A New AI Tool Is Helping Researchers Come Up With COVID-19 Vaccine

Coronavirus has terrorized the world for almost a year now. While the world is grappling, South Africa recently alarmed the people with a new strain of COVID that has affected the nation. As the fight continues with vaccines that have come into the picture, a research team in the University of Southern California Viterbi School of Engineering developed a new way to counter mutations of the deadly coronavirus. This team created an artificial intelligence (AI) program that not only stops but also fastens the development of vaccines to stop the virus that has killed thousands of people and got several economies to the ground. This AI tool's machine learning system can finish vaccine design cycles within minutes as opposed to several months or years.

Paul Bogdan, associate professor of electrical and computer engineers at USC Viterbi and corresponding author of the study says, "This AI framework, applied to the specifics of the virus, can provide vaccine to candidates within seconds and move them to clinical trials quickly to achieve preventive medical therapies without compromising safety. Moreover, this can be adapted to help us stay ahead of the coronavirus as it mutates around the world."

AI Tool in Action

The team at the University of Southern California collected data from the IEDB, Immune Epitope Database, the Virus Pathogen Resource, and the National Center for Biotechnical Information. IEDB contains over 600,000 known epitomes from 3,600 different species.

When applied to SARS-COV-2, the AI model quickly eliminated 95% of the elements that could be potential COVID treatments and highlighted the best options. In totality, this AI tool predicted 26 potential vaccines that could fight this deadly virus. Out of those 26, the researchers identified 11 to develop a multi-epitope vaccine, to attack the viral spike proteins that are responsible for replication.

The current process to fight coronavirus requires growing the pathogen in labs, deactivating it, and injecting it into the people. This traditional method is time-consuming and the risk of the virus getting mutated in the meanwhile is high.

"The proposed vaccine design framework can tackle the three most frequently observed mutations and be extended to deal with other potentially unknown mutations", said Bogdan.

Some scientists think the mutations may reduce the effectiveness of the Pfizer and Moderna vaccine, which are now in circulation. To this, Bogdan said that if the global outbreak of this virus becomes uncontrollable by the vaccines in circulation, or if new vaccines are needed to fight the virus, then this AI-assisted method will provide quick results to design other preventive mechanisms.

As of now, the AI tool uses only B-cell and T-cell epitopes to come up with results. If given more datasets and more possible combinations, a stronger and quicker vaccine can be designed by the tool. It is estimated that this method can give accurate predictions with over 700,000 different proteins.

COVID-19 is responsible for a global pandemic and 1.88 million deaths worldwide. It brought the lives of many people into shambles and destroyed economies, facilitating a huge setback. Any solution, like this AI-powered tool, gives a ray of hope and helps scientists and medical health providers deal with this deadly issue. While science and R&D are playing a vital role in the production of millions of vaccine doses, it's being shown that artificial intelligence can be put to use to deploy an attack against this virus.