As viruses like COVID-19, influenza, and HIV continue to affect global populations through new mutations, researchers at the Massachusetts Institute of Technology (MIT) have reworked a preexisting algorithm to predict new forms of the viruses. An algorithm originally designed to analyze and predict patterns of language may now be essential in overcoming new viral strands.
The model has been trained to analyze genetic sequences and predict new ones that follow the biological rules of the protein structure. A virus cannot replicate without a correct sequence; therefore, it relies on shifting protein structure to avoid antibodies. The model can detect these alterations in the genetic makeup of the virus that allow for a viral escape or allow it to go undetected by the immune system.
The only information needed to study these mutations is a relatively small amount of sequence data. The model was trained using 60,000 HIV sequences, 45,000 influenza sequences, and 4,000 coronavirus sequences.
The sequences of these three viruses are among the fastest to mutate, making it impossible to produce a single vaccine with lasting effects in the long term. The model determined that the spike protein in SARS-CoV-2 called the S2 subunit is the least likely to result in viral escapes, however the time it takes for the novel virus to mutate is still unknown.
Professor of Mathematics and Head of the Computation and Biology Group in MIT's Computer Science and Artificial Intelligence Laboratory, Bonnie Berger, and her colleagues published a study in Science that states they identified possible targets for HIV, influenza, and SARS-CoV-2. The group has also flagged viral genetic sequences in the new strand appearing in the United Kingdom and South Africa, however this analysis has not been confirmed.
Continued research and the help of this new strategy could lead to a vaccine capable of detecting new mutations in viruses like COVID-19. MIT’s researchers are working to create a vaccine that informs and prepares antibodies for future mutations, preventing future pandemics.
MIT is also collaborating with others to develop a treatment capable of destroying cancerous tumors using their algorithm, which may also be applied to the creation small-molecule drugs to treat diseases such as tuberculosis. As the study of sequence data continues, researchers work to strengthen the immune system, prepare for future illnesses, and create effective vaccines.