Axial - Ginkgo Bioworks

Surveying great inventors and businesses

Axial invests and partners in early-stage life sciences companies. If you or someone you know has a great idea or company in life sciences, Axial would be excited to get to know you and possibly invest in your vision and company — info@axialsprawl.com

To simplify this newsletter, if you want to receive:

Image result for ginkgo bioworks logo

Ginkgo Bioworks is one of the canonical synthetic biology companies of this era. Led by five incredible founders working together going on two decades, Ginkgo is bringing new technologies to life sciences and just as importantly, going to the edge to explore new business models. The key purpose of this case study is to understand how the history of Ginkgo and synthetic biology in general is driving the boldness of the company and many of its key decisions. Synthetic biology is littered with countless blowups, but the potential for the field to transform human health, the environment, and beyond is so high diving head first and being bold is needed.

I remember hearing about a small company out in Boston Harbor in 2011. Some of my friends who I was doing summer lab research with were always telling me how Ginkgo built out an awesome automation system for bench work. Spending my summers doing PCRs and picking worms made me really envious of the capabilities Ginkgo had. Ever since then I began tracking the company.

Ginkgo has been 20 years in the making. The founders met in 2002 at MIT with two of the co-founders (Shetty and Che) in the Knight Lab and two (Kelly and Canton) in the Endy Lab. Synthetic biology was just forming. Just as physicists entered biology and transformed it - https://axial.substack.com/p/axial-what-is-life Engineers began entering biology in the early 2000s and are transforming the field. The founders first worked together on organizing the first synthetic biology conference. Together, they helped form iGEM (along with BioBricks) and OpenWetWare. After getting their PhDs in 2008, the group of five (Shetty, Che, Kelly, Canton, and Knight) formed Ginkgo to bring their work to more people. Raising capital was hard but a loan from Knight and a grant from the NSF - https://www.sbir.gov/sbirsearch/detail/192832 got the company to work. Then and now, this persistence and their cohesiveness as a team has been a major driver for Ginkgo’s success. Where others would not try, the five founders had the risk-taking ability to do something new. I’ve always thought of the Ginkgo founders similar to the Furious Five - both of them:

Ginkgo was premised on the idea that the organism is the product. Whereas companies like Amyris and Solazyme were focused on being fully integrated and investing large amounts of capital to build out 200K liter fermenters, Ginkgo learned from their mistakes and focused on a licensing business model. To focus on the organism as a business requires a better toolkit to design them. Around Ginkgo’s founding most companies were using metabolic engineering to design microbes to produce a specific molecule; the field was early and the work was artisanal by nature. Ginkgo set out to build a biofoundry to integrate biology, software, and hardware. For Ginkgo that means creating the compiler, debugger, and codebase for an organism. The idea is that the larger the foundry, the greater the economies of scale and lower costs per organism. More experiments creates more designs (i.e. larger codebase) that ultimately creates a virtuous cycle where overtime less work is needed to create a new organism. Then a larger biofoundry enables the company to engineer increasingly more complex cells (i.e. image below). That means Ginkgo can start at fragrances but end up developing cell therapies. Ginkgo would agree this creates a winner-take-all market where the company with the largest biofoundry can win the highest value deals and grow the most.

With this approach, Ginkgo is building out an Intel-like business model for biology. What really set their business model apart from competitors was the focus on 50/50 deals with partners where Ginkgo and their partners equally pay for the costs of a program. This is a deal structure that Regeneron actually pioneered in life sciences - https://axial.substack.com/p/axial-regeneron This approach requires a lot of courage and belief - you’re betting on yourself. For Ginkgo, they learned a lot of lessons from companies like Amyris especially inspired by the Amyris/Firmenich partnership. This early focus on 50/50 deals was a major driver for the company’s growth. Now, Ginkgo has pivoted, evolved, or however you want to put it to forming/investing companies or doing joint ventures to increase the demand for its biofoundry. Given this, figuring out what other businesses become possible because of Ginkgo’s biofoundry is important. What layers of abstraction have been built by Ginkgo? If Ginkgo works, then what?

