AI and Data Science lead the way – Outbreak Analysts are creating weapons to fight the Coronavirus
‘Crimson Contagion” – Simulation foreshadows reality
Through the first eight months of 2019, the US city of Chicago was witness to a strange simulation exercise. It involved a scenario where unsuspecting US tourists returning home from China carry along a deadly respiratory virus and spread it throughout the country.The findings were chilling. The model predicted a doomsday scenario with 110 million infections, 7.7 million hospitalizations, and 586,000 deaths in the US alone.
Conducted by the US Department of Health and Human Services, the simulation –named“Crimson Contagion”– used data modelling to visualize the flu pandemic originating in China,assuming Chicago as the host city from where it spreads.
Around midnight of December 30, 2019, the Toronto-based Canadian artificial intelligence (AI) start-up, BlueDot, picked up something unusual happening around a market at Wuhan in China – a cluster of unusual pneumonia conditions. Their data scientists immediately notified their clients via its Insights platform, validating its skills as a global early warning system for infectious diseases – a full nine days before the WHO declared it a global pandemic. The outbreak is said to have originated from the stalls at Wuhan’s booming wet market, selling freshly slaughtered animals to thousands of unsuspecting tourists during the Christmas holiday season.
Within three months of its completion, “Crimson Contagion” had become a reality!
Using AI and Data Science for drug discovery and spread forecast
While BlueDot raised the first red flag, British start-up Exscienta became the first company to put an AI-designed drug molecule to human trials earlier this year. BenevolentAI, based in London, is also using computational and experimental drug discovery to enable scientists uncover new ways to treat diseases and personalize drugs for patients.
Earlier this year, South Korean-based firm Deargen, involved in the “treatment of rare and intractable diseases through the integration of AI and collective intelligence”, also published a paper discussing the results of how strongly a molecule of interest will bind itself to any targeted protein using a deep learning model. By using simplified chemical sequences instead of 2D or 3D chemical structures, Deargen was able to identify four antivirals that may prove successful in binding the Coronavirus.
Identification aside, BlueDot has also been using AI to forecast its spread in real time. According to its website: “BlueDot disseminated bespoke, near-real-time insights to clients including governments, hospitals and airlines, revealing COVID-19’s movements. Our intelligence is based on over 40 pathogen-specific datasets reflecting disease mobility and outbreak potential.”
A new challenge for data scientists
Data Science today has been put in the trenches of technology in leading the global war against the Coronavirus outbreak. AI is being used to develop antibodies and vaccines – scan through millions of existing drugs to see if any of those could be used – and design new drugs to fight the germ. According to analyst firms, this has opened a yet-unforeseen potential for Data Scientists in a world where pandemics are going to recur every few years. In 2002 the world was hit by SARS, in 2009 it was the Swine Flu, followed by MERS in 2012 and Ebola in 2014.
Outbreak Analytics professionals are leading the war against the virus almost as much as healthcare professionals. They are gathering all available data on the epidemic – including confirmed cases, fatalities, test results, tracing contacts of infected people, maps of population densities and demographics, traveller flows and migration, availability of health-care services, drug stockpiles and other factors. The raw data is then being processed into compatible formats by machine-learning software trained to recognize patterns and clean up diverse data sources – and fed into algorithmic models. The models are designed to predict the number of new cases that are likely to arise in an exposed population, or peak infection rates under a given set of conditions, among other outcomes.
The world is witnessing an incredible rapidity of innovation as medical science, epidemiology and data science are converging to stop this deadly virus in its tracks. It’s only now a matter of time before our collective brains win this epic battle to save humankind.
