With the global rise of infectious diseases there is increasing demand for tools that enable us to better predict future disease threats. Ideally, such predictions would enable us to better target ongoing surveillance, and target research to understand when and why spillover infections are likely to occur. This seminar will give examples of predictive analyses that have been undertaken with the aid of machine learning algorithms, which were trained to identify particular wild species that may be the most likely carriers for human infectious pathogens. These analyses have also generated hypotheses about why some species seem to be so much better at carrying zoonotic pathogens compared to others. Some new and ongoing projects will be discussed that expand process-agnostic ecoinformatic methods with existing theory of disease dynamics and life history to build predictive capacity in disease ecology.