Coronavirus: new US data hub is ‘call to action’ for AI experts

The Covid-19 research platform brings together over 29,000 articles and is machine-readable to speed up insights to support the response to the pandemic including the development of a vaccine

A staffer works in the pop-up Huoyan Laboratory specialised in the nucleic acid test on the coronavirus in Wuhan. EPA 

The US rolled out a free open data set of available Covid-19 research to help scientists around the world generate new insights to help fight the pandemic.

The Covid-19 Open Research Dataset, or CORD-19 for short, brings together over 29,000 scholarly articles about Covid-19 for the global research community to use. On Monday, US chief technology officer Michael Kratsios called the resource a "call to action" for artificial intelligence experts who can employ machine learning techniques to surface insights or connect the dots in the enormous cache of data. Sharing information across research communities can speed up the response to the Covid-19 pandemic, according to experts.

As of writing, 185,067 cases have been reported across 155 countries, with 80,236 recoveries and 7,330 deaths. The number of deaths outside China have surpassed those inside the country with the situation declared a pandemic by the World Health Organisation on March 11.

"As of now we are struggling to understand how [Covid-19] is behaving and mutating," Anamika Mehta, an applied molecular biology and clinical genetics specialist in Dubai, told The National. "Centralising knowledge will allow algorithms to be created to help public health officials predict and contain the spread."

The database can also be used to help develop a vaccine and surface effective treatments from clinicians around the world, she added.

Requested by the White House Office of Science and Technology Policy, CORD-19 represents the most extensive machine-readable coronavirus literature collection available for data and text mining to date, according to the White House. The platform was created by researchers and leaders from CZI, the Allen Institute for Artificial Intelligence (A2), Georgetown University’s Centre for Security and Emerging Technology, Microsoft Research, Cold Spring Harbor Laboratory, and the National Library of Medicine of the National Institutes of Health.

In addition to the more than 29,000 articles, the content will continue to be updated as new insights are published in peer-reviewed publications and in archival services, such as the pre-print servers bioRxiv, medRxiv, and others, according to the Chan Zuckerberg Initiative.

There is also now an open call for AI researchers to develop new techniques for text and data mining that will help the medical community comb through the mass of information faster, MIT's Technology Review reported.

Last week, in a similar move to develop more open source tools, the World Economic Forum partnered with the World Health Organisation to roll out a first-of-its-kind digital platform for the private sector's Covid-19 response.

The Pandemic Supply Chain Network is one of the first projects to launch on the Covid Action Platform. It will aim to ensure that essential health commodities are available and affordable. Other actions will include supporting business donations to the public health response and the development of vaccines, diagnostics, treatments and protective equipment, as well as tracking the economic impact of the virus and pursuing collaboration to address disruptions.

The Covid-19 dataset will help medical teams be more proactive, Ms Mehta said, and can continue to be a resource for public health officials as communities begin to recover and put in place policies to prevent future outbreaks.

“Once we overcome it, definitely we will be better prepared for the next one.”