Global EditionASIA 中文双语Français
Home / China / HK Macao

Macao-led research develops AI model to predict pathogenic variants of COVID-19

Xinhua | Updated: 2023-08-02 11:29
Share - WeChat

MACAO -- An international team led by researchers at the Macao University of Science and Technology in south China has developed an artificial intelligence (AI) model that can predict the pathogenic variants of COVID-19.

Named UniBind, the model can predict which variants of COVID-19 can increase the infectibility of the virus or help it develop resistance to antibodies or vaccines, through analyzing the over 6 million pieces of viral sequence data generated from global monitoring, according to the team.

The study was published in the latest edition of Nature Medicine, a monthly journal.

Zhang Kang, professor of medicine at the university who had led the research, said the model can integrate and analyze data from different experimental sources and modalities, unlike most existing AI methods that can only make predictions by analyzing a certain kind of experimental data.

The team said it had used UniBind to simulate over 30,000 virtual variants and correctly predicted the evolutions of current main strains such as XBB and BQ mutations of Omicron.

The model further predicted that top ranked mutations such as A475N and S494K are likely to possess high immune escape properties and may drive future viral evolutions.

Results also showed the model can accurately predict the affinity of different viruses and their mutations to different species, which is significant to discovering the intermediate hosts of epidemics and predicting viruses' trans-species transmission paths.

Copyright 1995 - . All rights reserved. The content (including but not limited to text, photo, multimedia information, etc) published in this site belongs to China Daily Information Co (CDIC). Without written authorization from CDIC, such content shall not be republished or used in any form. Note: Browsers with 1024*768 or higher resolution are suggested for this site.
License for publishing multimedia online 0108263

Registration Number: 130349