SCIENTISTS USE AI TO FIND DRUG TO FIGHT COVID
the world races to find a cure for Covid-19, a Mumbaibased data scientist and his team with the help of artificial intelligence (AI) found three molecular compounds, which they claim can be synthesised and tested further to find a novel drug to fight the Covid-19 pandemic .
Vikram Jha, 31, founder and chief executive officer of Pucho Technology Information Limited, along with his team has developed virtual drug molecules that can act as a potential drug to cure Covid-19. The team, comprising deep learning engineers Madhusudan Verma and Deepanshu Bansal, has generated three molecular structures that may help inhibit the virus’s growth.
“One of the generated drugs is functionally similar to the three medicines that are approved by the US FDA (Food and Drugs Administration) as emergency drugs — Remdesivir, Umifenovir and Elbasvir,” said Jha.
In a bid to come up with possible drugs to fight Sars-cov-2, the virus that causes Covid-19, the scientists first trained the AI model to learn the grammar of the subject language—in this case, medicinal chemistry — which led the AI to suggest possible solutions to neutralise different viruses. Following this, the super computers generated around 200 chemical structures. Of these, three were found to have drug likeness properties which can help in inhibiting the growth of Sars-cov-2.
However, considering the virus keeps mutating, it is important to find target protein that doesn’t alter much after mutation to neutralise the virus. During their research, the team found out about a critical enzyme of the virus called 3Clpro which helps for the growth of virus. If this enzyme gets inhibited, it can help to stop the growth of the virus. They used a process called docking to predict the inhibition. This is being done with the help of supercomputers to fasten the process by almost 1,000 times as the process depends on the computing power of the respective supercomputers.
“We started working on it in around mid-march and it took us almost 45 days to find the lead molecules. If one had not done this by computer-aided design, it would have taken more than three years, further delaying the process,” said Jha. Now, the team is working to convert the virtual molecules into drugs by using active pharmaceutical ingredients as feedstock. Once the drug passes the tests and clinical trials, it can be submitted for emergency user authorisation approval to drug regulatory authorities.