Broadwick: a framework for computational epidemiology.

Area of Science:

• Epidemiology
• Computational Biology
• Mathematical Modeling

Background:

• Modern disease outbreaks generate vast datasets challenging to integrate into simple epidemiological models.
• Complex computational models are often required for efficient data integration, posing development challenges.
• Existing frameworks may lack systematic approaches and user-friendliness for handling extensive outbreak data.

Purpose of the Study:

• Introduce a novel, user-friendly framework for building and running epidemiological models.
• Facilitate systematic integration of disease data into transmission models.
• Reduce boilerplate code associated with developing complex epidemiological models.

Main Methods:

• Developed Broadwick, a modular, object-oriented epidemiological framework.
• Implemented packages for stochastic simulations.
• Integrated parameter inference methods including Approximate Bayesian Computation (ABC) and Markov Chain Monte Carlo (MCMC).

Main Results:

• Broadwick efficiently handles large epidemiological datasets.
• The framework supports customizable simulation and parameter inference algorithms.
• Demonstrated framework utility by developing an SIR model on a simple network.

Conclusions:

• Broadwick enhances researcher productivity by providing a unified platform for developing and sharing complex epidemiological models.
• The framework is suitable for pre-outbreak model development and managing large datasets.
• Broadwick facilitates the creation of adaptable and shareable epidemiological models.