Researchers Build New AI Model To Analyze And Predict Viral Escape

Virus!! Vaccine!! Mutation!! These are probably three most heard words in recent times. But how are they related? Though vaccines are a boon to humankind and can develop immunity, viruses can mutate fast enough to leave the vaccines impotent. This is called ‘viral escape‘. This is a big problem.

Experts across the world are afraid as this phenomenon can happen with coronavirus as well.

But researchers from MIT have a good news for us. A team of researchers from MIT have developed a new way to computationally model viral escape.

The study recently appeared in the Science journal.

According to Bonnie Berger (the Simons Professor of Mathematics and head of the Computation and Biology group in MIT’s Computer Science and Artificial Intelligence Laboratory), viral escape is a major hurdle in building a vaccine for influenza and HIV.

How does it work?

The model works by predicting which sections of viral surface proteins are more likely to mutate in a way that leads to viral escape. Moreover, the model can even identify sections that are less likely to mutate which gives us a perfect target for vaccines.

Interestingly this model is based on machine learning algorithms that were originally developed to analyze language.

The model was trained using 60,000 HIV sequences, 45,000 influenza sequences, and 4,000 coronavirus sequences.

Using the model, researchers identified possible targets for vaccines against influenza, HIV, and previous strains of SARS-CoV-2.

As the current paper was accepted for publication, the researchers have applied their model to the new strains of SARS-CoV-2. Though the new analysis has not yet been peer-reviewed, it has flagged viral genetic sequences that should be further investigated for their potential to escape the existing vaccines.

The Results

Researchers used the trained model to predict sequences of the coronavirus spike protein, HIV envelope protein, and influenza hemagglutinin (HA) protein as these were likely to generate escape mutations.

For influenza, the model revealed that the sequences least likely to mutate and produce viral escape were in the stalk of the HA protein. This is supported by recent studies as well.

The model’s analysis of coronaviruses suggested that a part of the spike protein called the S2 subunit is least likely to generate escape mutations.

The researchers found that the V1-V2 hypervariable region of the HIV envelope protein has many possible escape mutations, which is consistent with previous findings, and they also found sequences that would have a lower probability of escape.

Developing it further, the researchers are now working to identify possible targets for cancer vaccines that stimulate the body’s own immune system to destroy tumours.

Journal Reference:
Brian Hie, Ellen D. Zhong, Bonnie Berger, Bryan Bryson Learning the language of viral evolution and escape Science  15 Jan 2021 DOI: 10.1126/science.abd7331

Press Release: MIT Computer Science & Artificial Intelligence Lab