Monday, April 23, 2018

Company focus: Exscientia – PharmaTimes Magazine May 2018

With several high-profile collaborations under its belt, Exscientia is bringing AI tech straight to the heart of pharma R&D. We asked the company’s founder, Professor Andrew Hopkins, to explain how they’re hoping to make long, expensive drug development a thing of the past


What is your background?

I am the founder and CEO of Exscientia Ltd – the first company to automate drug design from idea to clinical candidate, using AI. Following a chemistry degree (Manchester) and a DPhil from Oxford, I spent ten years at Pfizer, where I was responsible for establishing new research foci including the concepts of druggability and network pharmacology. Subsequently, I went back into academia to develop these concepts and was one of the youngest professors when I was appointed as Chair of Medicinal Informatics at the University of Dundee. I have raised around £50 million for commercial and academic research activities in my role at the University. I founded Exscientia in 2012 as a spin-out from the University of Dundee. I am the author  of some of the most highly cited papers in modern  drug discovery.

What is the history of the company? What is the main drive behind the business?

The company is focused on dramatically improving productivity in the stage from initial chemical design to clinical candidate. This is the single most expensive part of drug development per launched drug; due to a perfect storm of high costs for each project (typically around $20 million), and even higher project attrition rates (the number of projects needed in reality to yield one marketed drug).

Exscientia’s goal is, starting from a project definition, to design and advance novel high-quality compounds (with excellent therapeutic properties) to clinical entry in one and a half years. It does this by applying AI to dramatically reduce the number of compounds required for analysis from typically 2,500 per project using traditional approaches, to just 500 when using AI-driven techniques.

Our approach is to fuse the power of AI with the discovery experience of seasoned drug hunters (key among which is Andy Bell, co-inventor of Viagra and other key molecules during his career). As a result, the company believes it is the first to automate drug design in a manner surpassing conventional approaches.

Since its establishment in 2012, Exscientia has achieved a series of collaborations, including more recently with Sanofi and GSK, which have helped to fuel expansion. Recently, Exscientia also opened a new  office in Oxford.

Tell us about your technology. How does it work? What is the history of the idea?

I recognised that there was a breadth of data in journals and patents which exceeded that available to any individual group or company. I also recognised that if organised appropriately, it should be possible to make connections in a way that was more systematic and strategic.

Many members of Exscientia worked on some of the first IT systems to curate data from multiple sources (such as SAR data from patents and literature) and applied it productively to drug discovery. The next crucial step was to develop an automated drug design system that could combine these data with a mass of other inputs, such as internal HTS screens, fragment screens, protein structure data, and other information – in order to efficiently design novel, patentable chemical entities for synthesis and assay. Together with a young and talented academic group, a new approach to drug design was developed and published in Nature in 2012. Over time, we have enhanced and refined this AI-driven approach, testing it in partnership with pharmaceutical partners on real projects, which also helped fund our advancement without the need for external investment.

In these early partnerships, Exscientia has successfully and reproducibly completed projects to drug candidate stage within its target of 500 compounds and 12 months. For one partner, a candidate molecule was delivered within 12 months of project initiation. A key aspect of this speed is a tightly integrated Design-Make-Test cycle; here  the strategic selection of which compounds to make and test based on all data, coupled with rapid experimental turnaround (ideally two weeks), yields a rapid cycle of learning for each project undertaken.

The attractions of Exscientia’s approach and capabilities have resulted in several recent drug discovery collaborations with GSK, Sanofi, and Evotec, as well as a €15 million direct investment from Evotec, which becomes the company’s first external investor.

Can you give some specific examples of how you have collaborated with pharma companies?

In detail these collaborations are:

  • GSK (announced July 2017) – to discover novel and selective small molecules for up to ten disease-related targets nominated by GSK across multiple therapeutic areas (near-term milestone potential of £33 million)
  • Sanofi (April 2017) – up to €250 million strategic research collaboration and licence option agreement for bispecific small molecules targeting metabolic diseases (including diabetes and obesity)
  • Evotec (April 2016) – 50:50 co-development agreement to discover and develop small molecules focused on immuno-oncology. In September 2017, Evotec made a €15 million strategic investment in Exscientia becoming our first external investor.

How would you like to further develop the technology/business in the future?

A goal for the business is to routinely achieve the transformational productivity improvements we believe our AI-based approach can deliver, to build our own pipeline of novel drug compounds and to support our partners’ discovery efforts.

Being a participant in Elsevier’s Hive initiative will enable Exscientia to access Elsevier’s suite of information solutions and boost our existing capabilities. For instance, the Hive provides us with additional access to structured chemistry and activity data, ensuring that we have a comprehensive overview of what has been published already to support our ongoing projects. Similarly, the pathway and literature tools support our target identification and validation activities, and we look forward to partnering with Elsevier’s newest exciting initiatives around large-scale knowledge graphs.

AI is seeing huge amounts of interest from pharma at the moment. Do you think the potential matches the hype?

AI, if used correctly, has the potential to transform the productivity of drug discovery and deliver superior candidates into the clinic. This will change the way discovery is done. We are extremely encouraged by the results we have seen with our own programmes and the progress of molecules through the development process will provide validation of the benefits of an AI-based approach. The challenge is how to develop applications that apply AI productively, rather than questions about AI itself.

What can the industry do to ensure that the use of AI technologies such as yours is as successful as it can be?

To be successful, big pharma must invest in big pharma. Large companies need to look critically at their incumbent processes and the associated metrics and ask whether those are the right processes to take them towards the future of drug discovery. We believe that our collaborators are already thinking along these lines and this will help advance  the technology.

What are the biggest challenges companies working in AI are facing at the moment? How are you aiming to tackle them?

Delivery is everything. The technologies are becoming more sophisticated as humans learn how best to apply the tech, and how best to work with the tech. However, it is not AI alone that can solve all these problems. We believe that the combination of a human with AI as a ‘Centaur’ team, will be better than either human process or AI process alone. This has already been shown in other areas, such as chess, where a human with a computer algorithm can typically beat both a human or a computer, when alone. Drug discovery is far more complex than a game of chess – for this reason we think of our chemists as ‘Centaur chemists’, allowing the AI to perform the design work whilst the chemist sets the strategy.

There will need to be an overhaul in how big pharma rethinks its strategy when it comes to the use of AI. This is why we believe we have such a strong advantage.

What are some of the greatest opportunities for AI pharma companies in the UK at the moment?

Drug discovery is the last artisan industry. One of the key opportunities is also its challenge; the amount of data now available is beyond any human’s capability to hold in its mind at one time. The use of AI for drug design, where Exscientia concentrates its expertise, will have unprecedented opportunity.

Other companies are looking at patient stratification and personalised medicine. For example, in patient stratification for clinical trials, AI may be used to fit the patient to the.

Source link

No comments:

Pharmaceutical Validation Documentation Requirements

Pharmaceutical validation is a critical process that ensures that pharmaceutical products meet the desired quality standards and are safe fo...