A framework to predict zoonotic hosts under data uncertainty: A case study on betacoronaviruses DOI Creative Commons
Andrea Tonelli, Marcus S. C. Blagrove, Maya Wardeh

et al.

Methods in Ecology and Evolution, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 22, 2025

Abstract Modelling approaches aimed at identifying unknown hosts of zoonotic pathogens have the potential to make high‐impact contributions global strategies for risk surveillance. However, geographical and taxonomic biases in host–pathogen associations affect reliability models their predictions. Here, we propose a methodological framework mitigate effect data account uncertainty models' Our approach involves ‘pseudo‐negative’ species integrating sampling into modelling pipeline. We present an application on genus Betacoronavirus provide estimates mammal‐borne betacoronavirus hazard scale. show that inclusion pseudo‐negatives analysis improved overall validation performance our model when compared does not use pseudo‐negatives, especially reducing rate false positives. Results unveil currently unrecognised hotspots subequatorial Africa Americas. addresses crucial limitations association modelling, with important downstream implications assessments. The proposed is adaptable different multi‐host disease systems may be used identify surveillance priorities as well knowledge gaps pathogens' host‐range.

Language: Английский

A framework to predict zoonotic hosts under data uncertainty: A case study on betacoronaviruses DOI Creative Commons
Andrea Tonelli, Marcus S. C. Blagrove, Maya Wardeh

et al.

Methods in Ecology and Evolution, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 22, 2025

Abstract Modelling approaches aimed at identifying unknown hosts of zoonotic pathogens have the potential to make high‐impact contributions global strategies for risk surveillance. However, geographical and taxonomic biases in host–pathogen associations affect reliability models their predictions. Here, we propose a methodological framework mitigate effect data account uncertainty models' Our approach involves ‘pseudo‐negative’ species integrating sampling into modelling pipeline. We present an application on genus Betacoronavirus provide estimates mammal‐borne betacoronavirus hazard scale. show that inclusion pseudo‐negatives analysis improved overall validation performance our model when compared does not use pseudo‐negatives, especially reducing rate false positives. Results unveil currently unrecognised hotspots subequatorial Africa Americas. addresses crucial limitations association modelling, with important downstream implications assessments. The proposed is adaptable different multi‐host disease systems may be used identify surveillance priorities as well knowledge gaps pathogens' host‐range.

Language: Английский

Citations

0