The Royal Society of Chemistry published in Chemistry World recently an article about an AI-developed drug which looks set to enter stage three clinical trials. On the back of this news, LightOx’s CEO, Dr Sam Whitehouse, shares how he thinks AI will reshape drug development, from the lab to the clinic, and how LightOx is making use of emerging technologies.
AI as a strategic enabler
AI is rapidly becoming a strategic asset in drug development and a commonplace starting point for a lot of target identification and development for leads.
Today, it enables faster, data-driven decisions across discovery, design and development. Over the next decade though, I think AI will evolve from a supporting tool to an essential innovation tool – reducing R&D timelines, improving success rates and unlocking new therapeutic possibilities.
For investors, this means lower risks and faster ROI, but I hope it will also mean more options for patients.
AI can also make early intervention treatments more scalable and cost-effective. Early identification and patient stratification is so important. Being able to know who is at risk in a patient group with early-stage lesions is a challenge no one has got to grips with and it’s an area where I’d love to see AI making a positive difference.
For LightOx, this aligns with our view to deliver targeted oral cancer therapies that are not only effective but also accessible across diverse healthcare systems.
Beyond molecular modelling
We’re integrating AI into multiple stages of our pipeline, but this is not exactly new to us.
Medicinal chemists have been using molecular modelling to anticipate and optimise compound development since I was in the labs! AI advances this and helps us analyse complex biological datasets, refine our chemical libraries and prioritise high-potential candidates.
This enhances our ability to innovate quickly and cost-effectively, which is critical for a lean, agile biotech like LightOx.
With predictive modelling, we can simulate interactions and refine the molecule before lab testing – potentially reducing development time by several years. This significantly accelerated our development of LXD191, our light-activated oral cancer treatment, by looking at binding points with various proteins in the target cell types and optimising compound design.
Beyond molecular modelling, I believe the most transformative impact of AI tools lies in:
- Molecule design: Faster, smarter compound generation.
- Safety prediction: Early risk mitigation and off target analysis.
These areas directly influence cost, speed and success rates – key metrics for any investor evaluating biotech innovation.
Responsibility in AI-driven drug development
There are valid concerns about the use of AI, and we take them seriously.
Its use has to be based on high-quality, curated datasets and we must maintain transparency in why and how decisions are made when using these tools.
Regulatory-wise we still have to prove all our compounds work. Safety and efficacy don’t change, but our ways of working are evolving, and we see this as an opportunity to lead with integrity and innovation.
Where AI is used in diagnostics, I anticipate regulators will demand greater model explainability and data traceability and staying ahead of regulatory trends will be essential. Those in the diagnostics field will need to adopt a proactive approach to navigate approvals efficiently and maintain investor confidence.
In pharmaceutical development however, I think the real power lies in the front end of development and reducing time for new targets, but I am sure someone will tell me AI will be synthesising new compounds in no time!
LightOx’s lead candidate is LXD191 – a light-activated treatment designed to be applied directly to pre-cancerous oral lesions. Find out more about its development.
