Taken from an interview with Hannah Blake of Pharmaphorum
While the pharma industry may believe drug discovery has been fully optimized, David Leahy disagrees. As the CEO and co-founder of Molplex, David shares his vision with us – to drive down productivity costs from £20 million to £1 million for each new drug candidate. He believes that in the long-term, changing productivity methods and lowering costs will allow further, and better, development of small, niche, personalised medicines.
HB: Hi David, it’s great to speak with you. Can you please tell us a bit about your background as a drug discovery scientist?
I started in R&D at ICI Pharmaceuticals with a PhD in Physical Organic Chemistry trying to unravel the factors that reduced oral absorption. I worked with a very creative team of scientists developing what were then new computer-based drug design methods. While being actively involved in the technology development, I also worked in multiple drug discovery teams, most of which failed because we weren’t able translate potency into invivo activity while also avoiding toxicity. The company ran a lot of high level initiatives to try and solve these problems and I got brought in to work on many of them, but I became disillusioned when after several years none of the change initiatives had changed anything. So I quit and founded Cyprotex to pursue my ideas for automating ADMET screening so that we could front load projects with data to guide drug design, rather than work for several years only to fail at the final hurdle. After a difficult start, Cyprotex has grown to become the world’s largest specialist ADMET CRO.
Since leaving Cyprotex, I have been developing technologies aimed at exploiting the explosion of chemical and biological data generated by advances in laboratory automation to accelerate drug discovery, working independently and also for a time at a major UK University.
HB: What inspired you to found Molplex, and what does your current role, as CEO, entail?
The industry, investors and people in our business seem to have resigned themselves to believing that drug discovery has been fully optimised, that at £20M per new development candidate it remains very high risk and fundamentally uneconomic, but it is the best we can do. This cost-risk profile makes realising the goals of personalised medicines much harder of course. Management seem to have accepted defeat and are now shipping the work overseas where the costs are lower (at least for now) and keeping their fingers crossed that the Universities will somehow deliver the new development candidates they need to feed their pipelines. We think that the way we invent drugs might be at an optimum, but it’s a local optimum. Our view is that drug discovery by medicinal chemistry is an evolutionary dead end and we should reboot. We have spent nearly 6 years working out how to do that.
As CEO, my job is to work with my co-founder to develop our ideas into a compelling vision of a new, radically lower cost drug discovery business that will attract the support of investors as well as partners in business and Universities.
HB: What is the main aim of Molplex?
To reduce the average cost of drug discovery from £20M to less than £1M, and build a pipeline of quality drug candidates.
“Our view is that drug discovery by medicinal chemistry is an evolutionary dead end and we should reboot.”
HB: How do other approaches to drug discovery differ to yours?
Imagine a business was started to help people find information in online documents and to do this they trained people to be good at speed reading. When the customer wanted to find something they hired an expert speed reader, described what they were looking for and the speed reader then read through the documents and produced a report. Of course, they would have good software tools to prioritise documents, store data and produce the report nicely so the speed readers were more efficient. Individual speed readers also develop expertise and tacit knowledge over time that helps them know where to look, but eventually the only way you can do this more cost-effectively is through cheaper people. This is pretty much what we do by relying on medicinal chemists to drive drug discovery projects.
Molplex is as different from medicinal chemistry based drug discovery as Google is from a team of speed readers. Molplex technology automates the complex decision making done by drug discovery scientists allowing “always on, always updated” combinatorial search and optimisation through massive databases for all drug properties all in parallel. Our systems accumulate the tacit knowledge and expertise of drug discovery scientists and learn and improve over time. They are massively scalable and by extreme front loading of decisions on which compound, which series and which project to prioritise, we think we can deliver our goal of less than £1M per development candidate.
HB: Why is important to keep up-to-date with the latest technology for drug discovery?
Managers in industry have to spend more than 95% of their time on operational issues trying to optimise people, process and project performance. It’s easy to focus on soft skills training, process optimisation and shuffling organisations and systems to drive improvement in productivity. However, finding ways to get people to walk round the labs faster and cost-cutting campaigns will never deliver more than a few percentage points of improvements in productivity, as judged by the number of high quality compounds going into development. That may be useful, but is rarely transformative.
The killer problem in discovery isn’t really the cost of doing business, it’s the failure rate. Since 95% of discovery costs are wasted by failure to deliver high quality candidates, we should focus on what would give us better quality compounds, even if the cost per project increases. A 5% reduction in failure rate to 90% would double productivity; while a 5% cost reduction is just that. It is new technology and science that will deliver reduction in failure rates, not the process optimisation methodologies borrowed from other industries that the industry invests so heavily in. We are a long way from exhausting the power of emerging technologies to help us to find new drugs and to reduce downstream failure in the clinic. Keeping on top of technology is essential for spotting the opportunities that technology presents to do things differently.
HB: Do you think personalised medicine is just the next big trend, or do you think it will firmly cement its place in pharma?
We think personalised or stratified medicine is a very important development and will become normal practice. In fact, there are already good examples emerging. However, the economics of inventing drugs with a global reach is already stacked against us, so inventing a drug that targets a specific sub-population is even more difficult to justify commercially unless we can radically lower the costs and increase success rates. There are some advantages, of course, in being able to better stratify trials to target the right population and this may lead to fewer failures in the clinic, but our ability to discover the drug in the first place needs a step change in productivity if we are to deliver products into smaller, niche markets.
- Molplex raises £350,000 from the North West Fund for Biomedical (molplex.com)
- Driving down the cost of drug discovery (pharmaphorum.com)
Tags2.0 technologies AstraZeneca PLC autonomous agents BioMoby Biotechnology Biotechnology and Pharmaceuticals business Cancer Research UK ChEMBL chemical and biological data Cheminformatics Cloud Computing combinatorial search Cyprotex David Leahy drug design Drug development Drug Discovery Expert information technology Medicinal chemistry Microsoft Research Molplex on demand services Open Science Pharma pharmaceutical Pharmaceutical companies Pharmaceutical industry Pharmaceutical sciences Research Research and development semantic web Web 2.0 Web service