Global trade of coronavirus hosts: bringing geographically isolated hosts and viruses together risks novel recombination and spillover to humans
Lead Research Organisation:
University of Liverpool
Department Name: Evolution, Ecology and Behaviour
Abstract
Novel pathogenic coronaviruses - such as SARS-CoV and probably SARS-CoV-2 - arise by two coronaviruses co-infecting viruses a single host cell, and then 'swapping' parts of their genome. The result of this swapping (termed homologous recombination) is a novel daughter virus containing components of each parent virus. These viruses then circulate in reservoir animal populations before spillover to humans.
Our previous work has identified mammalian and avian hosts susceptible to each coronavirus and hosts biologically susceptible to multiple coronavirus strains (recombination hosts).
Here, using a novel ecological network approach, integrating presence, habitat, and behaviour indicating ecological traits of host species, we will build on our previous work to predict contact facilitated sharing of coronaviruses. Combining these predictions with epidemiologically-relevant spatial predictors, will allow us to predict geographical hotspots of coronavirus recombination (Work package 1), and therefore enable specific spatially-targeted surveillance and mitigation efforts.
Many coronavirus hosts interact with humans (e.g. through geographic/habitat overlap), or are used by humans as (e.g. pets/food). By enriching our novel network from WP1 with host species utilisation data from open-access sources, we will estimate the in situ likelihood of spillover from our previously identified hosts (Work package 2). Highlighting priority species and geographic hotspots for spillover mitigation efforts.
Wild animal trade has been implicated in the spillover of coronaviruses, including SARS-CoV-2, to humans. Trade could also readily facilitate homologous recombination in otherwise geographically isolated hosts and their respective coronaviruses. Utilising global animal trade and usages in our predictive framework will enable us to assess the impact of trade on spillover and recombination risk (Work package 3). Understanding the relative risk of wild host trade will allow insight into avoidable human influence on novel coronavirus generation.
Our previous work has identified mammalian and avian hosts susceptible to each coronavirus and hosts biologically susceptible to multiple coronavirus strains (recombination hosts).
Here, using a novel ecological network approach, integrating presence, habitat, and behaviour indicating ecological traits of host species, we will build on our previous work to predict contact facilitated sharing of coronaviruses. Combining these predictions with epidemiologically-relevant spatial predictors, will allow us to predict geographical hotspots of coronavirus recombination (Work package 1), and therefore enable specific spatially-targeted surveillance and mitigation efforts.
Many coronavirus hosts interact with humans (e.g. through geographic/habitat overlap), or are used by humans as (e.g. pets/food). By enriching our novel network from WP1 with host species utilisation data from open-access sources, we will estimate the in situ likelihood of spillover from our previously identified hosts (Work package 2). Highlighting priority species and geographic hotspots for spillover mitigation efforts.
Wild animal trade has been implicated in the spillover of coronaviruses, including SARS-CoV-2, to humans. Trade could also readily facilitate homologous recombination in otherwise geographically isolated hosts and their respective coronaviruses. Utilising global animal trade and usages in our predictive framework will enable us to assess the impact of trade on spillover and recombination risk (Work package 3). Understanding the relative risk of wild host trade will allow insight into avoidable human influence on novel coronavirus generation.
Publications
Blagrove MSC
(2022)
Reply to: Machine-learning prediction of hosts of novel coronaviruses requires caution as it may affect wildlife conservation.
in Nature communications
Wardeh M
(2021)
Predicting mammalian hosts in which novel coronaviruses can be generated.
in Nature communications
Wardeh M
(2021)
Divide-and-conquer: machine-learning integrates mammalian and viral traits with network features to predict virus-mammal associations.
