Disease Outbreaks and Digital Twin Models

This BioDT use case aims to implement a stochastic, spatially-explicit wild boar Sus scrofa - ASF predictive model that is both dynamic and periodic on a structured landscape . Given the input data for a specific study area (a habitat structural model, individual wild boar information, and barrier data). Model components are modularised, and the initial BioDT use case will run the model using a default configuration. The model returns a spatiotemporally dynamic prediction of ASF infection in the modelled wild boar population at weekly time intervals. Users will have access to, and ability to download, a standardised array of model outputs. Additionally, users should have the opportunity to provide additional data inputs—specifically spatiotemporal information about infected wild boar individuals and potential management/containment scenarios. Provision of these additional data will trigger the model to re-run under the influence of the new information. User provided input data and returned model predictions will be saved in a project facilitated repository.

Given the sensitive nature of disease modelling due to its direct relation to human health, much of the data associated with this use case will not be truly FAIR.  For example while wild boar geolocation data may eventually become publicly available (e.g., via GBIF), at the same time the correlated information regarding infection status may remain closed access. This will depend on the original data source. Viral control measures, or barrier scenarios, are unlikely to ever become FAIR, particularly considering that barrier data provided users will likely be exploratory in that potential policy makers will explore a variety of barrier options to predict (evaluate) the efficacy of proposed control measure measures.

The wild boar model itself is already published and documented; the specific implementation (e.g., module configuration) to be used in the BioDT project will be made available with project documentation.