Artificial Intelligence Can Help Identify The Next SARS-Like Virus

Artificial Intelligence Can Help Identify The Next SARS-Like Virus

January 11, 2022

"According to a team of researchers led by Georgetown University, artificial intelligence can accurately predict which viruses will infect humans."

An international research team led by Georgetown University scientists has proved the ability of artificial intelligence to anticipate which viruses could infect people — such as SARS-CoV-2, the virus that caused the COVID-19 pandemic — which animals would host them, and where they might arise.

"To discover the viruses, you must first profile their hosts, ecology, and evolution," said Colin Carlson, PhD, an assistant research professor at Georgetown University Medical Center. "Artificial intelligence enables us to transform data about bats into specific forecasts, such as where to seek for the next SARS."

Although statistical models are increasingly being used to prioritise which wildlife species to sample in the field, the predictions given by any model can be highly unreliable. Furthermore, scientists rarely track the accuracy of their projections, making it impossible to learn and construct more accurate models in the future. When these constraints are combined, there is considerable uncertainty about which models are best suited to the task.

This new study indicates that searching for closely related viruses may be challenging. With over 400 bat species worldwide predicted to host beta coronaviruses, a large group of viruses includes the viruses responsible for SARS-CoV (the virus that caused the 2002–2004 SARS outbreak) and SARS-CoV-2 (the virus that causes COVID-19). While the origin of SARS-CoV-2 is unknown, the spread of other viruses from bats is becoming a significant problem as agricultural development and climatic change continues.

The researchers discovered that models built on bat ecology and evolution were exceptionally good at predicting new hosts. In comparison, cutting-edge network science models that relied on high-level mathematics but lacked biological data performed about as well as or worse than expected at random.

For more information, refer to the article.

Dr Nivash Jeevanandam PhD,
Researcher | Senior Technology Journalist

Get a FREE Digital Marketing Review