Generic model scenarios¶
The EMOD generic model is explained in detail in Model overview. While the various components that comprise the model are explained with examples, it may be more useful to learn the model through hands-on implementation. The following sections will introduce sets of example files that illustrate how the generic model works on particular topics. All files are available in the downloadable EMOD scenarios > Scenarios > Generic folder and, in addition to the explanations below, each scenario will have a more detailed README file to cover relevant information.
EMOD supports different simulation types for various diseases and disease classes. The features present in the GENERIC_SIM simulation type, which can be configured to model a variety of different diseases, are inherited by the more specific simulation types, such as malaria or HIV. Because all disease-specific EMOD simulation types are based on the generic model, the following scenarios are helpful in learning the basics of modeling with EMOD, even if you intend to use one of the disease-specific simulation types.
For more information on the software architecture and inheritance, see Overview of EMOD software.
Contents
Compartmental models¶
Before working with an agent-based model like EMOD, you will likely be familiar with deterministic compartmental models. In these models, an ordinary differential equation (ODE) controls the rate at which the population transitions from each disease state (compartment) to another. These models assume that individuals within each compartment are identical and the entire population is well-mixed. Given the same inputs, a compartmental model will always produce the same outcomes.
While compartmental models are very useful, in many situations an agent-based model is a better tool for describing complex phenomena. In these models, each agent (such as a human or a vector) is simulated individually. Their behavior and interactions with one another are determined using decision rules. For a more detailed comparison of EMOD and compartmental models, see Compartmental models and EMOD.
The following scenarios under EMOD scenarios > Scenarios > Generic show how you can use EMOD to simulate disease scenarios that might also be simulated using a compartmental model. For example, a disease like measles could be modeled using a compartmental SIR model or it could be modeled using EMOD configured such that there is no incubation period or waning immunity after infection.
SI
SIS
SIR
SIRS
SEIR
SEIR_VitalDynamics
SEIRS
Density scaling of infection¶
By default, EMOD uses frequency-dependent transmission and the overall transmissibility does not change with population size or density. An infected individual will infect the same number of people either in a small village or in a metropolitan area. The EMOD scenarios > Scenarios > Generic >DensityScaling scenario shows how you can configure EMOD so population density has an effect on the transmissibility of disease. For a detailed description of the mathematics behind this functionality, see Population density and transmission scaling.
Heterogeneous transmission¶
By default, the probability of transmitting disease to another is the same within each node in an EMOD simulation. However, we know that some people may be at greater risk based on their behavior, environment, or biology. The Heterogeneous Intra-Node Transmission (HINT) feature enables you to specify values for heterogeneity in transmission based on individual properties in a WAIFW matrix. For detailed information, see Property-based heterogeneous disease transmission.
EMOD scenarios > Scenarios > Generic > HINT_AgeAndAccess illustrates a common scenario in which the population has properties applied based on age and accessibility. Transmission is higher among individuals of similar ages and accessibility levels. Interventions are targeted based on these properties.
EMOD scenarios > Scenarios > Generic > HINT_SeattleCommuting illustrates a more novel scenario that uses the HINT feature. To simulate migration in EMOD, you can run multi-node simulations where individuals have a certain probability of migrating to a different node at each timestep (typically set to one day). This works well when individuals move to another location permanently or seasonally. However, for daily movement like commuting for work, you can use HINT to make transmission higher for people in the same area during the workday. This scenario assigns individual properties that represent the area codes surrounding Seattle and configures higher transmission for people in the same area code.
Disease reservoirs¶
Generally with EMOD, you model a disease by introducing an outbreak at some point in the simulation. However, endemic settings may involve a non-human disease reservoir that periodically reintroduces the disease to the human population. The EMOD scenarios > Scenarios > Generic > Zoonosis scenario illustrates how to configure EMOD to simulate such zoonotic diseases. For detailed information, see Disease outbreaks, reservoirs, and endemicity.
Interventions¶
The EMOD scenarios > Scenarios > Generic > Vaccinations scenario introduces the concept of adding interventions to stop disease transmission. It compares the outcomes when a disease outbreak has no interventions applied, when the entire population is vaccinated, and when you target the vaccination to a particular segment of the population.
Because interventions are often very specific to the disease being modeled, the other simulation types introduce diagnostic tests, drugs, and health system elements relevant to that disease. Therefore, the scenarios for the disease-specific simulation types provide more examples of intervention configuration. For detailed information, see Creating campaigns.