The battle for evolution is the starkest battle: 60% of the roughly 400 emerging infectious diseases are zoonotic, and this figure is just increasing. COVID-19 has reminded the world we are just one step away from being caught by microbial evolution. Historically, the emergence of new epidemics has caused significant human and socioeconomic burden. Scientific evidence (Morse, 2004; Morse et al., 2012, Allen et al., 2017) provides insights into the factors that influence the epidemic emergence, the so-called drivers of emergence, and alerts on the impact of anthropogenic modifications on the environment. The scientific consensus agrees on the rather urgent need of developing a tool able to strengthen the prediction of potential upcoming epidemics.
Pandora team has engineered the Epidemic Emergence Monitor bolstered by the principle by which outbreak prediction is the best protection. A successful model for epidemics prediction should take into consideration how pathogens evolve to escape from their natural hosts and colonise human beings.
Epidemic risk is therefore a function of the virus/host contact frequency and its successful adaptation to humans in a sustained manner. Samely, the contact frequency depends not only on the mere physical proximity between the viral and human environments, but a successful adaptation to humans depends upon the virus/host relatedness, the virus’ host range and plasticity, the patterns of host/virus co-evolution, and the predicted virulence in humans,
The infectious diseases can be viewed operationally as a two step process, introduction and dissemination: 1) introduction of the new agent into a new host population and 2) dissemination of the agent into the new host population. We designate ‘microbial traffic’ as the process by which infectious agents become disseminated from isolated groups into new populations. The greater the microbial traffic, the higher the risk for disease emergence and potential epidemics.
The factors and determinants of disease emergence, or drivers of emergence, can be distributed into 3 groups: ecological, environmental and demographic. These place people closer to the natural host for a previously unfamiliar zoonotic agent or that promote the spread of the pathogen. As these factors are increasingly prevalent, and zoonotic epidemics keep appearing with massive economic and human burden, there is a need to monitor the variance of factors and determinants of disease emergence.
The main inspiration for choosing this challenge was our shared interest in helping to prevent the next pandemic by tackling it at its source. The best way of approaching this goal from a data-driven perspective was to identify likely future epidemic emerging hotspots upon which prevention or containment efforts should be focused.
For a global problem such as the one we have committed to address, the scope must be global too. Data collected from space is a unique tool to contribute to our aim, as it is unbiased by economy or frontiers. The Emerging Epidemics Monitor (EEM) makes extensive use of Earth observation data, amongst other sources, to address the risk of new zoonotic epidemics on a global scale.
Space-based assets for land cover, deforestation, fires, and floods have been combined with other global-scale datasets, such as mammalian biodiversity, population density, and wet markets location in order to compose various factors measuring local drivers of disease emergence:
These factors were again combined to arrive at a single magnitude measuring the risk of emergence of a new zoonotic disease at a given location: the Emerging Epidemics Monitor, a world map of the risk of zoonotic-origin epidemic emergence.
Some of the technologies that made this possible were:
One of the main technical difficulties faced during the project was the conversion of multiple datasets, with different formats, resolutions, and map projections into a shared standard for combining them. In the end, the exceptional Open Source data handling tools and libraries built around Python saved the day.
We are a highly multidisciplinary team, from aeroespacial engineers to data scientists, biochemists, senior UX designers, and machine learning engineers. Beyond our passion for innovation, we wholeheartedly believe in delivering real solutions: at Pandora, we strive for providing the community with a novel tool able to integrate the key drivers of epidemic emergence.
Data Sources
The main source for both space-imagery based and socioeconomic datasets come from NASA SEDAC (Socioeconomic Data and Applications Center).
Additionally, our project integrates data from the EarthEnv Global 1-km Consensus Land Cover and the University of Maryland Global Forest Change projects. Finally, researchers from the Institute of Global Health from the University of Geneva and other institutions kindly provided access to a dataset with locations of wet markets around the globe.
All papers of the data sources related to these datasets have been cited below.
International Union for Conservation of Nature - IUCN, and Center for International Earth Science Information Network - CIESIN - Columbia University. 2015. Gridded Species Distribution: Global Mammal Richness Grids, 2015 Release. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). https://doi.org/10.7927/H4N014G5. Accessed 30/05/2020.
Tuanmu, M.‐N. and Jetz, W. (2014), Consensus land cover. Global Ecology and Biogeography, 23: 1031-1045. doi:10.1111/geb.12182. Accessed 30/05/2020.
Hansen, M., Potapov, P., and Moore, R.: High-resolution global maps of 21st century forest cover change, Science, 342, 850–853, 2013. Accessed 30/05/2020.
