The promised democratization of biology plays out as industry frontrunners consolidate early leads and bio-preneurs consider the sheer breadth of opportunityToday, anyone who can code can win in the internet economy. Thanks to services such as Amazon’s AWS and Apple’s App Store, only basic coding skills are needed to develop and distribute lucrative apps. Similar enabling infrastructure is taking root in the field of synthetic biology, a scientific discipline that uses genetic tools to engineer microbes for a wide range of downstream use cases, from manufacturing the screens in our smartphones to producing the food we eat. A recent McKinsey report estimates that there are at least 400 unique applications of synthetic biology across fields as diverse as medicine, agriculture, food, and chemicals. The potential market opportunities for end products alone exceeds four trillion dollars. Who will tap into this biological goldmine?The creation of synthetic biology infrastructure, including low-cost genetic sequencing, automated cloud-accessed laboratories, and biology-as-a-service providers, could enable a democratized ecosystem similar to that seen in mobile app development as bio-preneurs identify profitable use cases for synthetic biology technology. But despite these empowering tools, the importance of big data and machine learning in synthetic biology — an engineering problem with thousands of genetic and metabolomic inputs — provides a counterforce that contributes to the development of a consolidated, “winner-take-all” ecosystem within the synthetic biology space.The balance of power between the democratization of biology’s toolset and the centralization of essential datasets and algorithms will dictate whether profits are accrued by a few synthetic biology titans or by a wide array of companies and entrepreneurs.
Synthetic biology “apps” come in two forms: a product produced by a microbe (such as silk or food protein), or the microbe itself (e.g., a bacterium that can substitute for traditional fertilizer). In both cases, there are typically three steps in the product development: First, identify the use case. Second, design the microbe. Third, manufacture the end-product. Once a would-be bio-preneur has identified an “app”, there are both biology-as-a-service providers and low-cost, turnkey equipment manufacturers that drastically lower the expertise and capital barriers to entry for each of these steps in the synthetic biology process. Organizations such as Ginkgo Bioworks will cover your microbial design needs, Culture Biosciences can optimize your bio-manufacturing process, and a slew of biomanufacturing organizations can deliver on the end-product manufacturing. As this enabling infrastructure develops, synthetic biology product development could be so abstracted away from the core biology skill-set to enable even those without any specialized training to pursue cutting-edge synthetic biology “apps” — at least, in theory.
This democratization is conceivable in theory. But early pioneers in the field have smartly capitalized on the digitization of biological data, creating the potential for a consolidation of value in the field. The design of a microbe for any given purpose — to produce a medicine or to sequester CO2 in the air — is a complex process incorporating the interactions of hundreds or thousands of genes, proteins, and metabolic pathways. For this reason, human-led engineering may provide a starting-point, but only recent advances in machine learning can truly optimize this process. Early leaders in the synthetic biology space, particularly Zymergen, have developed massive chemical, genomic, proteomic, and metabolomic datasets, as well as near-fully automated laboratories to conduct high-throughput experiments that generate even more data every day. These datasets are then fed into machine-learning algorithms that predict the best molecule for a given purpose and the best microbe to produce that molecule. As the datasets grow, the machine-learning algorithms are perpetually trained to offer stronger and more optimal predictions.This self-reinforcing feedback loop of experimentally-derived data and machine learning optimization has resulted in a moat that competitors will find tough to contend with. With this platform, Zymergen expects to be able to discover the best materials for a given use case, and the most efficient microbe for producing that material. And they expect to be able to do this more quickly and more cheaply than any competitor without similar data and algorithms to leverage. Given this consolidation of ability, how could a synthetic biology competitor (or an aspiring bio-preneur) ever hope to compete?
The Zymergen juggernaut may seem intimidating to would-be entrepreneurs in the space, but there is still strategic space for bio-preneurs to target. First, the sheer breadth of applications for biomanufacturing and synthetic biology broadly precludes a frontrunner like Zymergen from competing across all these use cases — at least for now. With early identification of target applications, bio-preneurs can stake out profitable niches by developing persuasive biology and sticky commercialization models to disincentivize any future entrance by a company like Zymergen. This approach enables early-movers to establish a defensible moat around the production of certain relatively commoditized goods.A second approach relies on developing expertise around a field which the machine-learning engines are not built to optimize. In industries such as medicine or food science, researchers can discover new microbes and molecules for a given use case that Zymergen may not be best equipped to predict. For example, companies discovering and designing new food items, such as Nature’s Fynd, are playing in a niche that has not yet been made vulnerable to Zymergen’s brand of machine-learning enabled disruption.In a similar vein, companies inventing new ways to compete with Zymergen could find a competitive advantage in certain synthetic biology verticals. To this end, companies digitizing new types of data, including the next-generation proteomics championed by Nautilus Biotechnology, could begin to accumulate their own datasets that are advantageous within a given use case. Finally, bio-preneurs would be well-advised to consider the old adage, “If you can’t beat ‘em, join ‘em”. For entrepreneurs pursuing ingenious use cases for synthetic biology or innovative downstream business models that thrust engineered biology into the mainstream, leaders like Zymergen may not be a competitor, but a powerful and willing partner in the engineering of biology.