Published 13:38 IST, April 29th 2020

Scientists use AI to crack novel coronavirus genome signature

Scientists, including one of Indian origin, have used artificial intelligence (AI) to identify an underlying genomic signature for 29 different DNA sequences of the novel coronavirus that causes COVID-19, providing an important tool for vaccine and drug developers.

Follow: Google News Icon
  • share
null | Image: self
Advertisement

Scientists, including one of Indian origin, have used artificial intelligence (AI) to identify an underlying gemic signature for 29 different DNA sequences of vel coronavirus that causes COVID-19, providing an important tool for vaccine and drug developers.

This new data discovery tool will allow researchers to quickly and easily classify a dely virus like SARS-CoV-2 in just minutes, according to researchers, including Gurjit Randhawa from Western University in Cana.

Advertisement

It provides a process of high importance for strategic planning and mobilising medical needs during a pandemic, y said.

study, published in PLOS ONE journal, also supports scientific hyposis that SARS-CoV-2 virus that causes COVID-19 disease has its origin in bats as Sarbecovirus, a subgroup of Betacoronavirus.

Advertisement

"ultra-fast, scalable, and highly accurate" classification system uses a new graphic-based, specialised software and decision-tree approach to illustrate classification and arrive at a best choice out of all possible outcomes, researchers said.

machine-learning method achieves 100 per cent accurate classification of vel coronavirus sequences and more importantly, discovers most relevant relationships among more than 5,000 viral gemes within minutes, researchers said.

Advertisement

Machine learning is an application of AI that provides systems ability to automatically learn and improve from experience without being explicitly programmed.

"All we needed was COVID-19 DNA sequence to discover its own intrinsic sequence pattern," said Kathleen Hill, a professor at University of Western Ontario in Cana.

Advertisement

"We used that signature pattern and a logical approach to match that pattern as close as possible to or viruses and achieved a fine level of classification in minutes -- t days, t hours but minutes," Hill said.

This classification tool has alrey been used to analyse more than 5,000 unique viral gemic sequences, including 29 vel coronavirus sequences available on January 27, researchers said.

Advertisement

Hill believes tool will be an essential component in toolkit for vaccine and drug developers, front-line health-care workers, researchers and scientists during this global pandemic and beyond.

13:38 IST, April 29th 2020