With three major moats, Ginkgo is getting a lead in the field (Zymergen is there too and that’s another case study):

  1. The biofoundry

  2. The underlying designs of organisms generated

  3. The business model

Ginkgo has evolved from something small to something big. A small group of 5 scientists built out a unique business with an incredible culture that is a consequence of their persistence and mission. The founders’ continued boldness is deserving from all the large and small battles they’ve won. Now the company is in the position to bridge the gap between the potential of biology and products. It’s taken ~12 years to get here. For the business, this can power billions of dollars of sales. For society, new consumer products, new medicines, and a lot more can come to consumers and patients. Ginkgo is a company where its past more than its future dictates its decision making. Ginkgo has always taken a lot of business model risk. Ginkgo has always believed a biofoundry is monopolistic. Ginkgo has always believed that other companies should commercialize their organism designs. 

Key findings

  1. Progress in life sciences over the last 40 years has been tremendous. Genentech cloning insulin. Agtech stacking genes. Biology products from medicines to plants have been increasing in genetic complexity. Ginkgo is building out the infrastructure to push the boundaries here: engineer tens of genes in a microbe to make a specific molecule, using mammalian cells for food or therapies, create cells for environmental reclamation, and ultimately bring new functions to biology.

  2. A big issue in synthetic biology has been the overpromise of market size. The numbers are big, really big - energy, food, healthcare. This is trillions of dollars in economic output. But this doesn’t mean that biology captures all of it. The trick is figuring out which spots in these markets are more accessible than others.

  3. A key driver for Ginkgo’s ability to secure over 10 partnerships within a few years by ~2017 and grow that number steadily since then has been the use of 50/50 deals where Ginkgo shares the R&D costs equally with their partners and receives a royalty rate that depends on the product. This 50/50 deal structure led to more partnerships in a shorter period of time versus competitors leading to a lot of upfront R&D payments that likely led to more capital invested. For every business, the little things matter. You can walk into a movie theater in Amsterdam and buy a beer. In Paris, a Quarter Pounder with Cheese is a Royale with Cheese. For Ginkgo, the founder’s conviction of using 50/50 deals really set them apart from the competition.

Technology

Ginkgo’s origin story is intimately connected to the history of synthetic biology. The field was founded around the early 2000s. I would say the touchstone event was the invention of the repressilator by Elowtiz and Leibler -  the work showed the ability to predictively engineer a new behavior into a cell, a circuit with a triple negative-feedback loop of sequential repressor–promoter pairs. Essentially, the duo created a genetic circuit with autologous features to an electrical circuit. As one of my professors, Vlad Denic, put it - “you don’t understand something until you can control it.” The concept of bringing electrical engineering principles to biology was transformative. By focusing on reusable parts and using mathematical models to describe cellular behavior, synthetic biologists brought a new approach to life sciences where results can be compared to a model and iterated upon. In 2004, the field’s first major conference was held, Synthetic Biology 1.0 - https://openwetware.org/wiki/Synthetic_Biology:Synthetic_Biology_1.0, which the founders of Ginkgo actually helped organize. Scientists from biology to computer science showed up and helped set standards and an interdisciplinary culture in synthetic biology that still exists. These events were major influences on Ginkgo. Ideas around making biology just as much an engineering discipline as electrical engineering were being explored. A major movement toward creating a standard set of genetic parts was pushed. Labs across the world were building out layers of abstraction to engineer an organism. The time between the invention of the repressilator and Ginkgo’s founding (~8 years) was the intellectual starting point for the business to focus on building a biofoundry where synthetic biology could be scaled.

What is a biofoundry? That’s the key question to understand what Ginkgo has built out. Metabolic engineering, modifying cellular pathways to increase yields of certain products, has been around since the 1990s. Then the idea to use standard genetic parts to introduce new behaviors in cells emerged in the 2000s. But this work was artisanal: there was little-to-no automation, and production and assembly of DNA was a limiting factor. Metabolic engineering was essentially a series of elegant demonstrations to overproduce natural metabolites and products. Synthetic biology enabled the production of non-natural molecules. But there was still a scaling problem. Around the time of Synthetic Biology 1.0, there was increasing awareness that a framework and standard tools were needed to allow for the universal application of these tools across a wide set of hosts. Instead of building bespoke tools for each host. In essence, a biofoundry allows for the transfer of biological tools between hosts (i.e. from microbes to men). A biofoundry has to bring software, automation, and a whole set of features to make biology experiments reliable and reproducible. So it’s just a set of experiments instead of an ambiguous term that can be misused. This is true for any platform in life sciences - they’re just a set of experiments. For a biofoundry, the key experiments are:

  1. Picking the host whether it’s E. coli, S. cerevisiae, or something else

  2. Going onto a computer and designing some DNA to put into the host. You’re probably using something like Benchling - https://axial.substack.com/p/axial-benchling Over time, more genetic parts become available - think a Sears catalog of parts for particular hosts and the corresponding behavior or product produce - and this process becomes a lot easier and more fun. In this step, there is nitty gritty work like optimizing retro-biosynthesis algorithms, promoter tuning, and finding the right gene copy number.