in Nature communications
| Description | The generation and emergence of three novel respiratory coronaviruses from mammalian reservoirs into human populations in the last 20 years, including one which has achieved pandemic status, suggests that one of the most pressing current research questions is: in which reservoirs could the next novel coronaviruses be generated and emerge from in future? Coronaviruses are a family of RNA viruses, which can cause an array of diseases. In humans, these range from mild cold-like illnesses to lethal respiratory tract infections. Seven coronaviruses are known to infect humans1, SARS-CoV, MERS-CoV and SARS-CoV-2 causing severe disease, while HKU1, NL63, OC43 and 229E tend towards milder symptoms in most patients. Homologous recombination is a natural process, which brings together new combinations of genetic material, and hence new viral strains, from two similar non-identical parent strains of virus. This recombination occurs when different strains co-infect an individual animal, with sequences from each parent strain in the genetic make-up of progeny virus. Homologous recombination has previously been shown in Coronavirues, and is implicated in the generation of SARS-CoV-2. The most fundamental requirement for homologous recombination to take place is the co-infection of a single host with multiple coronaviruses. However, our understanding of which hosts are permissive to which coronaviruses, the prerequisite to identifying which hosts are potential sites for this recombination (henceforth termed 'recombination hosts'), was extremely limited. Here, we utilise our similarity-based machine-learning pipeline to address this significant knowledge gap. Our approach predicted associations between coronaviruses and their potential mammalian hosts by integrating three points of view, encompassing: (1) genomic features depicting different aspects of coronaviruses; (2) ecological, phylogenetic and geospatial traits of potential mammalian hosts; and (3) characteristics of the network that describes the linkage of coronaviruses to their observed hosts. Using this pipeline, we demonstrated that there is currently a large underappreciation of the potential scale of novel coronavirus generation in wild and domesticated animals. Specifically, we predicted there are 11.5-fold more coronavirus-host associations, over 30-fold more potential SARS-CoV-2 recombination hosts, and over 40-fold more host species with four or more different subgenera of coronaviruses than have been observed to date (at >0.5 mean probability cut-off). We discussed the large number of candidate hosts in which homologous recombination of coronaviruses could result in the generation of novel pathogenic strains, as well as the substantial underestimation of the range of viruses which could recombine based on observed data. Give that coronaviruses frequently undergo homologous recombination when they co-infect a host, and that SARS-CoV-2 is highly infectious to humans, the most immediate threat to public health is recombination of other coronaviruses with SARS-CoV-2. Such recombination could readily produce further novel viruses with both the infectivity of SARS-CoV-2 and additional pathogenicity from other Coronaviruses. Taking only observed data, there are four non-human mammalian hosts known to associate with SARS-CoV-2 and at least one other coronavirus, and a total of 504 different unique interactions between SARS-CoV-2 and other coronaviruses. However, when we add in our model's predicted interactions this becomes 126 SARS-CoV-2 hosts and 2544 total unique interactions; indicating that observed data are missing 31.5-fold of the total number of predicted recombination hosts, and 5.05-fold increase of the predicted unique associations. Furthermore, we predicted that there are 11.5-fold more general coronavirus-host associations, and over 40-fold more host species with four or more different subgenera of coronaviruses than have been observed to date. Our results demonstrate the large underappreciation of the potential scale of novel coronavirus generation in wild and domesticated animals. With the greater understanding of the extent of mammalian host reservoirs and the potential recombination hosts we identify here, a targeted surveillance programme is now possible which would allow for this generation to be observed as it is happening and before a major outbreak. Such information could help inform prevention and mitigation strategies and provide a vital early warning system for future novel coronaviruses. |
