tetano
Editor, Senior Moderator
Transbound Emerg Dis
. 2025 Sep 30:2025:3959370.
doi: 10.1155/tbed/3959370. eCollection 2025. Predictive Modeling of Global SARS-CoV-2 Infection Risk in Animals: Unveiling Potential Reservoirs and Informing Health Policy Synergies
Ruying Fang[SUP] 1 2 [/SUP], Luqi Wang[SUP] 2 [/SUP], Xin Yang[SUP] 2 [/SUP], Yiyang Guo[SUP] 2 [/SUP], Bingjie Peng[SUP] 2 [/SUP], Yinsheng Zhang[SUP] 1 2 [/SUP], Dilinuer Kamili[SUP] 2 [/SUP], Sirui Li[SUP] 3 [/SUP], Yunting Lyv[SUP] 3 [/SUP], Sen Li[SUP] 2 [/SUP], Shunqing Xu[SUP] 4 [/SUP]
Affiliations
Reports of SARS-CoV-2 infections in animals have increasingly raised concerns about potential natural reservoirs for the virus. However, our understanding of the global distribution and drivers of animal infection risk remains limited. To bridge this knowledge gap, we conducted extensive data mining from various sources and developed machine learning (ML) models to estimate the global probability of SARS-CoV-2 infections in animals. We trained and evaluated three ML models, mapping the distribution of infection risk in well-documented regions and projecting risk in areas with sparse infection records. Our models pinpointed high-risk areas in Europe and the United States, where infection records are scattered, as well as in the southern regions of Brazil and Asia, which have sparse infection records. Notably, our projections indicated overlaps between predicted high-risk areas and the known distribution of white-tailed deer, American minks, and Asian small-clawed otters. Anthropogenic factors were found to be more predictive of animal infection than biophysical factors, highlighting the importance of accessibility, population density, and COVID-19 mortality rates. These findings suggest the potential for synergies between public and animal health policies.
Keywords: SARS-CoV-2; animal infection; distribution; machine learning; mapping.
. 2025 Sep 30:2025:3959370.
doi: 10.1155/tbed/3959370. eCollection 2025. Predictive Modeling of Global SARS-CoV-2 Infection Risk in Animals: Unveiling Potential Reservoirs and Informing Health Policy Synergies
Ruying Fang[SUP] 1 2 [/SUP], Luqi Wang[SUP] 2 [/SUP], Xin Yang[SUP] 2 [/SUP], Yiyang Guo[SUP] 2 [/SUP], Bingjie Peng[SUP] 2 [/SUP], Yinsheng Zhang[SUP] 1 2 [/SUP], Dilinuer Kamili[SUP] 2 [/SUP], Sirui Li[SUP] 3 [/SUP], Yunting Lyv[SUP] 3 [/SUP], Sen Li[SUP] 2 [/SUP], Shunqing Xu[SUP] 4 [/SUP]
Affiliations
- PMID: 41063922
- PMCID: PMC12503988
- DOI: 10.1155/tbed/3959370
Reports of SARS-CoV-2 infections in animals have increasingly raised concerns about potential natural reservoirs for the virus. However, our understanding of the global distribution and drivers of animal infection risk remains limited. To bridge this knowledge gap, we conducted extensive data mining from various sources and developed machine learning (ML) models to estimate the global probability of SARS-CoV-2 infections in animals. We trained and evaluated three ML models, mapping the distribution of infection risk in well-documented regions and projecting risk in areas with sparse infection records. Our models pinpointed high-risk areas in Europe and the United States, where infection records are scattered, as well as in the southern regions of Brazil and Asia, which have sparse infection records. Notably, our projections indicated overlaps between predicted high-risk areas and the known distribution of white-tailed deer, American minks, and Asian small-clawed otters. Anthropogenic factors were found to be more predictive of animal infection than biophysical factors, highlighting the importance of accessibility, population density, and COVID-19 mortality rates. These findings suggest the potential for synergies between public and animal health policies.
Keywords: SARS-CoV-2; animal infection; distribution; machine learning; mapping.