Abstraction Layers and the Foundry Era of Biotech R&D
Making biology easier to engineer means developing new strategies to manage biological complexity. The Ginkgo foundry is a new way to organize biotech R&D operations and abstraction layers a new way to organize biological thinking.
Transcript
Biology labs as we know them today are about 150 years old1. Maybe it's time for a change?
If you traveled back in time to the days of Louis Pasteur, you could visit a lab and it would be more-or-less familiar: long benches, fancy apparatus, human scientists moving test tubes around.
Science was happening before that, but the people and the places were organized differently. People were working in castles or in garden sheds. There was a lot of alchemy going on. It was kind of cool looking but also super weird. Like look at this thing. What even is that? Anyway sorry - not the point.
The point is that laboratories as we know them were an invention of a certain era in human history. They arose to meet a new technical need. Some time around the 1850s, science became complex enough that it was no longer practical to do it in a garden shed or with whatever this thing is. The laboratory itself is a kind of technology, a built environment designed for doing research and development.
Fast forward 150 years and biology laboratories are still built with the same basic design. Obviously, the equipment is different. Mouth pipettes became hand pipettes, then liquid handling robots. Those were important changes, but not fundamental to the organization of research. They're better tools for the same workshop, better parts for the same machine.
I believe that the foundry at Ginkgo Bioworks represents a fundamentally new way of doing biotech R&D. I think that it isn't a laboratory in the historical sense. The people, the work, the data, the thinking, are all organized differently. Louis Pasteur would not understand what is happening here. But I bet he would want to.
As with the original invention of the lab, Ginkgo's foundry is a response to increasing complexity. "We make biology easier to engineer" is our mission. At this moment in history, biology has become too hard for the lab.
Biology workflows are too big to perform by hand, so we need automation. Biology creates more data than we can use effectively, so we need AI models to process it. Biology is expensive, so we need economies of scale to bring down the unit costs. The world needs more biology, so we need to make it available as a service to the many companies that are building the bioeconomy.
What happens when an industry changes to meet the demands of increasing complexity? The modern history of tech can offer a model. As a synthetic biologist, this example is irresistible to me because we love to make comparisons between biology and computers. The analogy isn’t perfect, but I think it’s useful here.
Computer engineers manage complexity by organizing it into abstraction layers. Transistors, for example, combine to form logic gates, then integrated circuits, processors, and networks. Simpler devices are standardized and then composed to create more complex devices. This makes it easier to build new technologies at the top layer because you can trust that the deeper layers will handle the details.
Although the structure of biology is naturally occurring and not engineered, thinking in terms of abstraction layers can still be very helpful. DNA base pairs are organized into genes. Genes make RNA and protein. Proteins are organized into pathways that produce a phenotype. The phenotype of a cell determines how it behaves and what biological products it makes. You can build new products at the phenotype layer because you can trust a partner like Ginkgo to handle the DNA details.
In the tech world, new layers of abstraction can be transformative. They change where and how new technologies are developed. The first transistors were invented in a classic physics lab. But the industry didn’t stay there for long.
_(14568564798).jpg){kind=link}
As new layers emerged, they specialized, with new facilities, new kinds of human expertise, and new business models. Chip fabs, consumer electronics, server farms, GitHub.
Because of this history, the tech world has the habit of organizing complex projects with layers of abstraction. When tech developers bring new products to market, they don’t start by kitting out a lab and building their own transistors from scratch. They build on top of an existing tech stack.
I think that biology is moving in the same direction. Classical labs will continue to exist, but they will be only one of many spaces working together to solve the hardest problems. We’ll see increasing awareness of the need to create specialized facilities for managing different kinds of biological complexity.
That’s why Ginkgo, in my view, isn’t a lab. It’s a foundry for cell programming. Biological developers build their products on top of our platform, trusting us to handle the details at the DNA layer. The abstraction layers organize complexity and biology gets easier to engineer.