To write What’s Your Bio Strategy? we’ve interviewed dozens of academics, business leaders, and entrepreneurs defining synthetic biology. The following is an excerpt from one of those interviews. To receive this interviews in your email box, subscribe to our newsletter.
Among the people making biology easier to engineer, Tim Gardner stands out. Along with his PhD supervisor Jim Collins at BU, he wrote one of the seminal papers in the field in 2000, Construction of a Genetic Toggle Switch in Escherchia Coli – a functional biological switch, made from genetic parts.
Later, while at Amyris, he led the engineering of yeast strains and pioneered process technologies for the large-scale bio-manufacturing of renewable chemicals. He then founded Riffyn to create [process design, experiment design and analytic] tools to accelerate innovation in research and development.
Karl Schmieder: Tell us how you started Riffyn?
Tim Gardner: I’ve worked in biotech for twenty years – as a student, academic researcher, and founder. I held leadership positions in medical and industrial companies. Time and time again, I kept running into horrible research and development inefficiencies.
By that, I mean it takes too long to get products to market. There are too many errors and mistakes and too much effort wasted on fundamentally unproductive pursuits.
Not because the science is hard and not because biology is complex, but because researchers are blocked by the difficulty of communicating scientific results and information to each other in reproducible, scientific methodologies.
Scientists are still relying on a means of communications that was developed 400 years ago for a handful of scientists that had regular contact to debate narrow ideas. That eventually evolved into publications. Today it’s presentations, spreadsheets but those technologies are inadequate for communicating complex, sophisticated ideas.
As a result, people don’t have a clear understanding of what their colleagues are doing, how they’re doing it, and whether the results are trustworthy. The bottomline is
I started Riffyn to solve this with computer aided design (CAD), statistical data and an analytics approach to science experimentation. That way, you start at designing an experiment, improving it iteratively and communicating it as structured, unambiguous, visual data based on real-time data acquisition. Riffyn tells you how an experimental design is working, helps you make it better, and assures you get the results you want so that anyone can look at your data and get the same results. This is the same way you would design parts for a car.
We believe the impact is significant. We have examples where we can cut time to market in half and double the productivity of R&D organizations. When you do that, you reduce the capital expenditures of new products and you increase the certainty that you’ll get a better result.
John Cumbers: How does Riffyn impact an R&D organization?
Tim Gardner: When business decision-makers have to decide whether to invest in a new product, they look at time and investment. If they know they can develop a biological product in two years with a million dollar investment, they’re more likely to build that product. If it’s going to take four years and cost $10 million, they’re never going to make that product – especially since it might cost five times as much when they factor in the uncertainty. They’ll never get your return on investment.
To increase the size of the bio-based economy, substituting bio-based products for those that would be made from chemistry and petroleum, we need to reduce the cost of developing those products. If we can do that then developing more specialized products is acceptable. We’ll stop looking for the billion dollar blockbusters, which are few and far between. We’ll have a lot more entrepreneurial success and investors will be happy because we’re delivering on the promise of the biobased economy.
Karl Schmieder: Can you elaborate more on what Riffyn enables?
Tim Gardner: Riffyn is about integrating data from people, instruments, automation systems, databases, spreadsheets – anywhere you’re capturing data. We take all that data and organize it around your experimental design processes.
Right now, we don’t control robots or provide automation, but we have an API that facilitates automated workflows. Our idea isn’t to automate the world – it’s to assure the capital investments you’ve made are working more efficiently and all the data is captured in a uniform, repeatable structure.
For us, that results in a software system oriented toward the scientists in a lab and the director – not toward automation.
Our thesis is that the solution to faster, better, cheaper drugs, and faster, better, cheaper bio-based products is predictability of information. It’s about integrating information and making better informed decisions. It’s not necessarily about fancy robots or magical tools.
John Cumbers: It sounds like you’re taking both a scientific and engineering approach to solve a problem.
Tim Gardner: I’ve always taken an engineering approach to science. But I think it’s hubris to think engineers can approach science without learning how science works, or how a scientist works.
John Cumbers: People often compare the tech industry to the biotech industry. The tech industry started on hardcore science. Creating the microprocessor, you could argue, was a scientific challenge. You could call the explosion of computing and the Internet, 95% engineering. The economic value of the 1960s through the 1990s could be credited to a boom in engineering. When you now look at Google, Facebook, Netflix and Palantir, the need is for scientists – data scientists and statisticians. The tech industry is now going back to science – big data, machine learning, analytics. I wonder if we’ll see the same thing with biology. We’re now hitting the period that requires engineering. We’ll see productivity gains and a lot of great things made with less science involved. Then we’ll swing back toward a period of science and analytics to build on top of that.
Tim Gardner: You make a great observation.
The idea that scientists are being paid more or are delivering more value or are in greater demand is not entirely true. It’s hard to hire engineers.
The place where you can do magic is where you clear the cloud of uncertainty and turn that into something usable. That’s why the data scientists are in demand and hold such a powerful role, because they have the potential to reduce uncertainty.
I would argue though that a lot of engineering is more empirical than we give it credit. There’s a lot of experimentation that goes on in engineering. The airplanes might be viewed as an engineering problem but Orville and Wilbur [Wright] were performing fluid dynamics experiments to understand how to engineer propeller in a wing. My hope is not that you see science take over engineering – or engineering take over science – but that you see the synthesis of them.
Karl Schmieder: What do you think non-biotech businesses need to know about synthetic biology?
Tim Gardner: When it comes to health and you’re solving a biological problem, you need to know biology.
The environment is also a biological system. So, if you don’t understand the nitrogen cycle and the transformation of matter by biological processes, you can’t come up with good solutions to environmental problems.
Let’s take the most extreme example, radiodurans, an organism that can absorb the radiation equivalent of a nuclear blast. It’s a crazy polyextremophile. When it absorbs radiation, it shreds its genome into tiny fragments that it stitches back together to keep going.
There are organisms that can detect light or transform electricity into energy for survival. Or, think about how incredibly efficient muscles are compared to the hydraulics or batteries that you might put into a robot. It’s extraordinary.
If we want to use those properties to make the world a more efficient, higher performing, more enjoyable place, then we need to learn how to learn from nature.
# # #