MIT develops machine learning model to quicken release of COVID-19 vaccine


OptiVax tests potential vaccines for efficacy and population coverage, then designs vaccines for testing, all with artificial intelligence.

Image: Shutterstock

Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a new combinatorial machine learning system that could both decrease research time needed for a COVID-19 vaccine and make it more effective, researchers said. 

The platform, called OptiVax, focuses on developing peptide vaccines, which are a different approach from common whole virus, DNA, and RNA vaccines currently among the more than 100 vaccines in development.

Peptide vaccines are a relatively recent development in the vaccination game that are designed around one specific short amino acid string, called a peptide, that can be found in the target disease. Peptide vaccines use a synthetic version of the peptide that is created in a laboratory and not harvested from the disease itself. 

Traditional vaccines have a larger amount of genetic information in them that isn’t useful in developing resistance and can lead to unwanted immune responses and dangerous reactions—it’s these genetic elements that peptide vaccines are designed to eliminate, MIT said. 

The peptides included in a peptide vaccine are, ideally, the most effective at building an immune response without unnecessary material, and are effective across a wider range of individuals.

SEE: The new normal: What work will look like post-pandemic (TechRepublic Premium)

In the case of COVID-19, MIT researchers aren’t trying to develop a pure peptide vaccine, but are trying to create a complementary peptide that can improve traditional vaccines targeting the spike proteins that cover the virus.

“”Based on our analysis [of a common COVID-19 vaccine], we developed an augmentation to improve its population coverage by adding peptides. If this works in animal models, the design could move to human clinical trials,” said CSAIL researchers Ge Liu and Brandon Carter. 

OptiVax works by first testing a wide variety of peptide fragments to figure out which would be the best candidates for a vaccine. The data gathered by those machine learning-powered tests are scored on a wide range of criteria, including how the peptides react to various human genetic samples that are sorted by geographic location.

SEE: Return to work: What the new normal will look like post-pandemic (free PDF) (TechRepublic)

A completed OptiVax test would end with a vaccine designed by the machine learning model that would “maximize population coverage in different geographical regions, and from the number of peptides displayed per individual to improve the chances the person will become immune,” the researchers said.

The MIT researchers behind OptiVax said that the next step after finding a proper peptide vaccine is animal testing, after which they would move on to human tests if a clinical trial is warranted. 

OptiVax isn’t limited to usefulness in fighting the COVID-19 pandemic, either: The CSAIL team said that it could be used as a model for developing and testing vaccines for a variety of diseases using machine learning.

Also see



Source link