Emerging viral diseases rarely arise from nowhere. Most originate in animal reservoirs—wild or domestic species where the virus persists without causing severe disease. Yet identifying which animal populations might harbour a dangerous pathogen, or which new host a virus might successfully invade, remains extraordinarily difficult. Field-based ecological mapping is expensive, time-consuming, and often unsafe. Researchers from Makerere University and the Interdisciplinary Consortium for Epidemics Research have developed a theoretical computational approach called genography and genomography—a method to predict, in-silico, which species across the animal kingdom are most likely to be suitable hosts for a given virus.

Using comparative genomic analysis and BLAST algorithms calibrated for distant evolutionary relationships, genography scans for “evolutionary acceptors” in host genomes: regions of homology to viral genes that may indicate where viral integration could occur. Drawing on evidence from influenza, filoviruses (Ebola and Marburg), and other RNA viruses, the framework lays out algorithms for interpreting these signals and estimating which host species sit closest to a virus’s natural reservoir—and therefore, which ones pose the greatest spillover risk.

This collection of papers and working notes documents the development of genography as a surveillance and prediction tool, with implications for understanding how viruses cross species barriers and for designing interventions before outbreaks reach humans.

The full manuscript is available to read below.