By Steve McCann | October 18, 2024
For Ben Thomas, it was the living creatures part that almost turned him off biology completely. It’s not that Thomas has any issue getting his hands dirty — at one point, in fact, he wound up mucking about on a farm. But all the blood and guts and single-celled organisms felt too conceptually messy for Thomas’s analytical mind, and he veered away from his post-secondary life science studies into cooler disciplines like corporate finance, where the only cells he needed to wrangle could be found on spreadsheets.
Somewhat ironically, Thomas’s love of data took him right back where he started. After a few jogs in the road, he realized that, as he puts it “computing and biology had become one.” In 2021, Thomas founded AMPLY Discovery, a Belfast-based company that uses artificial intelligence and machine learning to accelerate the drug discovery process. AMPLY excels at identifying and procuring medicinally promising molecules from nature, which are subjected to rigorous testing. With the help of AI, this happens exponentially faster than a comparable process in a traditional lab. This method can home in on multiple candidates at once, and is cheaper and less labour-intensive than more conventional approaches — a game-changer in a field that is notorious for its glacial pace.
Here, Thomas shares insights about drawing inspiration from the natural world, applying stock-market principles to drug discovery and ensuring that potentially life-saving medications make it to the people who need them most.
How does a person make the leap from corporate finance to precision drug discovery?
Almost everybody in my extended family was a biologist. I did biology at university, then rebelled against the family tradition and went into computing, then the financial industry — a real corporate job. I worked there until 2008, when I had a crisis and felt I was wasting my life making rich people slightly richer. I quit my job and became a farmer in France.
Why farming?
I thought I would do all of the tangible stuff a human being has to do. Could I grow my own food? Kill a pig? Build a barn? And I wanted to learn a foreign language, live in a different place — that sort of stuff.
That vocation doesn’t seem like a natural precursor to what you’re doing now.
I eventually came back to the U.K., went into international politics, then pivoted again. The important part of this story is that I realized computing and biology had become one: Biology wasn’t squidgy, wet creatures and bacteria anymore. So I did a computational biology PhD.
What is computational biology?
It’s a process that digitally captures biological data from life forms and stores it on a hard drive. This changed the dynamic of biology from a discipline where you had to roll your sleeves up and put on a lab coat to one where you can make discoveries purely within a digital space. It gave me the opportunity to apply my skills in computational areas to real living things. The clever bit is that there’s technology that lets you print molecules out from scratch, like an inkjet printer.
How exactly do you print molecules?
I’ll take 100 to 1,000 potential molecules that our system has identified and literally have them printed out as a little white powder in a tube. Let’s say I think a molecule will kill a particular type of bacteria. We go into the lab, culture the bacteria and try to kill it for real. You then tell the AI program what it got right, what it got wrong. That means with every cycle, it gets more and more powerful.
So, does that mean nature is no longer part of the mix?
Finding novel things in nature is how we’ve always found drugs. When penicillin was discovered back in the 1920s, Alexander Fleming found a fungus that started growing on his petri dishes, and it was killing the bacteria around it. So what we’ve done is take that idea of discovering promising molecules in nature, and digitizing and printing those compounds. For example, the Peruvian poison arrow frog, which lives in these damp conditions in the rainforest, produces compounds to prevent bacteria from growing on its skin. So we get that organism, digitize it, take a snapshot of the data, and test it. So it’s not really any different from, like, people going off into the rainforest on a canoe and finding a new sort of plant or something like that. It just means you can make that process vastly quicker, because you’re doing it all digitally.
Can you walk me through the key differences between your approach and how pharmaceuticals are traditionally developed?
We’re actually a bit like — you know the tradwife movement, where people live by the old codes of society? OK, yeah. So, going and finding something in nature and appropriating it was how we’ve always found drugs. Then it went into a phase where it was like, “We don’t need nature — we’ll just generate molecules ourselves, using chemistry.” Our approach is actually a return to more of a traditional drug discovery chain, but we’re slapping AI over it in the digital portion. We’re sort of saying that the earth is essentially acting like a massive AI, a kind of computer with all the interactions between forms of life — everything is trying to defend itself or kill other things for an advantage. We’re appropriating those interactions and connecting those to the way we’ve always made drugs, rather than assuming that as human beings we can shortcut the process and get a computer to do it.
Speaking of computers — how does your finance background factor into what AMPLY is doing today?
I applied the concept of stock trading to biological data. There’s all sorts of metrics that go into deciding what stocks to buy. Essentially, you treat these big hunks of digital, biological data like a market you’re exploring, as though you’re a biological fund manager. It’s like a digital prospecting system.
Does faster drug discovery accelerate the process of bringing treatments to market?
You sound like my investors. You can’t — and shouldn’t — shortcut clinical trials. We’re no different than any other drug discovery company in that sense. I can’t bypass the massive hurdle of getting a drug into a market, but I can very quickly find new candidates for an underserved disease that big pharma companies don’t want to direct their money toward.
Any major breakthroughs?
We have thousands of novel molecules that do amazing things. We’ve made huge strides in particular types of cancer, specifically triple-negative breast cancer. But making a drug isn’t just finding an active molecule that does what you want it to do. There’s also the formulation of the drug and releasing it at a particular rate. Hopefully, some of the work we’ll be doing in Canada is finding a partner to help us with the formulation component.
What are your hopes for your IUK partnership with MaRS?
In Toronto, there are a lot of specialist hospitals that deal with cancer, and there are some Canadian companies that are very good at drug formulation and delivery. There’s also a lot of interest because the U.K. shot itself in the face with Brexit, and has isolated itself globally, so we’ve shifted to thinking more cross-border. Canada is a really good fit, because there are similarities in the health system and the language — and truth be told, if you can get a foothold in Canada, it does make it a bit easier to access U.S. funding.
Is there any particular disease you’d like to help cure?
I have a childhood friend who is being severely affected by triple negative breast cancer and my father was basically killed by leukemia, so one of those two. If I could solve that then I would be happy to have left some sort of mark on the world.
This interview was edited for clarity and length.
Learn more about how MaRS and Innovate UK are bringing new breakthroughs to Canada.
Photo courtesy of Amply; background image source: iStock