Can AI help in preventing spread of Coronavirus?

Published in

World AI

5 min readFeb 9, 2020

A look into the genetic sequence of the Coronavirus reveals that Coronaviruses possess the largest genomes (26.4 to 31.7 kb) among all known RNA viruses, with G + C contents varying from 32% to 43%.

Traditionally, viruses have been characterized and classified by culture, electron microscopy, and serological studies. Using these phenotypic methods, coronaviruses are defined as enveloped viruses of 120–160 nm in diameter with a crown-like appearance. The name “coronavirus” is derived from the Greek κορώνα, meaning crown.

Molecular clock analysis using various gene loci revealed that the time of most recent common ancestor of human/civet SARS related coronavirus to be in the years ranging from 1999–2002.

The large genome has given the Coronavirus extra plasticity in accommodating and modifying genes. This property allows it to adjust itself to latch onto newer host organisms. This characteristic of the coronavirus also makes it hard and complicated to find an effective cure. The virus can necessarily mutate itself to navigate around the immune system response of its host, thereby resulting in longer survival rates of the infection itself and adding to its ability to spread quickly by adapting itself to various host organisms.

Evolutionary rate and divergence

In 1992, Sanchez et al., in a Virology publication in 1992, analyzed 13 enteric and respiratory TGEV related isolates and estimated the mutation rate of TGEV to be 7x10^-4 nucleotide substitutions per site per year. 1n 2005, using linear regression, maximum likelihood, and Bayesian inference methods, Vijgen et al. estimated the rate of evolution in BCoV to be 4.3x10^-4 nucleotide substitutions per site. This study was with a confidence interval of 95%.

Subsequently, when various novel human and animal coronaviruses were discovered, evolutionary rates and divergence time in the Coronaviridae family were estimated by multiple groups using different approaches. Although Bayesian inference in BEAST is probably the most widely accepted approach and was used by most researchers, the use of different genes and datasets by various groups has resulted in a considerable difference in the estimated history of coronaviruses. One group, using the helicase gene for analysis, has estimated the life history of coronaviruses to be as short as about 420 years. However, across studies, a divergence of up to 100 years is observed.

Origin

The genetic analysis finds the origin of the Coronavirus to be Bats. The virus is 80% similar to the SARC virus, which, too, originated from Bats. The virus was first reported in Wuhan, the sprawling capital of Central China’s Hubei province.

The first human infection from the virus seems to have come from the Huanan wholesale seafood market in Wuhan. Post this discovery, the market was closed down and tapped to prevent people from entering it.

Transfer to Mankind

Vincent Munster, a virologist at Rocky Mountain Laboratories said, only the beta coronaviruses can make the jump to humans and nest in our respiratory tracts. This is so as not all Coronaviruses have the same shape.

The circular shell of the Coronavirus is peppered with spike-shaped proteins. These spikes help it to attach to host cells. If the viruses spike shape doesn’t fit receptors on a potential host, the virus cannot attach onto the host to infect it. However, when the virus mutates itself, the shape of the proteins gets altered. This alteration sometimes allows the virus to dock onto new hosts that it previously could not dock onto. The current structure of the Coronavirus has a mutation that helps its protein spikes to latch onto receptors in the human respiratory system. The virus thereby attacks the respiratory system, causing patients to build respiratory problems, with sometimes the infection escalating to being lethal.

Current Spread

The virus is spreading at an alarming rate. It started with less than 100 people in Wuhan in Dec 2019, and by Feb 2020, it has affected more than 30,000 across the globe. Most reported infections are from within Mainland China. While the virus is spreading rapidly, it’s lethal properties are not very severe. Many can fight it off using their natural immunity.

Courtesy: The New York Times (spread as of 09 Feb 2020)

Live spread stats maintained and updated daily by The New York Times can be found here.

Stop it or slow it?

The virus is readily transmitted from person to person interaction. There is simply nothing that can be done to stop its spread. However, taking measures such as shutting down public transport systems, quarantine the infected, staying at home, and restricting cross country travel, can slow down its spread. It merely means we need to limit person to person interactions to the best extent possible.

Wuhan, where the virus was first reported, has implemented strict measures to lock down a city of 11 million residents to slow the spread. The hope, of course, was to neutralize the spread, but all boundaries seem to be porous. The drastic measures have slowed the spread, but have not stopped it.

AI in disease control

Science has experimented with the use of Artificial Intelligence in disease control but is far from maturing it to be used in real-life scenarios. Understanding the genetic sequence of a virus can be outright cumbersome and time consuming to bio-scientists. A well trained AI model, on the other hand, can perform this task much faster.

When a virus like a Coronavirus is fast to spread, time is of the essence. Situations like these beg the need to advance the field of AI faster so that the next epidemic can be prevented by AI even before it has begun. It is not technology, but the intelligence of a machine that will help us in the future.