The race for a cure

In TCS’ Innovation Lab in Hyderabad, India, a team of TCS scientists have identified 31 molecular compounds that hold promise towards finding a cure for COVID-19. This effort is part of the many worldwide mission-critical activities that TCS is engaged in, working with private enterprise and governmental groups. It represents a crucial breakthrough supporting the larger worldwide endeavor towards combating the coronavirus.

Notably, TCS leveraged its prowess in Artificial Intelligence (AI) as a key part of this discovery process.

“The use of AI has considerably shortened the initial drug design process from several months to only a few days,” said Dr Gopalakrishnan Bulusu, a principal scientist involved in the project.

Artificial Intelligence for a Real Health Crisis

The group of scientists in the Life Sciences Research unit of TCS Innovation Labs began by deciding to strategically apply their mind to de-novo drug design – a current focus area of their research. But first, they had to set up an AI model. The fundamental strength of AI is that it can rapidly evaluate multiple scenarios with a multitude of parameters while problem-solving. But any AI model must first be trained to learn the grammar of the subject language -- in this case, medicinal chemistry -- before it can start suggesting possible scenarios that could build towards a solution.

It is important to remember that the molecular universe comprises zillions of molecules, and the world of chemistry has probably looked at just about a 100 million of these. The next step was to ask the AI model about the specific case; in this instance, the SARS-Coronavirus-2 (SARS-CoV-2), the virus that has spawned the disease, COVID-19.

“We knew that the SARS-CoV-2 has a protease protein that is responsible for viral replication. What followed next was to ask the model to generate novel small molecules de novo which have protease inhibiting capability and could bind the target protease protein with high affinity,” said Dr Bulusu. “We filtered the suggestions of the AI model to a set of 1,450 molecules, and further shortlisted 31 that we determined would be good to start with and that could possibly be synthesized for further testing,” he further explained.

While these molecules capture the features of protease inhibitors, they are predicted to be much better binders than the existing protease inhibitors under clinical trial. The ¹results from this research -- put together by Dr Bulusu, Dr Arijit Roy, Dr Navneet Bung and Ms. Sowmya Krishnan -- have been published in the preprint open access chemistry archive, ‘ChemRxiv’. “The use of AI has considerably shortened the initial drug design process from several months to only a few days,” Dr Bulusu added.

Collaborating to take this discovery forward

Following the preprint research being made public, the TCS team is working closely with India’s Council for Scientific and Industrial Research (CSIR) that has agreed to provide its labs for the synthesized testing of these 31 molecular compounds. Much remains to be done before the process can move from drug design to drug discovery and finally, drug development at scale. One of the greatest challenges the international scientific community faces in relation to COVID-19 research is that not everyone has all the information. There are a lot of pieces that need to come together. Furthermore, given the risk involved in isolating and containing the virus itself, testing is that much more challenging.

TCS’ scientific ethos has been to work at discovery and to learn from the outcomes of every experiment. In that sense, no scientific research endeavor is ever a failed exercise. “We have to keep trying, keep testing, and keep learning from every experiment,” Dr Bulusu concluded.