Ramankutty, N., A.T. Evan, C. Monfreda, and J.A. Foley. 2010. Global Agricultural Lands: Pastures, 2000. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). https://doi.org/10.7927/H47H1GGR. Accessed 30/05/2020.
Ramankutty, N., A.T. Evan, C. Monfreda, and J.A. Foley. 2008. Farming the Planet: 1. Geographic Distribution of Global Agricultural Lands in the Year 2000. Global Biogeochem. Cycles 22 (1): GB1003.https://doi.org/10.1029/2007GB002952. Accessed 30/05/2020.
Center for Hazards and Risk Research - CHRR - Columbia University, and Center for International Earth Science Information Network - CIESIN - Columbia University. 2005. Global Flood Hazard Frequency and Distribution. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). https://doi.org/10.7927/H4668B3D. Accessed 30/05/2020.
Dilley, M., R.S. Chen, U. Deichmann, A.L. Lerner-Lam, M. Arnold, J. Agwe, P. Buys, O. Kjekstad, B. Lyon, and G. Yetman. 2005. Natural Disaster Hotspots: A Global Risk Analysis. Washington, D.C.: World Bank.https://doi.org/10.1596/0-8213-5930-4. Accessed 30/05/2020.
Center for International Earth Science Information Network - CIESIN - Columbia University. 2018. Gridded Population of the World, Version 4 (GPWv4): Population Density, Revision 11. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). https://doi.org/10.7927/H49C6VHW. Accessed 30/05/2020.
Yetman, G., S.R. Gaffin, and X. Xing. 2004. Global 15 x 15 Minute Grids of the Downscaled GDP Based on the SRES B2 Scenario, 1990 and 2025. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC).https://doi.org/10.7927/H4NC5Z4X. Accessed 30/05/2020.
Gaffin, S.R., C. Rosenzweig, X. Xing, and G. Yetman. 2004. Downscaling and Geo-spatial Gridding of Socio-economic Projections from the IPCC Special Report on Emissions Scenarios (SRES). Global Environmental Change 14 (2): 105-123. https://doi.org/10.1016/j.gloenvcha. 2004.02.004. Accessed 30/05/2020.
Giglio, L., J. T. Randerson, and G. R. van der Werf. 2018. Global Fire Emissions Indicators, Grids: 1997-2015. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC).https://doi.org/10.7927/H400002V. Accessed 30/05/2020.
Giglio, L., J. T. Randerson, and G. R. van der Werf. 2013. Analysis of Daily, Monthly, and Annual Burned Area Using the Fourth-Generation Global Fire Emissions Database (GFED4). Journal of Geophysical Research 118 (1): 317-328. https://doi.org/10.1002/jgrg.20042. Accessed 30/05/2020.
Kogan, N. E., Bolon, I., Ray, N., Alcoba, G., Fernandez-Marquez, J. L., Müller, M. M., Mohanty, S. P., & Ruiz de Castañeda, R. (2019). Wet Markets and Food Safety: TripAdvisor for Improved Global Digital Surveillance. JMIR public health and surveillance, 5(2), e11477. https://doi.org/10.2196/11477
Scientific Papers
Morse S. Factors and determinants of disease emergence. Rev. sci. tech. Off. int. Epiz., 2004, 23 (2), 443-451.
https://drive.google.com/open?id=15DV8demLnjRy8g9yJyFWz5b8mRbsP-SL
Morse SS, Mazet JAK, Woolhouse M, et al. Prediction and prevention of the next pandemic zoonosis. Lancet 2012; 380: 1956–65.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3712877/pdf/main.pdf
Allen T, Murray KA, Zambrana-Torrelio C, et al. Global hotspots and correlates of emerging zoonotic diseases. Nature Communications 2017; 8:1123. https://www.nature.com/articles/s41467-017-00923-8.pdf
Shengal RNM. Deforestation and avian infectious diseases. J Exp Biol. 2010 Mar 15; 213(6): 955–960.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2829318/
Brock PM, Fornace KM, Grigg MJ, et al. Deforestation and disease dynamics. Royal Society 2018; 286:1894
https://royalsocietypublishing.org/doi/full/10.1098/rspb.2018.2351
Afelt A, Frutos R, and Devaux C. Bats, Coronaviruses, and Deforestation: Toward the Emergence of Novel Infectious Diseases? Front Microbiol. 2018; 9: 702.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5904276/
Yanga S, Santillana M, and Koua SC. Predicting influenza outbreaks using google data. PNAS 2015, 112:47.
https://www.pnas.org/content/pnas/112/47/14473.full.pdf
Hansen M.C., Potapov P., Moore R, et al. High Resolution Global Forest Change. Science 2013, 342(6160):850-853. https://www.researchgate.net/publication/258529161_High-Resolution_Global_Maps_of_21st-Century_Forest_Cover_Change