Efficiency welcomed in a competitive market
Biotech buzzwords do not come much bigger than those of artificial intelligence (AI) and machine learning (ML). With the promise of expediting routes to market and reducing costs, such platforms may very well find themselves cemented as critical components in future drug development toolboxes. The application of ML approaches in the pharma industry has now matured to a stage where the first purely ML generated candidates have entered clinical trials. However, we are still a way off from using solely ML to uncover completely new disease mechanisms of action or targets, which many would consider as the holy grail of applications. Leading the charge towards this ambition are focused biotechs leveraging ML that are not only applying their platforms to bolster internal pipelines but striking deals with big pharma, which are likely to be the main clients, utilising ML to boost their own portfolios.
The pharma industry continues to evolve and, based on current macro pressures, we are likely to see different approaches to increase efficiency and expedite drug discovery. In our view, these efficiencies will be partly driven by an increased focus on ML technology, finding applications from the discovery through to clinical trial stages of development. Albeit still in their relative infancy, ML platforms and technology applications aim to improve complex drug development by scouring vast pharmaceutical datasets to identify trends or patterns that facilitate decision-making. Due to current funding challenges, methods that claim to de-risk drug candidate selection and reduce time to market are likely to gain heightened interest. It is not just technologies that are modernising either. In efforts to minimise animal testing, regulators are beginning to support the use of new methods to validate drugs before they enter the clinic, marked by the recent passing of the FDA Modernization Act 2.0. Change may not come quickly; however, the passing of the act will certainly help promote the use of alternative approaches that, in time, may predict drug safety and efficacy with even greater accuracy and cost efficiencies.
On the face of it, the concept of harnessing big data to improve drug development efficiency seems like a no brainer. However, to truly disrupt the industry, awareness will need to be raised to drive change away from historically manual processes and more traditional mindsets that believe nothing can truly trump the knowledge of a medicinal chemist with 30 years’ experience under their belt. Additionally, drug development failures are estimated to cost as much as c $700m, with failure rates of c 90%; if ML technologies produce results, it would certainly promote their usage. While the approval of drugs stemming from ML might seem in the distant future, we believe credibility and success in the more immediate term will be based on those platforms that bring drugs into clinical trials or that unveil new disease targets. Not only would this help support the wider adoption of such nascent technologies, but it would provide both big pharma and investors alike with opportunities to get in on the ground floor, before ML starts realising its potential of accelerating development and delivering clinical successes.