Predicting potential zoonotic hosts

Using Complex Inference Networks to predict zoonotic multi-host systems

Data resources used via GBIF : 37,297 species occurrences
Mexican gray squirrel (Sciurus aureogaster)

Mexican gray squirrel (Sciurus aureogaster) predicted and confirmed to be a Leishmania host. Photo by Alfonso Gutiérrez Aldana licensed under CC BY-NC 4.0.

Zoonoses are diseases that can be transmitted from animals to humans, accounting for more than 60 per cent of human infectious diseases. Host animal range can be a crucial factor in predicting disease risk and potential for interventions, but some zoonoses are complex with not one, but multiple potential hosts and subsequent large ranges.

In this study, researchers from Mexico used Complex Inference Networks to predict hosts of Leishmania parasites, causal agent of Leishmaniasis, that transmit from sand-flies to mammalian hosts, of which only eight have been previously identified.

Using GBIF-mediated occurrences of sand-flies and mammals in Mexico they constructed a database of co-occurrences and inferred potential vector-host interactions. Ranking potential hosts by probability, they collected wildlife specimens, tested them for parasite infection, and found 22 positive species, including thirteen species of bats and one squirrel identified as hosts for the first time.

The study shows how innovative modelling of species occurrences can provide useful information about interactions and be applied to complex problems such as multi-host diseases.


Stephens CR, González-Salazar C, Sánchez-Cordero V, Becker I, Rebollar-Tellez E, Rodríguez-Moreno Á, Berzunza-Cruz M, Domingo Balcells C, Gutiérrez-Granados G, Hidalgo-Mihart M, Ibarra-Cerdeña CN, Ibarra López MP, Iñiguez Dávalos LI and Ramírez Martínez MM (2016) Can You Judge a Disease Host by the Company It Keeps? Predicting Disease Hosts and Their Relative Importance: A Case Study for Leishmaniasis. PLOS Neglected Tropical Diseases. Public Library of Science (PLoS) 10(10): e0005004. Available at: