May 27, 2022 – Scientists at the Broad Institute of MIT and Harvard and the University of Massachusetts Medical School have developed a machine learning model that can predict which SARS-CoV-2 viral variants are likely to cause surges in COVID-19 cases.
Pardis Sabeti, an institute member at the Broad, a professor at the Center for Systems Biology and the Department of Organismic and Evolutionary Biology at Harvard University, and a professor of immunology and infectious diseases at Harvard T.H. Chan School of Public Health, was part of the research team that developed the model, called PyR0, which was described in a May 24, 2022 paper in the journal Science.
The model, which was trained using more than six million SARS-CoV-2 genomes from the GISAID (Global Initiative on Sharing Avian Influenza Data) database, can estimate how genetic mutations will impact the fitness of a particular coronavirus variant, according to a May 25 article in Genetic Engineering & Biotechnology News. When researchers tested the model using viral genomic data from January 2022, it predicted the rise of the BA.2 variant, which caused surges in many countries in March 2022. The model would also have identified the alpha variant (B.1.17) by late November 2020, a month before the World Health Organization dubbed it a variant of concern.
Information about which mutations help a variant survive can also help experts identify vaccine targets—because those mutations are likely to be the ones that remain in the virus over time.
“This kind of machine learning-based approach that looks at all the data and combines that into a single prediction is extremely valuable,” Sabeti said. “It gives you a leg up on identifying what’s emerging and could be a potential threat.”
Read the Genetic Engineering & Biotechnology News article: Which SARS-CoV-2 Variant Will Cause the Next Wave? An AI Tool Predicts
Read Broad Institute coverage: Computer model predicts dominant SARS-CoV-2 variants