| Exploitation Route | With the greater understanding of the extent of mammalian host reservoirs and the potential recombination hosts we identify here, a targeted surveillance programme is now possible which would allow for this generation to be observed as it is happening and before a major outbreak. Such information could help inform prevention and mitigation strategies and provide a vital early warning system for future novel coronaviruses. As an example of the utilisation of our model from the perspective of likely future viral homologous recombination events, Banerjee et al. bioinformatically identified potential genomic regions of homologous recombination between MERS-CoV and SARS-CoV-2. They highlighted a significant risk of the highly human-to-human transmissible SARS-CoV-2 acquiring the considerably more pathogenic phenotypes of MERS-CoV. The work presented here identifies 102 potential recombination hosts (excluding humans and laboratory rodents) of the two viruses. Together, our work and Banerjee et al., we provide evidence for both the possible production of a potentially severe future recombinant coronavirus and identify the hosts in which this threat is most likely to be generated. We recommend monitoring for this event. |
| Sectors | Environment,Healthcare |
| URL | https://www.nature.com/articles/s41467-021-21034-5 |
| Description | 74 news stories/radio interviews have been written/performed about this work. Through collaborators Species360, we are currently discussing a policy document with the WHO regarding trade of animal hosts of the WHO blueprint diseases |
| First Year Of Impact | 2021 |
| Sector | Agriculture, Food and Drink,Environment,Healthcare |
| Impact Types | Policy & public services |
| Description | Vector in the machine: How accurately can mosquito transmission of viruses be predicted by machine learning? |
| Amount | £181,248 (GBP) |
| Organisation | Medical Research Council (MRC) |
| Sector | Public |
| Country | United Kingdom |
| Start | 10/2022 |
| End | 09/2026 |
| Title | Data and code for: "Monkeypox virus shows potential to infect a diverse range of native animal species across Europe, indicating high risk of becoming endemic in the region." |
| Description | Background: Monkeypox is a zoonotic virus which persists in animal reservoirs and periodically spills over into humans, causing outbreaks. During the current 2022 outbreak, monkeypox virus has persisted via human-human transmission, across all major continents and for longer than any previous record. This unprecedented spread creates the potential for the virus to 'spillback' into local susceptible animal populations. Persistent transmission amongst such animals raises the prospect of monkeypox virus becoming enzootic in new regions. However, the full and specific range of potential animal hosts and reservoirs of monkeypox remains unknown, especially in newly at-risk non-endemic areas. Methods: Here, our pipeline utilises ensembles of classifiers comprising different class balancing techniques and incorporating instance weights, to identify which animal species are potentially susceptible to monkeypox virus. Subsequently, we generate spatial distribution maps to highlight high-risk geographic areas at high resolution. Findings: We show that the number of potentially susceptible species is currently underestimated by 2.4 to 4.3-fold. We show a high density of susceptible wild hosts in Europe. We provide lists of these species, and highlight high-risk hosts for spillback and potential long-term reservoirs, which may enable monkeypox virus to become endemic. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2022 |
| Provided To Others? | Yes |
| Impact | N/A |
| URL | https://figshare.com/articles/software/Blagrove_et_al_2022_poxvriuses_data_and_code/20485332 |
| Title | Divide-and-conquer: data and codes |
| Description | Data and codes associated with: Divide-and-conquer: Wardeh, M., Blagrove, M.S.C., Sharkey, K.J. et al. Divide-and-conquer: machine-learning integrates mammalian and viral traits with network features to predict virus-mammal associations. Nat Commun 12, 3954 (2021). https://doi.org/10.1038/s41467-021-24085-w. Abstract: Our knowledge of viral host ranges remains limited. Completing this picture by identifying unknown hosts of known viruses is an important research aim that can help identify and mitigate zoonotic and animal-disease risks, such as spill-over from animal reservoirs into human populations. To address this knowledge-gap we apply a divide-and-conquer