Machine Language to Predict the Language of Viral Mutation

As the world continues to battle with Covid-19 and its new strain, scientists around the globe are working hard to unravel the mystery behind viral mutations. The new strains of coronavirus reported in some countries, may obstruct the development of vaccines.

Scientists have considered the idea of using natural language to predict viral escape and analyze the type of mutation that result to viral escape. According to a report published on Science Magazine, the process involves the deployment of artificial intelligence (AI) technique on natural language processing.

Viral escape triggered scientists to develop an algorithm, initially meant for human natural language, in the identification of escape mutations with the ability to cause infection. Viral mutations which results to infection escape have precluded the development of a common antibody for prevention of influenza.

The model is based on the use of human language to predict how viruses evolve and escape the human immune system. Upon entering a human cell, the virus duplicates itself and the copies are transferred to other cells. Some of the copied viruses undergo mutation and produce and make the virus untouchable by the human immune defense.

The research conducted found out that viruses depicting viral escape mutations produced grammatically correct sequences but their semantics were different thus facilitating them to evade the immune system. Viruses which had the ability to evade the immune system remain a major impediment in the development of vaccines.

The idea of viral sequences was earlier raised by MIT computational biologist Brian Hie, who thought about the comparison between written language and viral sequence. He reasoned that every viral sequence had some sort of grammatical rules. As per his argument, mutations in violation of the grammar had to die.