Setting AI to work for the good of human health
In science fiction, machines are often a force for evil: invading, conquering, and killing. But at BenevolentAI in London, Dr Ngai Yi Mok (Croucher Scholarship 2005) and his colleagues are harnessing the power of machines for good, using artificial intelligence (AI) in their efforts to counter diseases through drug discovery and development.
It may sound glamorously futuristic, but Mok, a computational chemist, mostly works at his monitors, using powerful computers. “I find it really exciting. I have always had a passion [for exploring] new drugs and new treatments for diseases, which would obviously make people’s lives better and give people [a] better quality of life,” he said.
In trying to identify novel treatments, Hong Kong-born Mok is doing what he has been trained for from his days as an undergraduate and doctoral student. “I’m applying my knowledge to real-life problems. It’s very satisfying.”
Mok undertook his Hong Kong Advanced Level Examination (HKALE) at the city’s well-known Queen’s College. Both parents were scientists so he was exposed to the world of pharmaceuticals and drug treatments, drawing on that in 2001 when he undertook his bachelor’s degree in chemistry and pharmacology at the University of Leeds in the UK. His studies offered exploration of both pure chemistry and biomedical sciences, and his final-year project involved computational chemistry, in which he developed a strong interest.
He continued to a master’s degree in chemical biology, broadening his understanding of how chemistry can influence biology, and a Croucher-funded PhD in computational and medicinal chemistry.
Although his work at BenevolentAI, which he joined in 2018, is not the same as his PhD project, Mok’s career has been a continuum, leading up to this point. Now, he said, his work involves applying computational approaches to the design and optimisation of drug molecules, using what he learnt and trained for as a multidisciplinary researcher to seek solutions to problems so that human health can be improved.
Mok’s PhD project involved both computational chemistry – designing a drug molecule to treat Alzheimer’s disease – and synthetic medicinal chemistry, attempting to make what he had designed and to test it.
After completing his PhD, he furthered his research in computational chemistry applications on neglected and tropical diseases, including malaria and African sleeping sickness, at the University of Dundee’s Drug Discovery Unit. He then took on new cancer treatments at the Institute of Cancer Research in London.
“What I want to highlight from the different disease areas of the research organisations I have worked in is the transferable nature of the skill set, using computational chemistry to aid the design of new drug molecules,” Mok said.
He has published extensively on his work on applying computational techniques to finding new ways of designing drug molecules that can zero in on a biological target, usually proteins. This included publishing, with colleagues, a study on protein shapes of a potential drug target through their collaborative designs of selective inhibitors. The article appeared in the Journal of Medicinal Chemistry in 2019.
Explaining his work for the layperson, he uses the lock and key analogy. “Imagine the biological target, for example a protein, as a lock. In a lot of diseases, it is [the] malfunction or dysfunction of one or multiple proteins that leads to the underlying mechanism of the disease,” he said.
“So, designing new drug molecules is really designing a bespoke key, that will either unlock or lock up, to disable the protein responsible for the disease. Computational chemistry is a powerful approach to designing these specific keys that can control the individual locks.”
Mok has also published on efforts to understand the vast chemical space that could be accessed to build the keys that may control the locks: “We have to find the right elements and the right building blocks to design these bespoke keys.” In 2016, he published in Future Medicinal Chemistry, with Institute of Cancer Research colleagues, on the structural shape of molecules in drug design and ways of exploiting their three-dimensionality.
But some of these tasks are huge, with daunting timescales. So Mok and his BenevolentAI colleagues are using AI to speed the process. “The idea is learning from the literature what has been explored before and how we may apply that work to new challenges so, with the aid of AI, we will be able to combine and inspire novel solutions,” he said.
Designing the keys involves identifying the correct combinations of the different shapes and sizes of the building blocks from a huge number of possible permutations. “With the aid of AI, we would be able to narrow down the most probable solutions and to progress those looking most promising for evaluation, rather than all the possibilities – a very big number,” Mok said. “The technology and approach can theoretically be applied to any biological target once we have identified and verified the lock that is responsible for the disease mechanism.”
He emphasises teamwork, saying while he works at a relatively early stage of the drug discovery process, it is important to explain its benefits to the public, including patients. “What I really like about computational chemistry is you never really work in isolation.” For example, laboratory chemists synthesise and biologists evaluate the efficacy of the designed drug molecules, whereas software developers and data scientists develop the technology, and computational chemistry interacts with all these specialists.
Mok’s love of his work was “not quite a light bulb moment”, more a gradual process influenced by working in the UK, a country at the forefront of pharmaceutical research. He has seen the research field expand and opportunities grow for this new approach to developing drug treatments.
“It has given me the confidence in further pursuing it. The desire to keep at the forefront and identify solutions to the problem, which is to find new treatment for untreated diseases, is not a routine job. Every day there is something new. I’m not suggesting that it is easy. Every day has new challenges and I think that keeps us innovative.”
Ngai Yi Mok received his bachelor’s, master’s and doctoral degrees from the University of Leeds in the UK. As an undergraduate Mok studied Chemistry and Pharmacology, going on to pursue Chemical Biology at master’s level, followed by a PhD in Computational and Medicinal Chemistry. After his PhD studies, he worked in the Drug Discovery Unit at the University of Dundee and the Institute of Cancer Research before taking up his current position at BenevolentAI. He received a Croucher Scholarship in 2005.
To view Dr Mok’s Croucher profile, please click here.