approach which separates viral, mammalian and network features into three unique perspectives, each predicting associations independently to enhance predictive power. Our approach predicts over 20,000 unknown associations between known viruses and susceptible mammalian species, suggesting that current knowledge underestimates the number of associations in wild and semi-domesticated mammals by a factor of 4.3, and the average potential mammalian host-range of viruses by a factor of 3.2. In particular, our results highlight a significant knowledge gap in the wild reservoirs of important zoonotic and domesticated mammals' viruses: specifically, lyssaviruses, bornaviruses and rotaviruses. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| Impact | The multi-perspective host-pathogen predictive framework undperins the following research awards: Where coronaviruses hide, where novel strains are generated, and how they get to us: Predicting reservoirs, recombination, and geographical hotspots (NE/W002302/1); Global trade of coronavirus hosts: bringing geographically isolated hosts and viruses together risks novel recombination and spillover to humans (BB/W00402X/1); and Predicting mammalian and avian reservoirs of coronaviruses: identifying current reservoirs and co-infection hosts in which future novel coronavirus could be generated (BBSRC IAA COVID - 168478) |
| URL | https://doi.org/10.6084/m9.figshare.13270304 |
| Title | Predicting mammalian hosts in which novel coronaviruses can be generated - codes |
| Description | Novel pathogenic coronaviruses - such as SARS-CoV and probably SARS-CoV-2 - arise by homologous recombination between co-infecting viruses in a single cell. Identifying possible sources of novel coronaviruses therefore requires identifying hosts of multiple coronaviruses; however, most coronavirus-host interactions remain unknown. This novel method, deploys a meta-ensemble of similarity learners from three complementary perspectives (viral, mammalian and network), topredict which mammals are hosts of multiple coronaviruses. The results predict that there are 11.5-fold more coronavirus-host associations, over 30-fold more potential SARS-CoV-2 recombination hosts, and over 40-fold more host species with four or more different subgenera of coronaviruses than have been observed to date at >0.5 mean probability cut-off (2.4-, 4.25- and 9-fold, respectively, at >0.9821). |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| Impact | These models undperin the following research awards: Where coronaviruses hide, where novel strains are generated, and how they get to us: Predicting reservoirs, recombination, and geographical hotspots (NE/W002302/1); Global trade of coronavirus hosts: bringing geographically isolated hosts and viruses together risks novel recombination and spillover to humans (BB/W00402X/1); and Predicting mammalian and avian reservoirs of coronaviruses: identifying current reservoirs and co-infection hosts in which future novel coronavirus could be generated (BBSRC IAA COVID - 168478) |
| URL | https://figshare.com/articles/software/covs-recombination-hosts/13110896 |
| Title | covs-recombination-hosts |
| Description | Data and codes associated with Wardeh et al - Predicting mammalian hosts in which novel coronaviruses can be generated.authors:Maya WardehMatthew BaylsMarcus Blagroveplease email maya wardeh: maya.wardeh@liverpool.ac.uk with any queries or requests. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| Impact | N/A |
| URL | https://figshare.com/articles/software/covs-recombination-hosts/13110896/1 |
| Description | Sapienza University of Rome |
| Organisation | Sapienza University of Rome |
| Country | Italy |
| Sector | Academic/University |
| PI Contribution | Exploring the mammalian virome to detect patterns of compatibility between mammal species and viruses at a global scale, identifying eco-biological profiles of viral carriers along the fast-slow continuum of mammalian life-history. |
| Collaborator Contribution | Provided insight into virus-mammal interactions, and role virus traits have on the transmission/spill-over of viruses. |
| Impact | Publication: Identifying patterns along the fast-slow continuum of mammalian viral carriers (in prep/under review) |
| Start Year | 2022 |
| Description | Species360 - Impact of global trade in wildlife on virus spread. |
| Organisation | IDAs og Berg-Nielsens Studie-og støttefond |
| Country | Denmark |
| Sector | Charity/Non Profit |
