27, 28) and temperature forcing (6, 29)), and ordinary (30, 31), delay (32, 33), and partial differential equations (34). However, only a subset of these incorporate realistic development time distributions (6, 29, 31).…”

“…A common technique for developing pseudo-stage-structured models is to employ a series of identical sub-stages with exponentially-distributed dwell times to yield an ordinary differential equations (ODE) system with an Erlang-distributed time for development from one life-stage to the next (also known as the linear chain trick) (8, 12). This approach has been generalised for a broader set of development time distributions (12) and adapted for stochastic modelling (31, 36, 37). When using the linear chain trick, one can transform variations in the rate of transition through the sub-stages into variations in the Erlang distribution characteristics (8, 31).…”

“…This approach has been generalised for a broader set of development time distributions (12) and adapted for stochastic modelling (31, 36, 37). When using the linear chain trick, one can transform variations in the rate of transition through the sub-stages into variations in the Erlang distribution characteristics (8, 31). However, due to the fixed number of sub-stages, change in the ratio of the mean to standard deviation is not permitted.…”