We've updated our Privacy Policy to make it clearer how we use your personal data.

We use cookies to provide you with a better experience. You can read our Cookie Policy here.

Advertisement
Harnessing the Power of Robotics To Transform Drug Discovery
Industry Insight

Harnessing the Power of Robotics To Transform Drug Discovery

Harnessing the Power of Robotics To Transform Drug Discovery
Industry Insight

Harnessing the Power of Robotics To Transform Drug Discovery

Credit: Arctoris

Want a FREE PDF version of This Industry Insight?

Complete the form below and we will email you a PDF version of "Harnessing the Power of Robotics To Transform Drug Discovery"

First Name*
Last Name*
Email Address*
Country*
Company Type*
Job Function*
Would you like to receive further email communication from Technology Networks?

Technology Networks Ltd. needs the contact information you provide to us to contact you about our products and services. You may unsubscribe from these communications at any time. For information on how to unsubscribe, as well as our privacy practices and commitment to protecting your privacy, check out our Privacy Policy

Arctoris Ltd recently unveiled its new state-of-the art facility – home to the company’s next-generation, fully automated drug discovery platform. I was lucky enough to see the platform first-hand, and had the pleasure of speaking with Martin-Immanuel Bittner MD DPhil and Tom Fleming MChem, two of the cofounders of Arctoris.

Martin and Tom discuss the evolution of drug discovery, the challenges faced by those working in the field, and share how AI, machine learning and robotics are impacting the discovery and clinical success of drugs. 

Laura Lansdowne (LL): What led you to found Arctoris? Could you elaborate on the company’s mission and goals?

Martin-Immanuel Bittner (MIB):
The ideas on how to transform how drug discovery research is performed emanated during my PhD in Oxford. Having worked as a clinician in Germany, I was rather surprised by what life in the lab looked like. Rather than the focus on the intellectual challenges of new hypotheses or identifying new cures, one of the main challenges facing PhD students, postdocs and even professors was a practical one: the need to spend considerable amounts of their time, seven or eight hours every day, manually pipetting small amounts of liquids in a rather repetitive, time-consuming way. The way in which life sciences experiments are performed has not fundamentally changed for many decades, and addressing this could release researchers’ time to focus on other aspects of scientific discovery – whether that was reading literature or understanding more about molecular pathways.

When reading more about drug discovery and its challenges, it became clear that shortcomings in the data quality produced by manual laboratory techniques is an impediment to scientific progress. Independent third parties, such as the in-house teams of pharmaceutical companies, trying to reproduce findings in reputable journals only succeeded in 10–20% of cases.

With massive failure rates, low chances of success in drug discovery, and escalating costs, biotechnology companies, pharmaceutical corporations and academic centers cannot afford to continue to rely on an inconsistent, inefficient manual approach to experimentation.

So, during our doctoral studies, together with my two cofounders we thought this traditional way of doing lab research was ripe for disruption, and that there was potential to harness the benefits of automation and robotics, and use them to bring greater reproducibility, standardization, throughput and speed to the life sciences community.

Tom Fleming (TF):
Well, with drug discovery, imagine how much more efficient it could become, if you could rely on 90% of results, rather than being skeptical about 90%. There is a thirst for quality data that is not being satisfied – and both small companies and big pharma are experiencing this. The data that has been collected historically is uncertain in terms of value and reliability, and that is what our partners and clients are realizing – they need to go out and find new and better ways of producing data.

Regardless of revenue or size of a biotech or pharma company, drug discovery teams are finding that their databases are often unstructured and unreliable.

With the explosion of machine learning and artificial intelligence (AI), and the need for data to feed these approaches, the importance of high-quality data is becoming even greater.

LL: Could you highlight some of the latest drug discovery trends?

MIB:
From a clinical perspective, what we are seeing right now is a huge trend towards personalized medicine and targeted therapies. The drug discovery community is trying to develop new medicines that are not based on the “one size fits all” approach – one drug for a million patients. In fact, it is the complete opposite – ideally 1 million different drugs, one for each patient. This involves going far deeper in terms of characterizing diseases more thoroughly, to find drugs that are really going to benefit that one patient or a particular, very well-defined patient group.

The current approach to drug discovery – and the current costs of £2–3 billion required to develop just one new drug – are completely unsustainable.

TF:
The recent explosion of AI and machine learning in drug discovery has been accompanied by an explosion in the number of start-ups being founded with that as their remit.

This very young, new sector is realizing the value of high-quality experimental and real-world data, to validate and develop their computational models.

They often use existing open access data repositories to mine data and train their models – but that has a limit. To take the next step, they have started turning to Arctoris because our fully automated platform can “plug in” to their ecosystem to generate new, high-quality data for them. Ultimately, we are seeing that we can help AI and machine learning-driven companies transition from purely offering software as a service to becoming independent, highly innovative biotech companies in their own right.

LL: How have advances in automation and AI impacted the setup of the modern drug discovery lab?

TF:
So, automation is very trendy, but in fact, it's been around for decades. Typically, it was only applied to the first step in drug discovery, where a big pharmaceutical company would conduct a high-throughput screen – testing potentially millions of drug molecules against a single target. They would generate a vast amount of data that ended up being just the tip of the iceberg – missing key data insights, which would not fully support decision-making.