| PI Contribution | This partnership aims at quantifying the impact of global trade in wild animals on the potential spread of emerging infectious zoonses prioritized by the WHO Research and Development Blueprint Strategy. Our group provided data on animal species in which these zoonoses has been found to date; and the ecological role of these animals in the transmission of these pathogens (such as reservoirs; dead-end; and amplifying hosts), as well as information on how these pathogens manifest in the animal host (mortality, morbidity, minor symptoms, or no disease). |
| Collaborator Contribution | Our partners are leading on the analyses to identify geographical patterns of trade in the animals identified above, and the key data gaps that need to be resolved to fully assess risks from international wildlife trade and put this information in the context of other drivers of zoonotic diseases. |
| Impact | NA |
| Start Year | 2020 |
| Title | Blagrove_et_al_2022_poxvriuses data and code |
| Description | Data and code for: "Monkeypox virus shows potential to infect a diverse range of native animal species across Europe, indicating high risk of becoming endemic in the region." Authors: Marcus SC Blagrove*1, Jack Pilgrim1, Aurelia Kotsiri1, Melody Hui1, Matthew Baylis1, Maya Wardeh*1,2 * = corresponding authors 1) Institute of Infection, Veterinary and Ecological Sciences, University of Liverpool, Liverpool, UK 2) Department of Mathematical Sciences, University of Liverpool, Liverpool, UK Abstract: Background: Monkeypox is a zoonotic virus which persists in animal reservoirs and periodically spills over into humans, causing outbreaks. During the current 2022 outbreak, monkeypox virus has persisted via human-human transmission, across all major continents and for longer than any previous record. This unprecedented spread creates the potential for the virus to 'spillback' into local susceptible animal populations. Persistent transmission amongst such animals raises the prospect of monkeypox virus becoming enzootic in new regions. However, the full and specific range of potential animal hosts and reservoirs of monkeypox remains unknown, especially in newly at-risk non-endemic areas. Methods: Here, our pipeline utilises ensembles of classifiers comprising different class balancing techniques and incorporating instance weights, to identify which animal species are potentially susceptible to monkeypox virus. Subsequently, we generate spatial distribution maps to highlight high-risk geographic areas at high resolution. Findings: We show that the number of potentially susceptible species is currently underestimated by 2.4 to 4.3-fold. We show a high density of susceptible wild hosts in Europe. We provide lists of these species, and highlight high-risk hosts for spillback and potential long-term reservoirs, which may enable monkeypox virus to become endemic. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | None yet |
| URL | https://figshare.com/articles/software/Blagrove_et_al_2022_poxvriuses_data_and_code/20485332/2 |
| Title | Blagrove_et_al_2022_poxvriuses data and code |
| Description | Data and code for: "Monkeypox virus shows potential to infect a diverse range of native animal species across Europe, indicating high risk of becoming endemic in the region." Authors: Marcus SC Blagrove*1, Jack Pilgrim1, Aurelia Kotsiri1, Melody Hui1, Matthew Baylis1, Maya Wardeh*1,2 * = corresponding authors 1) Institute of Infection, Veterinary and Ecological Sciences, University of Liverpool, Liverpool, UK 2) Department of Mathematical Sciences, University of Liverpool, Liverpool, UK Abstract: Background: Monkeypox is a zoonotic virus which persists in animal reservoirs and periodically spills over into humans, causing outbreaks. During the current 2022 outbreak, monkeypox virus has persisted via human-human transmission, across all major continents and for longer than any previous record. This unprecedented spread creates the potential for the virus to 'spillback' into local susceptible animal populations. Persistent transmission amongst such animals raises the prospect of monkeypox virus becoming enzootic in new regions. However, the full and specific range of potential animal hosts and reservoirs of monkeypox remains unknown, especially in newly at-risk non-endemic areas. Methods: Here, our pipeline utilises ensembles of classifiers comprising different class balancing techniques and incorporating instance weights, to identify which animal species are potentially susceptible to monkeypox virus. Subsequently, we generate spatial distribution maps to highlight high-risk geographic areas at high resolution. Findings: We show that the number of potentially susceptible species is currently underestimated by 2.4 to 4.3-fold. We show a high density of susceptible wild hosts in Europe. We provide lists of these species, and highlight high-risk hosts for spillback and potential long-term reservoirs, which may enable monkeypox virus to become endemic. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | None yet |
