NetLogo User Community Models
## WHAT IS IT?
This model simulates the transmission and dispersal dynamics of sylvatic yellow fever in a closed population of humans, monkeys, and mosquitoes. The user will be able to simulate and identify in which situations the virus is more (or less) likely to spread among the human population living close a forest where monkeys and mosquitoes are infected.
Briefly, yellow fever is an infectious disease whose etiological agent is the yellow fever virus (YFV), which is transmitted to humans and non-human primates through the bite of infected mosquitoes. Illness ranges from a fever and body aches to severe liver disease with bleeding and jaundice (yellowing skin). In tropical and subtropical areas of Africa and South and Central America, the jungle (sylvatic) cycle of the disease has been observed, with the virus circulating in forested areas where wild mosquitoes transmit YFV to monkeys and sometimes to humans that visit, work or live close the forest. Outbreaks of yellow fever occurred regularly in urban areas of the Americas until the first half of the twentieth century, transmitted to humans by the Aedes aegypti mosquito. With the development of the YFV vaccine and the control programs to eliminate the Aedes aegypti, the urban circulation of yellow fever was disrupted. However, the occurrence of YFV outbreaks in forest areas in the vicinity of cities, with the onset of human cases, raises great concern about the possibility of the virus circulating again in urban areas. This concern is reinforced by the fact that the urban vector Aedes aegypti is a widespread and abundant mosquito in most cities of the South and Central America, and in many regions the human population has low vaccination coverage.
By manipulating the model parameters related to human, monkey, and mosquito populations the user can simulate scenarios of higher risk for the virus to spread through the human population.
Overall, this model helps the user to understand:
## HOW IT WORKS
The world is divided into three environments: forest (green patches), border (pink patches), and urban (gray patches).
There are four types of agents: humans (person shape), non-human primates (circle shape), urban-mosquitoes (triangle shape), and forest-mosquitoes (square shape). Humans and urban-mosquitoes move randomly, interact, and 'live' in the urban patches, while monkeys and forest-mosquitoes move randomly, interact, and 'live' in the forest. All different types of agents can move and interact on the border (patches between urban and forest). A choose proportion of humans can enter and move inside the forest.
All agents can become infected and transmit the virus at rates that will depend on the selected parameters.
Mosquitoes follow an SEI (Susceptible - Exposed - Infected) transmission model. Susceptible individuals are exposed to the virus and may enter a latent period (14 days) in which the virus has already infected the individual but cannot yet be transmitted. After this, the infectious period begins and the mosquito can transmit the virus to susceptible monkeys and humans. The mosquito will remain infectious for the rest of its life. For simplicity, we chose to keep the mosquito population constant. Thus, when a particular mosquito dies, another one is born in the susceptible state.
Humans and monkeys follow an SEIR (Susceptible - Exposed - Infected - Recovery) model. Susceptible humans can become infected by urban or forest mosquito bites. The exposed or latent phase will last 4 days and after this, an infectious period lasting 6 days will begin. During the infectious period, infected humans can transmit the virus to urban and forest mosquitoes. After the infectious period, humans have an 85% chance of recovering and becoming immune to the virus. If they do not recover, they will enter the toxic phase (the severe form of the disease), which lasts 7 days. At this stage, individuals remain ill but there is no further transmission of the virus to mosquitoes. At the end of the toxic phase, individuals will have a 50% chance of recovering and becoming immune; otherwise, they will die. For monkeys, the only difference from humans is that after the infectious period everyone will enter the toxic phase of the disease that will last 10 days. After this, only 20% of monkeys will survive and become immune (this high lethality is similar to what is observed for howler monkeys in forests of South America).
The presence of the virus in the populations is represented by the colors of individuals. Five colors are used: susceptible = white; exposed = red; infected = black; immune = blue; toxic-phase = yellow.
The graph YELLOW FEVER PREVALENCE shows the cumulative percentage of the population of unvaccinated humans and monkeys that were infected with YFV. The graph MOSQUITO INFECTION RATE shows the percentage of infected individuals per day in the populations of forest and urban mosquitoes. The three monitors above the graphs show the total humans that were infected, the infected humans that died, and the total monkeys that died during the outbreak.
When there are only susceptible or immune individuals in the world the simulation stops.
## HOW TO USE IT
The SETUP button creates individuals according to the parameter values chosen by the user. The outbreak will start from a certain percentage of infected forest-mosquitoes that will depend on the INITIAL-INFECTED-MOSQUITOES parameter.
Once the simulation has been setup, push the GO button to run the model. GO starts the simulation and runs it continuously until GO is pushed again or the virus is no longer circulating between individuals.
Each time-step can be considered to be a day.
The following is a summary explanation of the sliders in the model.
INITIAL-MONKEYS (initialized to vary between 5 - 50): The total number of monkeys the simulation begins with.
MONKEY-MOBILITY (0.1 - 2): This indicates how mobile the monkey is. The higher the value, the greater the mobility of the individual in each time step. Individuals move randomly by this assigned value.
VACCINE-COVERAGE (0 - 100%): The proportion of human individuals vaccinated against YFV. Vaccinated individuals will already start the simulation as immune.
FREE-WALKERS (0 - 100%): The proportion of human individuals free to entering and walking through the forest. This parameter takes into account that some individuals in the human population will have greater contact with the forest than others and, in this case, will be more exposed to the virus infection if they are not immune.