Beyond that, follow up of those “hits” would be performed using manual setups. They would then be subject to manual laboratory techniques – so that's as far as the automation would go. As a result, the consistency, precision and cost-effectiveness that is a part of the first step of drug discovery hasn’t yet been realized as part of the later stages.

At Arctoris, we are extending the value of automation further down the drug discovery pipeline, all the way through to the pinnacle stages – molecule development, hit-to-lead, lead optimization, and candidate selection. We are bringing consistency, standardization, digitalization and reproducibility to a wider range of drug discovery teams across biotechnology companies, pharmaceutical corporations and academic centers.

The other point to mention is that there has been a huge technical barrier, and cost barrier, to automation. Historically, this has limited most biotech start-ups or academic labs from accessing it.

The approach at Arctoris is trying to break down those cost barriers and allow any biotech start-up to be able to generate and access valuable drug discovery insights, at just the click of a button, without the need for millions and millions of pounds of investment.

LL: You specialize in the automation of cell-based and biochemical experiments, could you highlight some of the experimental capabilities you offer and touch on how you are able to support multiple stages of the drug discovery process?

TF: 
We developed and operate a cell-based, molecular biology, and biochemical platform, that already offers a broad range of the most frequent types of experiments required in drug discovery, including cell imaging, PCR, potency measurements and kinetic profiling. Our goal is to be able to eventually offer the full breadth of all relevant assays from these three constituent areas of drug discovery research – that is the focus of our on-going growth and expansion. Taken together, we are enabling researchers to get a much richer set of data about their candidates so that our clients and partners can make better decisions earlier on in the drug discovery process.

LL: Why should global drug discovery teams consider working with experts who put automation, robotics and AI at the heart of every experiment?

MIB:
I think there are two very important elements to this question.

Right now, for example, it is commonplace that a large pharma company might have 10, 20, 30 different sites with scientists working with slightly different reagents, slightly different assays, slightly different methods, which means any data generated is not easily comparable as it isn’t standardized.

However, a scientist based in Boston ordering an experiment today using our platform and at another site in London ordering the same experiment a year later will get the same results, because it was performed under exactly the same conditions, with the same methods and in a highly structured and standardized way.

Ultimately, if data can't be pooled, it is not usable for machine learning applications. This is of growing concern in the drug discovery industry, as the unique opportunities offered by machine learning are more and more appreciated.

But there is a saying within the machine learning community: “garbage in, garbage out” – simply put, any model can only be as good as the input data, and that is crucial. Our platform offers researchers the opportunity to use machine learning in a productive and fruitful way, because the input data is actually of high quality.

LL: Could you tell us more about the creation and continued development of the world’s first fully automated, robotic laboratory, dedicated to drug discovery?

MIB: 
The first step in building the company was research and development (R&D) when we soon realized that the general principle of lab automation, whereby a single experiment is run a million times, was far narrower than what we wanted to build. We instead wanted to build a laboratory capable of running dozens of different experiments for dozens of different clients on one integrated system. So, we worked with one of the world's top experts in robotic control software, and we developed our own robotic system with software capabilities that could support the greater flexibility and modularity we needed.

Based on our early R&D work, we then built our first prototype facility. This prototype demonstrated all of our capabilities and we were able to validate our platform technology. Based on this prototype we have now moved into the facility you see today – the world’s first fully automated, robotic laboratory – which applies these validated, proprietary technologies to a broad range of assays, including cell-based, biochemical and molecular biology.

We have moved from R&D, to prototype, to a production-grade facility that now supports drug discovery teams anywhere in the world.

LL: The term “lab of the future” is used a lot. How close are we to achieving the lab of the future or do you think we are there already?

MIB:
 The term “lab of the future” can mean many different things; for example, helping scientists to collect and report data more accurately with an electronic lab notebook or an innovative piece of equipment.

To take the “lab of the future” to the next level, we are building a laboratory that allows every scientist in the world, regardless of where they are, to access state-of-the-art equipment, and cutting-edge methods.

You could be an established professor in an advanced life sciences hub in Boston or San Francisco  – or an early career researcher in a setting with far more constrained resources; right now your chance of drug discovery success is heavily influenced by that. From our conversations with scientists in Sub Saharan Africa or South-Est Asia we know that access to laboratory infrastructure is often limited. The team at Arctoris has now built a platform that is hopefully going to level the playing field in drug discovery by providing this access.

TF:
I think the term lab of the future exists because scientists have realized that there is a huge opportunity to apply advances that are happening in other industries – cloud technology, computing technology, Internet of Things, connected systems.

There are so many benefits that haven’t reached the lab yet, and that is what we want to achieve.

One of the main benefits will be democratization of science. Currently roughly 75% of life sciences research stems from just 5 countries – typically the wealthiest countries – due to the prohibitive costs of R&D. Our vision of the future is that drug discovery breakthroughs and biotech companies will be popping up all over the world, simply because all it takes is a couple of smart people, an idea, and the passion to take that forward. We would like to enable this.

Martin-Immanuel Bittner MD DPhil and Tom Fleming MChem, cofounders of Arctoris, were speaking with Laura Elizabeth Lansdowne, Senior Science Writer for Technology Networks.

Meet The Author
Laura Elizabeth Lansdowne
Laura Elizabeth Lansdowne
Managing Editor
Advertisement