Tech & Innovation in Healthcare

Reader Questions:

Using Machine Learning to Uncover the Next COVID-19 Variant

Question: As the COVID-19 public health emergency (PHE) continues, vaccines and boosters are being formulated to protect the public. However, new variants and subvariants, like BA.4 and BA.5, are being discovered.

Are researchers using technology to uncover the next variant or subvariant of COVID-19?

Arkansas Subscriber

Answer: You’re right — new variants and subvariants of COVID-19 are being discovered by scientists, and the new strains are proving to be a challenge to helping put an end to the PHE. However, researchers in Australia are using technology to help identify new and dangerous COVID-19 variants.

Australia’s national science agency, the Commonwealth Scientific and Industrial Research Organisation (CSIRO), has developed a machine learning tool, called VariantSpark, to quickly identify new variants. The researchers worked with Intel and RONIN, an ACT-based cloud system provider, to develop the artificial intelligence (AI) tool. In a study, the researchers used the technology to analyze approximately 10,000 COVID-19 samples. The researchers used VariantSpark to examine the ribonucleic acid (RNA) of an entire variant, rather than the spike protein, which allowed the team to identify variants for monitoring a week before the variants were recognized by health organizations.

The research team hopes this study could create an early warning system when determining new and emerging variants. They are also aiming to use this approach to develop vaccines that could help prevent future pandemics.

Other Articles in this issue of

Tech & Innovation in Healthcare

View All