In this article, Dr Alan D. Roth explores the role AI is playing in drug development, from its initial inception in the discovery phase all the way to clinical trials and beyond to patient use.
Investment in AI and machine learning is at an all-time high and doesn’t look to be slowing down any time soon. It’s no secret that some of the biggest technology players, Microsoft, Meta and others are investing vast amounts of money into their AI development. This is having a significant effect on industries around the world, from manufacturing to marketing, and the pharmaceutical sector is no different.
Recent figures indicate that the ‘digital health’ market is projected to rise to $260bn by 2029. Additional reports show that the global market for AI drug discovery will reach approximately $13bn by 2032, while investment in AI-based solutions in clinical research is forecast to eclipse $7bn by 2030.
What these reports show is that AI/machine learning is no longer the technology of tomorrow – it is ready to make a significant contribution to drug development today.
AI drug discovery
During the initial phase of a drug’s creation, which involves inventing and developing the novel chemical compounds which eventually become drug candidates for clinical investigation, AI is already having a significant impact.
Usually at this initial stage, novel compounds can have an unpredictable efficacy, risk profile and biological impact on a disease. But AI changes this dramatically.
The technology allows us to be far more accurate and efficient when creating molecular designs. Our AI and machine learning drug-discovery platform, Synth AI®, enables us to search for new drug candidates fulfilling three key metrics simultaneously: they possess the required biological activity against disease, can be made in the lab using known chemical processes, and are amenable to be produced at scale, making them more financially viable for investors and manufacturers.
When developing drugs, it is critical that every pound or dollar in their long journey is used as effectively and efficiently as possible, but unfortunately the industry has always had an intrinsic problem with wastage. Chemists could develop a drug all the way up to mouse trials only to find it has an adverse side effect which wasn’t apparent in the test tube, or perhaps it just doesn’t kill the disease as effectively as they had hoped. However, with AI, the enhanced accuracy means that treatments start their journey at a more advanced point along the development progression and therefore are far more likely to achieve validation also at later stage of the development process.
In fact, we have recently discovered that our lead oncology candidate, which targets colon and breast cancer among many others, already displays similar efficacy, dose response and risk profile to drug candidates that have been optimised extensively in the lab. This is highly unusual and shows in practice how AI has aided us in the development of this potential therapeutic; it highlights the impact which increased, AI-based accuracy at this early stage can have on drug development.
The knock-on effect
At Oxford Drug Design we specialise on the pre-clinical discovery and development of a drug, and we have seen how the use of AI is having a significant impact across the entire development journey.
Once a drug has been optimised and validated through mouse or other animal trials, it is ready for the clinic where it will be tested in humans. This is the critical stage of the drug’s development – where clinicians find out whether all of the pre-clinical hard work has paid off – and whether the drug is safe and effective for widespread human use. This is not a short process however and takes several years of meticulous testing to complete.
During these trials there is an enormous amount of data for scientists to collect, monitor and analyse, including vital signs such as heart rate, temperature, and metabolism, as well as metrics relating to how effective the drug is at killing a disease though efficacy and dose response parameters.
AI can clearly be used at this clinical stage in a number of ways too. Data analysis is one of the technology’s most impactful uses and in a medical setting it can track vast amounts of data simultaneously, forecasting potential risks to patient wellbeing. The technology can also provide real-time feedback on how the drug is interacting with the disease, formulating detailed reports which track results over time and generating predictions on potential long-term impacts of a drug. This accurate reporting mitigates the possibility for human error, while also improving efficiency within trials.
This illustrates how AI is already effecting a wider digital transformation journey within healthcare and pharmaceuticals. For companies like Oxford Drug Design, it is delivering important, palpable benefits, allowing us to use funding more efficiently, minimising wastage and boosting accuracy within our molecule discovery. AI and machine learning are acting as a driver for this digital revolution, enhancing the coming decades of scientific advancement and delivering better outcomes for patients, physicians and investors.