| URL | https://figshare.com/articles/software/Blagrove_et_al_2022_poxvriuses_data_and_code/20485332/1 |
| Title | Blagrove_et_al_2022_poxvriuses data and code |
| Description | Data and code for: "Monkeypox virus shows potential to infect a diverse range of native animal species across Europe, indicating high risk of becoming endemic in the region." Authors: Marcus SC Blagrove*1, Jack Pilgrim1, Aurelia Kotsiri1, Melody Hui1, Matthew Baylis1, Maya Wardeh*1,2 * = corresponding authors 1) Institute of Infection, Veterinary and Ecological Sciences, University of Liverpool, Liverpool, UK 2) Department of Mathematical Sciences, University of Liverpool, Liverpool, UK Abstract: Background: Monkeypox is a zoonotic virus which persists in animal reservoirs and periodically spills over into humans, causing outbreaks. During the current 2022 outbreak, monkeypox virus has persisted via human-human transmission, across all major continents and for longer than any previous record. This unprecedented spread creates the potential for the virus to 'spillback' into local susceptible animal populations. Persistent transmission amongst such animals raises the prospect of monkeypox virus becoming enzootic in new regions. However, the full and specific range of potential animal hosts and reservoirs of monkeypox remains unknown, especially in newly at-risk non-endemic areas. Methods: Here, our pipeline utilises ensembles of classifiers comprising different class balancing techniques and incorporating instance weights, to identify which animal species are potentially susceptible to monkeypox virus. Subsequently, we generate spatial distribution maps to highlight high-risk geographic areas at high resolution. Findings: We show that the number of potentially susceptible species is currently underestimated by 2.4 to 4.3-fold. We show a high density of susceptible wild hosts in Europe. We provide lists of these species, and highlight high-risk hosts for spillback and potential long-term reservoirs, which may enable monkeypox virus to become endemic. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | None yet |
| URL | https://figshare.com/articles/software/Blagrove_et_al_2022_poxvriuses_data_and_code/20485332 |
| Description | AI used to 'predict the next coronavirus'. |
| Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | BBC Radio 4, Inside science |
| Year(s) Of Engagement Activity | 2021 |
| URL | http://www.bbc.co.uk/news/science-environment-56076716?fbclid=IwAR0AIa3il2XTl6QbrIlP7aCovc79To-tjPIa... |
| Description | Potential for New Coronaviruses May Be Greater Than Known. |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | Newspaper report in NYT |
| Year(s) Of Engagement Activity | 2021 |
| URL | https://www.nytimes.com/2021/02/16/science/Covid-reemerging-viruses.html?fbclid=IwAR0AIa3il2XTl6QbrI... |
| Description | Quand l'IA part à la chasse au prochain coronavirus chez les mammifères. |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | Newspaper article in France24 |
| Year(s) Of Engagement Activity | 2021 |
| URL | https://www.france24.com/fr/%C3%A9co-tech/20210218-quand-l-ia-part-%C3%A0-la-chasse-au-prochain-coro... |
| Description | The Coronavirus Menagerie - New York Times coverage |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | Newspaper report in NYT. |
| Year(s) Of Engagement Activity | 2022 |
| URL | https://www.nytimes.com/2022/02/22/health/coronavirus-animals.html |
| Description | The Search for Animals That Could Carry the Next Deadly Virus - Wall Street Journal Interview |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | Media interview with Wall Street Journal. |
| Year(s) Of Engagement Activity | 2021 |
| URL | https://www.wsj.com/articles/the-search-for-animals-that-could-carry-the-next-deadly-virus-116166736... |
| Description | Virologists use AI to work on next pandemic outbreak. |
| Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | Radio interview in New Zealand |
| Year(s) Of Engagement Activity | 2021 |
| URL | https://www.rnz.co.nz/national/programmes/first-up/audio/2018783912/virologists-use-ai-to-work-on-ne... |