The coronavirus has spawned numerous variants that are more contagious than their predecessors. This fight against the pandemic might seem never-ending, but scientist will not falter in the quest to save more lives. With the help of biotech advancement and research excellence, people can outsmart viruses.

A research team from The Chinese University of Hong Kong (CUHK)’s Faculty of Medicine (CU Medicine) has developed a computational framework that can estimate COVID-19 vaccine effectiveness (VE) against SARS-CoV-2 variants by considering genetic distance (GD). The model was verified with independent observed data and showed extremely promising results, with prediction accuracy as high as 95%. Timely information about VE is crucial for successful deployment of medical resources and improved public health responses.

VE-GD computational framework provides a rapid method to predict protective effect of vaccine against novel genetic variants with 95% accuracy

In a research study published in Nature Medicine, CUHK researchers analysed nearly 2 million SARS-CoV-2 sequences and 49 clinical trials and observational studies, and developed new algorithms to rapidly evaluate the VE of different types of vaccines against symptomatic COVID-19 infection using genetic distance. The team was led by Professor Maggie Wang Hai-tian and Professor Benny Zee Chung-ying, both from CU Medicine’s Jockey Club School of Public Health and Primary Care (JCSPHPC). They found that the GD between the receptor-binding domain of the spike protein of the circulating viruses and the vaccine strain is highly predictive of vaccine protection. The larger the genetic mismatch, the less effective the vaccine would be. Their method demonstrated 95% VE prediction accuracy through genome analysis, validated on multiple independent datasets.

The predicted VE based on GD largely mirrors the observed VE, with accuracy as high as 95%.

Rapid algorithms are months ahead of traditional studies, and can facilitate selection of the optimal antigen

The traditional approach to observing VE is very time-consuming. The results are only available after people have been vaccinated and a portion of the population has contracted the disease. For mutated variants, scientists must go back to square one and redesign the study all over again. The delay can hinder clinicians, healthcare providers and policymakers from pivoting swiftly.

Professor Maggie Wang, Associate Professor from CU Medicine’s JCSPHPC, said that the team were among the earliest groups to have anticipated the threat of the Omicron variant, which was first reported in Africa in November 2021. They were very shocked when they estimated a very low VE using the model. The first VE report for Omicron using traditional studies arrived three months later and proved their prediction right: the variant indeed evaded the immunity generated by current vaccines.

This timesaving bioinformatics platform can estimate VE much more quickly than conventional methods. This could give us the upper hand to improve health measures and design vaccines with the optimal antigen.

Professor Benny Zee, Director of the Centre for Clinical Research and Biostatistics at CU Medicine’s JCSPHPC, said, “Vaccine manufacturers can use this technology to select candidate antigens and inform clinical trial design. Healthcare workers and policymakers can estimate the scale of an upcoming epidemic caused by new variants with information about the predicted VE.”

(From left) Ms Cao Li-rong, PhD student; Professor Benny Zee, Director of the Centre for Clinical Research and Biostatistics; and Professor Maggie Wang, Associate Professor, The Jockey Club School of Public Health and Primary Care, CU Medicine.

The full text of the research paper can be found at:
Rapid evaluation of COVID-19 vaccine effectiveness against symptomatic infection with SARS-CoV-2 variants by analysis of genetic distance