INITIAL-PEOPLE (30 - 300): The total number of humans the simulation begins with.
HUMAN-MOBILITY (0.1 - 2): This indicates how mobile the person is. The higher the value, the greater the mobility of the individual in each time step. Individuals move randomly by this assigned value.
INITIAL-INFECTED-MOSQUITOES (0.5 - 10%): The probability that each individual of the forest-mosquitoes population has to start the simulation in the infectious state.
MOSQUITO-MORTALITY-RATE (0.01 - 0.5): The probability of a mosquito dying in each time-step after 15 time-steps (minimum life span considered in the model). To simplify the model, the mortality rate is considered equal for urban and forest mosquitoes.
MOSQUITO-DAILY-BITING-RATE (0.2 - 1): The daily rate at which mosquitoes seek for a blood meal. For example, a value of 0.25 means that in a potentially infectious contact there is a 25% chance the mosquito will bite the host.
INITIAL-URBAN-MOSQUITOES (100 - 1000): The total number of urban mosquitoes the simulation begins with. For each individual who dies a new individual is born, so the final number of mosquitoes will be equal to the initial number.
INITIAL-FOREST-MOSQUITOES (100 - 1000): The total number of forest mosquitoes the simulation begins with. For each individual who dies a new individual is born, so the final number of mosquitoes will be equal to the initial number.
U-MOSQ-INFECTION-COMPETENCE (0 - 1): The probability that a susceptible individual from the urban-mosquitoes population has to become infected by biting an infectious human or monkey.
F-MOSQ- INFECTION-COMPETENCE (0 - 1): The probability that a susceptible individual from the forest-mosquitoes population has to become infected by biting an infectious monkey or human.
U-MOSQ-TRANSMISSION-COMPETENCE (0 - 1): The probability that an infectious individual from the urban-mosquitoes population has to transmit the virus to a susceptible human or monkey by biting them.
F-MOSQ-TRANSMISSION-COMPETENCE (0 - 1): The probability that an infectious individual from the forest-mosquitoes population has to transmit the virus to a susceptible monkey or human by biting them.
## THINGS TO NOTICE
Note that the outbreak begins with a rapid increase in the number of infected monkeys, which is called epizootics. This event is followed by an increase in the number of infected forest mosquitoes as monkeys are currently serving as virus amplifiers. As monkeys die or survive and become immune, the rate of infection in wild mosquitoes tends to shrink to zero, while the prevalence curve of infected monkeys tends to stabilize.
There are two possibilities of starting an outbreak in humans. A first possibility is that susceptible urban mosquitoes become infected by biting infectious monkeys on the border and transmitting the virus to susceptible humans. The second possibility is that susceptible humans become infected by being bitten by infectious forest mosquitoes and transmit the virus to susceptible urban mosquitoes. In both cases, there is a possibility that an outbreak will start among the human population, and the spread of the virus will depend on the selected parameters (mainly vaccine coverage).
During outbreaks the number of humans who die is less than the total number of infected because around 85% of infected humans tend to recover and become immune, the others 15% enter the toxic phase and around 50% survive. However, these proportions present some variations in outbreaks observed in different places and times, which usually depends on the genetic lineage of the virus, among other factors.
There are many species of non-human primates that can become infected with YFV. This model considers some known infection parameters for howler monkeys, which are primates quite susceptible to YFV and have high mortality rates during outbreaks. Therefore, the number of monkeys killed during simulations will tend to be proportionally higher compared to humans.
## THINGS TO TRY
Try to identify which situations may favor an outbreak in the human population by changing one parameter at a time and then using different combinations of these parameters. Which parameters can be considered the most important?
Define a set of parameters and try to identify the minimum proportion of immunization coverage needed to prevent a YFV outbreak from spreading to the human population.
In addition to vaccination coverage, what other parameters (or set of parameters) can prevent the virus from leaving the forest and spreading in urban areas?
## EXTENDING THE MODEL
This model assumes that there is no seasonal variation in mosquito abundance. However, this is a very important factor in the dynamics of YFV outbreaks. Try to include seasonal variations in mosquito populations to see how virus transmission behaves over time.
Some of the fixed parameters of the model (such as latency and infectious period in monkeys and humans) can be changed to become means of a statistical distribution.
Different mortality and bite rates could be included for urban and wild mosquitoes.
The inclusion of other mosquito species that act as intermediate vectors and a virus reservoir animal could add complexity to the model and perhaps come a little closer to reality.
## NETLOGO FEATURES
In this model, the potential infectious contact between mosquitoes and primates occurs based on a Moore Neighborhood pattern. This means that the infectious individual will select a susceptible individual (if any) from the eight patches adjacent to his.
## RELATED MODELS
epiDEM Basic, epiDEM trave and control, HIV, Virus and Virus on a Network are related models.
## CREDITS AND REFERENCES
If you mention this model or the NetLogo software in a publication, we ask that you include the citations below.
For the model itself:
* Medeiros-Sousa, Antônio Ralph (2019). Sylvatic YF model. School of Public Health, University of São Paulo.
Please cite the NetLogo software as:
* Wilensky, U. (1999). NetLogo. http://ccl.northwestern.edu/netlogo/. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL.
(back to the NetLogo User Community Models)