Preprint

Beyond the scientific results, this work shows how modern approaches can open the space of possibilities:

• creating bridges between modeling, AI and the understanding of disease dynamics,
• accelerating inference and scaling analyses to broader, real-world questions,
• capturing uncertainties more effectively and improving forecasting.
  This hybrid vision - AI × spatially structured data analysis × mechanistic models applied to integrated health, host, genetics, habitat and socio-economic data - opens large research avenues and strong innovation potential for preparedness, understanding, and management of pathogen spread in a changing world.
  
  This convergence is central to the projects we are building, from national collaborations to European initiatives, and to the future scientific profiles we want to foster. Our ambition is to contribute actively to this growing ecosystem, where epidemiology, data science and space-aware approaches can meet to give timely and relevant insights to stakeholders.
  
  Francesco’s preprint is one example of how integrating innovative methods can help renew our analytical toolbox and generate the data and insights needed to support operational, territorially grounded, data-driven One-Health and agro-ecological management.

Référence article :
Francesco Pinotti, Julien Thézé, Xavier Bailly, Guillaume Fournié. Simulation based-inference of epidemiological and phylodynamic models via Neural Posterior Estimation. 2026. ⟨hal-05477583⟩