Grant: Discovery Grants Program - Individual
York University
$27,000.00
Computational Methods for Complex Infectious Disease Dynamics: Model Validation with Data Assimilation
FY 2024-25
Summary
Modelling and computational methods, using domain-specific data, are key to dissect spatiotemporal spread of infectious diseases (IDs), forecast, and inform intervention policies. Agent-based models (ABMs) provide a comprehensive framework to quantify the risk and impact of IDs, and to explore the interplay between heterogeneities and evolutionary mechanisms at both the individual and population levels. Validity and testability of these models rest on calibration to time-varying observations that are affected by numerous factors, especially for (re)emerging pathogens with an evolving nature affecting human behaviour. I propose a research program to bridge important methodological gaps in computational modelling with the goal of advancing scientific tools to:
1. Develop and validate the underlying theoretical constructs in ABMs, with the aim of providing predictive indicators of real-time data pertaining to epidemiological observations.
2. Create a data assimilation strategy, using non-linear particle filtering (PF) methods, to integrate data features (e.g., random effects) in ABMs and examine factors influencing the outcomes of ID interventions.
Long-term goals: to establish novel and computationally feasible methods to ensure (i) the adequacy of ABMs and their underlying theoretical basis (i.e., conceptual validation) in characterising the real-world observations; (ii) the error-free design of the intrinsic model structure with prescribed rules (i.e., internal validation); and (iii) the accuracy of the models in matching data (i.e., external validation).
The generic framework of an ABM will be developed based on four main components: (1) state variables (2) parameter space; (3) structural design; and (4) testability analysis. Theoretical validation will be explored via analyses of equation-based models with sufficiently detailed structures that could replicate the essential component of ABMs. To empirically validate this framework, we will develop adaptive PF methods, to generate posterior distributions for the state of the system by conducting statistical inference. New computational methods will be designed to alleviate the pitfalls of sampling methods in PF and facilitate recursive model calibration to real-time observations.
This framework will be adapted to specific diseases (e.g., Influenza, COVID-19) using historical data. Combined with a review of ID interventions, the testability of the model will be assessed to infer optimal parameters (e.g., transmission probability) and generate observed trends in epidemics. This architecture will result in algorithms able to efficiently assimilate real-time data, allowing for orchestrating a set of evidence-based policies derived from evaluating interventions and counterfactual scenarios. This program will also offer exceptional training opportunities to highly qualified individuals to acquire strong leadership and crucial skills, and contribute to the Canadian and global society.
Program
Discovery Grants Program - Individual
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
RGPIN-2023-05082
Name nom
Moghadas, Seyed
Department département
Mathematics and Statistics
Organization id
38
Institution établissement
York University
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2024
Competition year année de concours
2023
Award amount
27000
Program id
RGPIN
Program name en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
DISCOVERY RESEARCH
Group fr
RECHERCHE AXÉE SUR LA DÉCOUVERTE
Committee code
1508
Committee name en
Mathematics and Statistics
Committee name fr
Mathématiques et statistique
Area of application code
405
Area of application group en
Environment
Area of application group fr
Environnement
Area of application en
Modelling and mathematical simulation of natural processes
Area of application fr
Modélisation et simulation mathématique des processus naturels
Research subject code
2960
Research subject group en
Applied mathematics
Research subject group fr
Mathématiques appliquées
Research subject en
Mathematical modelling
Research subject fr
Modélisation mathématique
Field of research list names en
Not available
Field of research list names fr
Non disponible
Socioeconomic objective list names en
Not available
Socioeconomic objective list names fr
Non disponible
Application title
Computational Methods for Complex Infectious Disease Dynamics: Model Validation with Data Assimilation
Keywords
computational modelling; infectious disease dynamics; agent-based simulations; data analytics; statistical analysis; public health interventions; equation-based epidemic modelling; stochastic processes; particle filtering; sampling methods
Application summary
Modelling and computational methods, using domain-specific data, are key to dissect spatiotemporal spread of infectious diseases (IDs), forecast, and inform intervention policies. Agent-based models (ABMs) provide a comprehensive framework to quantify the risk and impact of IDs, and to explore the interplay between heterogeneities and evolutionary mechanisms at both the individual and population levels. Validity and testability of these models rest on calibration to time-varying observations that are affected by numerous factors, especially for (re)emerging pathogens with an evolving nature affecting human behaviour. I propose a research program to bridge important methodological gaps in computational modelling with the goal of advancing scientific tools to:
1. Develop and validate the underlying theoretical constructs in ABMs, with the aim of providing predictive indicators of real-time data pertaining to epidemiological observations.
2. Create a data assimilation strategy, using non-linear particle filtering (PF) methods, to integrate data features (e.g., random effects) in ABMs and examine factors influencing the outcomes of ID interventions.
Long-term goals: to establish novel and computationally feasible methods to ensure (i) the adequacy of ABMs and their underlying theoretical basis (i.e., conceptual validation) in characterising the real-world observations; (ii) the error-free design of the intrinsic model structure with prescribed rules (i.e., internal validation); and (iii) the accuracy of the models in matching data (i.e., external validation).
The generic framework of an ABM will be developed based on four main components: (1) state variables (2) parameter space; (3) structural design; and (4) testability analysis. Theoretical validation will be explored via analyses of equation-based models with sufficiently detailed structures that could replicate the essential component of ABMs. To empirically validate this framework, we will develop adaptive PF methods, to generate posterior distributions for the state of the system by conducting statistical inference. New computational methods will be designed to alleviate the pitfalls of sampling methods in PF and facilitate recursive model calibration to real-time observations.
This framework will be adapted to specific diseases (e.g., Influenza, COVID-19) using historical data. Combined with a review of ID interventions, the testability of the model will be assessed to infer optimal parameters (e.g., transmission probability) and generate observed trends in epidemics. This architecture will result in algorithms able to efficiently assimilate real-time data, allowing for orchestrating a set of evidence-based policies derived from evaluating interventions and counterfactual scenarios. This program will also offer exceptional training opportunities to highly qualified individuals to acquire strong leadership and crucial skills, and contribute to the Canadian and global society.