SoNA3BS

An Agent-Based Model to Study High-Resolution Spatial Heterogeneity in Aedes aegypti-Borne Diseases Through the Use of Network Theory 



const char* authors[] = {
    "Candidate: Héctor Manuel Sánchez Castellanos",
    "Advisor: Edgar Emmanuel Vallejo Clemente",
    "Co-Advisor: John M. Marshall"
};


Agenda

I. Background (5 min)
II. Model Description (5 min)
III. Validation Experiments (15 min)
IV. Spatial Heterogeneity Experiments (10 min)
V. Conclusions, Limitations and Future Work (5 min)

I. Background

1. Motivations 2. Hypotheses

1. Epidemiological Motivation

Dengue: 390 million annual cases and rising.
Zika: Cases all accross the Americas and widespread microcephaly fear.
Chikungunya: 693,489 suspected cases and 37,480 confirmed only in the Americas.


Human distribution and movement are drivers in Aedes-borne diseases spread.
Traditional control interventions are insufficient (and novel need efficacy predictions).



1. (ABM+Networks) Motivation



2. Main Hypothesis

The creation and use of an agent-based framework in conjunction with network theory analytical tools can provide epidemiological insight on how the spatial relations of human-mosquito interactions affect the potential spread of Aedes aegypti-transmitted diseases.


2. Sub-Hypotheses

1. A computational agent-based model that adheres to existing knowledge of Aedes aegypti biology can be created and run in reasonable computational times on general purpose computers (on the order of days).
2. The model's adult mosquito population dynamics matches an independently-created model's number of adult females within the range of 15°C to 32.5°C temperatures and up to 20 breeding sites within an absolute error rate of 1.5.
3. The model can track bites between humans that give rise to emergent vectorial-contact networks; and these networks provide insight into possible epidemiological processes in different spatial layouts.

II. Model

1. Agents' Definition 2. Mosquitoes 3. Humans

1. Agents' Definition

Reactive Architecture


2. Mosquito Life Stages




2. Mosquito: Aquatic Stages



2. Mosquito: Adult Stages




3. Humans


Types: Homebound, Visiting, Worker
Probabilistic Visits (diurnal): Uniform, Gravity-Based

III. Validation

1. Methodology Description 2. Experiments Descriptions
3. Results 4. Analysis 5. Conclusions

1. Validation by "Proxy"





1. ODE Model



2. Experiments Descriptions



2. Factorial Responses



3. Seasonality Responses



4. Modelling Differences



4. Collapsing ODE into ABM



4. Collapsing ODE into ABM



5. Experiments' Conclusions

Within the range of our scope, the level of error falls within specifications.
The differences can be explained by modelling discrepancies.
There should be between-scenario consistency
(offset in bites should be fairly uniform given same conditions).

*Note: ABM has the advantage of taking into account heterogeneity.

IV. Networks and Spatial Heterogeneity Experiments

1. Experimental Setups 2. Population Dynamics Results
3. Networks Results 4. Conclusions


1. Experimental Setups




2. Population Dynamics Results




2. Networks Emergence



2. Networks Emergence



3. Networks Analysis



4. Experiments' Conclusions

Spatial heterogeneity is paramount for the study of Aedes-borne diseases in human populations.
Transitional bites are strongly related to spatial distribution.
Mosquito population sizes are not enough to explain the details of transmission processes.
By studying vector-contact networks we can better understand potential epidemiological processes.

V. Conclusions & Future Work

1. Conclusions 2. Contributions 3. Limitations and Future Work
4. Publications & Questions

1. Conclusions

The use of ABMs coupled with network theory analysis provides us with a framework to analyse complex spatial interactions in human-mosquito systems.
-ABMs can help us re-create complex real-life scenarios and optimise the use of resources.
-Network theory can provide insight on how behaviours alter the patterns of pathogen transmission.



2. Main Contributions

Engineering: Tools to evaluate, compare and optimise the efficacy of mosquito-control measures; along with the way human and mosquito movement affect the epidemiological connections between individuals.

Scientific: The means to perform the reduction of a biological process into a networks domain of application.

Application: The recreation of a human population in the Mexican town of Catemaco, Veracruz; in which we make use of the full-fledged simulation and analysis tools developed in this project to analyse the way the spatial distributions of humans and mosquitoes affects the way they connect epidemiologically and the implications this could have in mosquito-borne pathogens' spread.


3. Limitations & Future Work

Pathogens



3. Limitations & Future Work

Spatial Scale



3. Limitations & Future Work

Biological Calibration and Extensions



4. Publications & Questions