Spatial Agent-Based Simulation Modeling in Public Health

S. M. Niaz Arifin, PhD, is Research Assistant Professor in the Department of Computer Science and Engineering at the University of Notre Dame. A member of The Society for Computer Simulation and American Society of Tropical Medicine and Hygiene and the recipient of The American Society of Tropical Medicine and Hygiene Travel Award in 2011, his research interests include agent-based modeling and simulation, public health, data warehousing, and geographic information systems.Gregory R. Madey, PhD, is Research Professor in the Department of Computer Science and Engineering at the University of Notre Dame. A member of The Society for Computer Simulation, Institute of Electrical and Electronics Engineers Computer Society, and American Society of Tropical Medicine and Hygiene, his research interests include agent-based modeling and simulation, cyberinfrastructure, bioinformatics, biocomplexity, e-Technologies, open source software, disaster management, and health informatics.Frank H. Collins, PhD, is Professor in the Department of Biological Sciences at the University of Notre Dame. His research interests include genome level studies of arthropod vectors of human pathogens, the biology of malaria vectors with a focus on the development of molecular tools that will permit better resolution of questions about vector population ecology, ecological genetics, and the epidemiology of malaria transmission.

• List of Contributors xvList of Figures xviiList of Tables xxiPreface xxiiiAcknowledgements xxixList of Abbreviations xxxi1 Introduction 11.1 Overview 11.2 Malaria 31.3 Agent-Based Modeling of Malaria 41.4 Contributions 41.5 Organization 52 Malaria: A Brief History 72.1 Overview 72.2 Malaria in Human History 72.2.1 The Malarial Path: Ancient Origins 82.2.2 Naming and Key Discoveries 92.2.3 Antimalarial Drugs 92.2.4 Prevention Measures 102.3 Malaria Epidemiology: A Global View 102.3.1 The Malaria Parasite 112.3.2 Geographic Distribution 122.3.3 Types of Transmission 122.3.4 Risk Mapping and Forecasting 132.4 Malaria Control 133 Agent-Based Modeling and Malaria 173.1 Overview 173.2 Agent-Based Models (ABMs) 173.2.1 Agents 183.2.2 Environment 193.2.3 Rules 203.2.4 Software for ABMs 203.3 History and Applications 213.3.1 M&S Organizations 213.4 Advantages of ABMs 233.4.1 Emergence, Aggregation, and Complexity 233.4.2 Heterogeneity 243.4.3 Learning and Adaptation 243.4.4 Flexibility in System Description 243.4.5 Inclusion of Multiple Spaces 253.4.6 Limitations of ABMs 253.4.7 ABMs vs Mathematical Models 273.4.8 Applicability of ABMs for Malaria Modeling 283.5 Malaria Models: A Review 293.5.1 Mathematical Models of Malaria 303.5.2 Agent-Based Models (ABMs) of Malaria 333.5.3 The Spatial Dimension of Malaria Models 353.6 Summary 364 The Biological Core Model 394.1 Overview 394.1.1 Relevant Terms of Interest 404.2 The Aquatic Phase 414.2.1 Egg (E) 424.2.2 Larva (L) 434.2.3 Pupa (P) 454.3 The Adult Phase 464.3.1 Immature Adult (IA) 464.3.2 Mate Seeking (MS) 474.3.3 Blood Meal Seeking (BMS) 474.3.4 Blood Meal Digesting (BMD) 474.3.5 Gravid (G) 474.4 Aquatic Habitats and Oviposition 484.4.1 Aquatic Habitats 484.4.2 Oviposition 484.5 Senescence and Mortality Rates 504.5.1 Senescence 504.5.2 Mortality Models: Basic Mathematical Formulation 514.6 Mortality in the Core Model 514.6.1 Aquatic (Immature) Mortality Rates 524.6.2 Adult Mortality Rates 534.7 Discussion 534.7.1 An Extendible Framework for Other Anopheline Species 534.7.2 Weather, Seasonality, and Other Factors 544.7.3 Mortality Rates 544.8 Summary 545 The Agent-Based Model (ABM) 575.1 Overview 575.2 Model Architecture 585.2.1 Object-Oriented Programming (OOP) Terminology 585.2.2 Agents 605.2.3 Environments 625.2.4 Event-Action-List Diagram 625.3 Mosquito Population Dynamics 645.4 Model Features 665.4.1 Processing Steps Ordering 665.4.2 Model Assumptions 675.4.3 Simulations 695.5 Summary 696 The Spatial ABM 716.1 Overview 716.2 The Spatial ABM 746.2.1 Definition of Terms 746.2.2 