Mathematical Modelling Demonstration Software: Animal Epidemic!

Mathematical models are used within the SEFARI as tools to understand and predict the spread and control of infectious disease. Communicating the concepts and outputs associated with these models to the public, farmers and industry stakeholders is a significant challenge, since the formal epidemiological, mathematical and computational concepts are highly technical. Interactive simulations with effective graphical interfaces can support public understanding and engagement with mathematical models. We propose to develop a web app to facilitate interactive visualisation of exemplar disease systems. This app will support public engagement and knowledge exchange activities, promoting wider awareness and understanding of SEFARI work.

Objectives

1. Identify an infectious animal disease that is familiar to the general public, with potential to motivate a narrative which encapsulates key KTE messages.
2. Develop a simulation engine with capacity to generate realisations of disease outbreaks relevant to the selected narrative.
3. Construct an interactive narrative which, via ‘game-play’, allows users to interact with simulated spread/control of the disease outbreak, providing a platform to inform them about the control of disease.
4. Design interactive visuals that enable users to interact with the simulation in an engaging, probably ‘game-like’ manner e.g. by choosing which control measures to implement when trying to contain the simulated disease outbreak.
5. Develop a web app implementation of the narrative and visuals.
6. Identify an initial list of opportunities at which the software will be deployed and develop support materials as necessary.

Motivation/problem: Mathematical modelling is a widely used tool in animal and plant epidemiology, influencing policy-making and advice to industry. Effective communication of the concepts and outputs of mathematical models represents a challenge when the audience is not familiar with the use of such models. Obstacles must be overcome to support farmers in understanding why disease control policies have been implemented, where these policies have been informed by mathematical modelling. More generally, there is an exciting story to tell about mathematical modelling, how it can be applied to understand and control disease and why this work is beneficial to society.

The challenges of communicating mathematical modelling ideas to generalist audiences are not well addressed by the KTE requirements within the SRP and other SG-funded research. However, much of the science within the SRP depends on outputs from mathematical modelling, so we believe that it will be beneficial to develop a web app for interactive visualisation of mathematical models; initially focussing on a disease scenario that is familiar to the general public e.g. spread/control of a Foot and Mouth Disease outbreak.

Our proposed solution: We propose to build an interactive web app using the Java Broadwick platform (https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-016-0903-2), which was developed in the EPIC Centre of Expertise. Broadwick has the capacity to simulate sophisticated epidemiological models. A Broadwick simulation engine, to perform the disease simulations, will be developed by a contractor. Staff at BioSS will develop a web app front-end to the simulation engine, providing visualisation of the simulation and allowing the user to apply control measures via a Graphical User Interface to try to contain the simulated outbreak. The resulting software will be hosted on the BioSS website and could also be deployed on the SEFARI Gateway website.

The web app will initially be designed to simulate an outbreak of an infectious animal disease familiar to the general public. The user can choose a control strategy, subject to resource limitations, and then ‘play the game’ (run a simulation of disease spread/control). Once the simulated outbreak is complete the user is given a score to summarise the performance of the chosen control measures. The interaction between user and web app will lead to numerous learning opportunities e.g. comparison of different control measures and the limitations imposed by available resources. The target audience is the general public.

The software will be designed to implement the disease scenario outlined above but could subsequently be extended to implement alternative disease scenarios and narratives, highlighting other messages, and targeting other audiences e.g. farmers or vets.

Given the focus on animal disease, the project team has been chosen from BioSS, MRI and SRUC, the three SEFARI Institutes with more expertise in animal infectious disease epidemiology and modelling, and in engagement with the livestock sector. During the project, a modelling group will specify the required mathematical models, while a user experience group will focus on the ‘in-game narrative’ and contextual material. The project management group will oversee all aspects of delivery including management of the contractor.

External collaboration and funding
The work proposed in this project is highly relevant to the activities of the SG Centre of Expertise in Animal Disease Outbreaks (EPIC). In fact, several of the SEFARI staff contributing to this proposal work within EPIC. EPIC has agreed to support this proposal and will assist in the design and specification of the disease scenario and narrative, as well as in developing other scenarios for future use.

Besides producing an app which will be of use in its own right, we see this project acting as a pilot exercise which could help to leverage funding for future projects. For example, we think that there will be value in applying the app to simulations of more complex animal disease scenarios with more specialist target audiences such as farmers, vets or policy makers. There is also potential for future links with the SG Centre of Expertise in Plant Health.