  3. Synthesizing the DNA and assembling it (these two steps have gotten a lot easier, super-exponentially easier actually, over the last 2 decades; second image below)

  4. Transforming the host with the DNA

  5. Fermenting the host or using a cell culture system depending on the organism type. The value prop often is to replace plant extraction. Culture Biosciences is doing some great work here. This step is pretty comprehensive. An AWS for fermentation would be helpful. Substrate uptake is optimized, secondary metabolite overproduction might have to be blocked, biphasic induction could be used to shift flux toward product formation; a lot of heavy lifting here - some data-driven and some based on centuries of experience around fermentation.

  6. Recovery and purification of the end-product whether its spider silk - https://axial.substack.com/p/axial-bolt-threads or rose oil. This is where a lot of money can be gained or lost. Titer, yield, and productivity are the key metrics here.

For Gingko, the company has a lot of robots, orders a lot of DNA, and has a lot of people to do these experiments. For the company, from its first instantiation of the biofoundry to the current version, progress was the ability to automate across each step. More automation leads to more consistency between experiments, quicker turnaround times, and reduced labor costs. Think of it similar to automating the assembly for cars. Automation was a major competitive advantage to a particular business. In Ginkgo’s case, a more automated biofoundry allows the company to iterate across more experimental runs to build out a larger and larger codebase (Step 2), which makes figuring out which experiments to do much easier.

For Ginkgo’s biofoundry or any one in general, the key metrics are: how many tests can be done and at what cost (fully loaded including labor and equipment)? What organism types can be used?

Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5544601/figure/F1/

Source: NIH

But synthetic biology hasn’t yet brought a wide set of new materials and molecules to market. Even though biofoundries are being built out, they’ve yet to be used for commercial-scale products. As consumer demand more sustainable products and industries need to reconfigure their supply chains, Ginkgo and companies in synthetic biology are in the position to meet their needs. What’s really exciting is that over the last two decades, progress in synthetic biology has mainly been driven by the desire to push the field forward with some customer demand for things like ingredients and chemicals. As that demand increases, more progress in synthetic biology is likely to be made. Moore’s Law is not a technology law, it’s an economic one.

Progress in life sciences over the last 40 years has been tremendous. Genentech cloning insulin. Agtech stacking genes. Biology products from medicines to plants have been increasing in genetic complexity. Ginkgo is building out the infrastructure to push the boundaries here: engineer tens of genes in a microbe to make a specific molecule, using mammalian cells for food or therapies, create cells for environmental reclamation, and ultimately bring new functions to biology.

Market

A big issue in synthetic biology has been the overpromise of market size. The numbers are big, really big - energy, food, healthcare. This is trillions of dollars in economic output. But this doesn’t mean that biology captures all of it. The trick is figuring out which spots in these markets are more accessible than others. It’s like a cheetah hunting the weaker, slower gazelles first. As a business or a cheetah, you don’t want to make your life harder by picking the wrong target. For synthetic biology, the wrong target was the oil industry. The 4 major markets here are: industrial biotechnology, agriculture, human health, and consumer products with applications from chemicals (commodity and specialty), fragrances, food, pharmaceuticals, flavors, to feedstock.

Ginkgo has the potential to pursue multiple large markets. The purpose of this case study is to understand what Ginkgo did to be in the position they are in now. The first market Ginkgo chose to focus on was flavors and fragrances (F&F). To market themselves here, Ginkgo called their products cultured ingredients - this refers back to their underlying biofoundry where the fermentation step removes plant extraction that is part of the traditional process. When Ginkgo was setting up to commercialize their work around 2013/2014, the F&F market was generating ~$20B in sales with the 10 largest companies (top 4: Givaudan, Firmenich, IFF, and Symrise) controlling more than 80% of the market. Now, the market is worth more than $30B with the same underlying market structure. 