Landscapes 756.2.3 Landscape Generator Tools 766.3 Resource Clustering 796.4 Flight Heuristics for Mosquito Agents 816.5 Simulation Results 856.5.1 Model Verification 856.5.2 Landscape Patterns 866.5.3 Relative Sizes of Resources 876.5.4 Resource Density 886.5.5 Combined Habitat Capacity (CHC) 896.6 Spatial Heterogeneity 906.7 Summary 937 Verification, Validation, Replication, and Reproducibility 957.1 Overview 957.2 Verification and Validation (V&V): A Review 967.2.1 Acceptability Assessment and Quality Assurance (QA) 967.2.2 Verification and Validation (V&V) 987.3 Replication and Reproducibility (R&R): A Review 1007.4 Summary 1018 Verification and Validation (V&V) of ABMs 1038.1 Overview 1038.2 Verification and Validation (V&V) of ABMs 1038.3 Phase-Wise Docking 1058.3.1 Assumptions for the ABMs 1058.3.2 Phase-Wise Docking Results 1078.4 Compartmental Docking 1108.4.1 Implementations of the ABMs 1118.4.2 Assumptions for the ABMs 1128.4.3 Compartmental Docking Results 1148.5 Summary 1169 Replication and Reproducibility (R&R) of ABMs 1219.1 Overview 1219.1.1 Simulation Stochasticity 1229.1.2 Boundary Types 1239.2 Vector Control Interventions 1249.2.1 Larval Source Management (LSM) 1259.2.2 Insecticide-Treated Nets (ITNs) 1269.2.3 Population Profiles for ITNs 1279.2.4 Coverage Schemes for ITNs 1279.2.5 Applying LSM in Isolation 1309.2.6 Applying ITNs in Isolation 1329.2.7 Applying LSM and ITNs in Combination 1329.3 Simulation Results 1349.3.1 Simulation Stochasticity 1349.3.2 LSM in Isolation 1349.3.3 Impact of Boundary Types 1379.3.4 ITNs in Isolation 1389.3.5 LSM and ITNs in Combination 1439.4 Replication and Reproducibility (R&R) Guidelines 1479.5 Discussion 1509.6 Summary 15210 A Landscape Epidemiology Modeling Framework 15510.1 Overview 15510.2 GIS in Public Health 15910.3 The Study Area and the ABM 16010.3.1 Features of the Spatial ABM 16110.3.2 Vector Control Intervention Scenarios 16210.4 The Geographic Information System (GIS) 16310.4.1 The GIS-ABM Workflow 16310.4.2 GIS Processing of Data Layers 16410.4.3 Feature Counts 16510.5 Simulations and Spatial Analyses 16510.5.1 Output Indices 16610.5.2 Hot Spot Analysis 16710.5.3 Kriging Analysis 16710.6 Results 16810.6.1 Mosquito Abundance 16810.6.2 Oviposition Count per Aquatic Habitat 17110.6.3 Blood Meal Count per House 17410.7 Discussion 17710.7.1 Stochasticity and Initial Conditions 17710.7.2 Model Calibration and Parameterization 17810.7.3 Emergence 17810.7.4 Complexity 17910.7.5 Data Resolution (Granularity) 17910.7.6 Spatial Analyses 18010.7.7 Habitat-based Interventions 18110.7.8 Miscellaneous Issues 18110.8 Conclusions 18211 The EMOD Individual-Based Model 185Philip A. Eckhoff and Edward A. Wenger11.1 Overview 18511.1.1 Motivation: Modeling of Malaria Eradication 18611.1.2 Questions that Arise in the Context of Malaria Eradication 18711.1.3 Spatial Heterogeneity and Metapopulation Effects 18811.1.4 Implications for Model Structure 19011.2 Model Structure 19311.2.1 Human Demographics and Synthetic Population 19311.2.2 Vector Ecology 19411.2.3 Vector Transmission 19511.2.4 Within-Host Disease Dynamics 19711.2.5 Human Migration and Spatial Effects 19811.2.6 Stochastic Ensembles 20011.3 Results 20111.3.1 Single-Village Simulations 20111.3.2 Spatial Simulations: Garki District 20211.3.3 Madagascar 20311.4 Discussion 206Appendix A Enzyme Kinetics Model for Vector Growth and Development 209A.1 Overview 209A.2 Stochastic Thermodynamic Models 210A.3 Poikilothermic Development Models 210A.4 The Sharpe and DeMichele Model 214A.5 The Schoolfield et al. Model 217A.6 Depinay et al. Model 219A.7 Summary 221Appendix B Flowchart for the ABM 223B.1 Flowchart for the Agent-Based Model (ABM) 223Appendix C Additional Files for Chapter 10 233Appendix D A Postsimulation Analysis Module for Agent-Based Models 239D.1 Overview 239D.2 Simulation Output Analysis: A Review 240D.3 The LiNK Model 243D.4 P-SAM Architecture 245D.5 Postsimulation Analysis and Visualization 247D.6 P-SAM Performance 250D.7 Conclusion 254References 255Index 279