A recent study has unveiled a novel artificial intelligence (AI) model capable of forecasting the emergence of Covid-variant waves. Conducted by a team from the Massachusetts Institute of Technology in the United States and The Hebrew University-Hadassah Medical School in Israel, the research utilized an AI algorithm to analyze 9 million genetic sequences of the SARS-CoV-2 virus across 30 countries.

COMMERCIAL BREAK
SCROLL TO CONTINUE READING

The model demonstrated an ability to predict approximately 73% of variants in each country that would result in at least 1,000 cases per 10 lakh people within three months, with accuracy exceeding 80% after two weeks.

The researchers incorporated data from the Global Initiative on Sharing Avian Influenza Data (GISAID), encompassing factors such as vaccination rates, infection rates, and other relevant variables.

The study, published in the journal PNAS Nexus, emphasized the importance of the early trajectory of infections, spike mutations, and the distinctiveness of mutations from the dominant variant during the observation period as key predictors of a variant's infectiousness.

Highlighting the limitations of current models in predicting variant-specific spread, the team proposed a risk assessment model based on machine learning. This approach leverages variant-specific genetic data and epidemiological information to enhance early detection and predict the future spread of newly identified variants.

The researchers suggested the potential extension of this modeling approach to other respiratory viruses, such as Influenza, Avian Flu viruses, or other Coronaviruses, with the aim of predicting the future course of various infectious diseases.

They also encouraged future research to explore how genetic and biological insights into a variant's infectiousness can be translated into predictive factors, evaluated based on available data.

(With input from PTI)