The thesis for Ginkgo to pursue the F&F market was that plant (and sometimes animal) extraction creates supply chain complexity that fermentation can solve. Rather than sourcing raw materials from a diverse set of farms (for the F&F industry, covering over 250K hectares of farmland), Ginkgo could standardize sourcing raw material with a microbe. Early on, Ginkgo focused on products like rose oil with Robertet and ingredients with Archer Daniels Midland. The market choice is sound in terms of value prop, and the strategy of focusing on high-value, low-volume compounds first makes sense too. The issue is that ~95% of all F&F are synthesized from petroleum and not from a plant. So Ginkgo’s supply chain value proposition addresses 5% of the market, but the bar in terms of unit economics Ginkgo has to achieve is at least an order of magnitude higher because it has to compete with petroleum-based products. This is a major challenge for any synthetic biology company to reach price parity (I have a pretty extensive set of financial models on a lot of these F&Fs I can share). Until consumers in unison start paying more money for petroleum-free products or Ginkgo reaches price parity for certain F&Fs, most of the market is not completely addressable. But as long as the company continues to improve its biofoundry, Ginkgo will be able to address 100% of the market one day. The potential for Ginkgo to use biology for citrus oils that go into your orange soda and redefine the F&F market is there. Evolva’s vanillin product is a good signal of progress. Ginkgo is facing this dynamic in other markets as well, and every synthetic biology company faces these market realities where even though the number looks big, it’s not all addressable for them right now.

Business model

What sets Ginkgo’s business apart from others is its culture. Ginkgo’s business model is a consequence of it being founder-driven. The company bets on itself. In the 2000s, synthetic biology companies were making promises of replacing the petrochemical industry and raising billions of dollars. They engineered the microbe, built the plants, and commercialized the product. Being vertically integrated was capital intensive and required broad expertise across synthetic biology, biomanufacturing, product commercialization, and when the companies failed to launch commercially-viable products, many of them died or pivoted to smaller markets. Ginkgo learned from this and decided to focus on what it’s good at - designing microbes and licensing their organisms to companies who can bring the resulting products to market. This approach is very similar to what Adimab achieved licensing humanized antibodies - https://axial.substack.com/p/axial-adimab Then Ginkgo uses it biofoundry to design organisms specific to a market problem and their partners pay a upfront R&D fee and royalties on commercialized products to Ginkgo. Rather than investing the capital to be vertically integrated, Ginkgo decided to focus on upstream organism design and license out their products. In a very similar manner as Intel.

A key driver for Ginkgo’s ability to secure over 10 partnerships within a few years, by ~2017, and grow that number steadily since then has been the use of 50/50 deals where Ginkgo shares the R&D costs equally with their partners and receives a royalty rate that depends on the product. This 50/50 deal structure led to more partnerships in a shorter period of time versus competitors leading to a lot of upfront R&D payments that likely led to more capital invested. For every business, the little things matter. You can walk into a movie theater in Amsterdam and buy a beer. In Paris, a Quarter Pounder with Cheese is a Royale with Cheese. For Ginkgo, the founder’s conviction of using 50/50 deals really set them apart from the competition.

Early on Ginkgo was focused on building an ecosystem. Just as the founders helped organize Synthetic Biology 1.0 that helped define the academic field of synthetic biology, the founders are using Ginkgo to define the commercial applications of synthetic biology. The future of Ginkgo is centered around building an ecosystem of companies that use Ginkgo’s biofoundry. The shift toward a holding company model is a consequence of the underlying market structures (i.e. consolidation and how much can biology address now). It’s a risky approach that other companies in the field wouldn’t dare to try. But the company’s boldness is a consequence of the founder’s knowledge base of how synthetic biology developed and what it needs to succeed and their willingness to always bet on themselves dialing up risk to the Nth degree. Ginkgo has been able to strike up a wide set of partnerships and have added joint ventures to its business model. Actually, one of the offshoots from Ginkgo could become larger than Ginkgo itself - the companies Intel powered ultimately became much larger than Intel. This future is both exciting and terrifying. The potential to transform industries and products with synthetic biology is high. But Ginkgo and a lot of companies in the field need to show the ability that synthetic biology can be used for commercial-scale products. Ginkgo is a case study of how the early-stages of a scientific field can influence the trajectory of a company. The company still retains the values of boldness, scientific rigour, and engineering principles to bring the power of synthetic biology (invented over the last 20 years) to consumers everywhere.

Image result for ginkgo bioworks yarkov

Source: Ginkgo Bioworks