RIYAZ SYED compares the relationship between disease-causing proteins in the human body and the drugs that treat them with a lock-and-key analogy. “The protein is the lock, and the drug is the key. As a drug discovery scientist, my job is designing the key that unlocks the target. This involves validating the target to ensure it is the correct lock, followed by designing molecules,” he explained while discussing his AI-powered innovations in drug discovery at the awards event in Hyderabad.
As humanity evolves, so does the spectrum of diseases affecting it. According to the 11th Revision of the International Classification of Diseases (ICD-11) by the World Health Organization, there are more than 17,000 unique codes for various diseases, injuries and causes of death. However, a significant gap exists in the availability of corresponding drugs for most of these conditions. This gap is largely because of the conventional drug discovery process, which typically spans 10 to 15 years and demands substantial financial investment from initial research to an approved treatment.
The past decade, however, has witnessed a transformative impact of AI on drug discovery, significantly reducing the time and cost of bringing new drugs to the market. Syed is part of a new wave of Indian drug researchers leveraging AI technologies―particularly machine learning models―to analyse vast datasets, identify potential drug candidates and predict their effectiveness and safety with remarkable speed. He notes that these technologies streamline critical stages of drug development, including target identification, virtual screening and toxicity prediction.
A seasoned drug hunter with 15 years of experience in medicinal chemistry, Syed has authored 50 publications and holds multiple patents. His firm, Centella AI, is trying to revolutionise the drug discovery process by reducing the time required for the design-make-test-analyse (DMTA) cycle by 25 per cent and cutting costs by 60 per cent.
He highlighted the scarcity of health data for drug research in India. “AI models need extensive datasets to identify patterns, but accessing such data was extremely challenging initially,” he said. “Many diagnostic companies I’ve interacted with have data and an interest in AI applications but lack clarity on what to do, how to do it, and most important, why to do it. Bridging this gap required significant effort to train and groom teams with dual-domain expertise.”