Making Biology Easier To Engineer, For Engineers
This was a softball question for Jason Kelly, cofounder and CEO of Ginkgo Bioworks, who simply replied, "You make biology easier to engineer!"
If you type "make biology easier to engineer" into Microsoft Word, then it throws a grammar flag suggesting you mean "make biology easier for engineers." That's not what you meant, but our dear friend Clippy might be onto something.
Chris Gibson, cofounder and CEO of Recursion Pharmaceuticals, evoked the magic of Bell Labs and the importance (and difficulty) of melding together diverse disciplines under one roof:
Like other interdisciplinary and rapidly evolving fields, synthetic biology demands a new kind of scientific culture. The most significant breakthroughs will happen at the intersection of physical and digital sciences, but it takes effort and intentionality to bring together biologists, chemists, engineers, and data scientists in a constructive way.
In the life sciences, our academic system often teaches people to build upon prior work and assume its truthfulness. Culturally, that's very different from technical fields like data science, where you're taught to challenge conclusions. There is a natural tension that exists when these schools of thought come together – we need to cultivate that to encourage divergent thinking and drive innovation.
People and culture came up frequently in responses. One CEO told me most of the smartest synthetic biologists he's ever hired were trained as software engineers, but the siren song of higher salaries and stock options from large tech companies can be difficult to fend off. It would be interesting to ponder if the natural tension between disciplines also creates challenges in attracting talent. Maybe layoffs in information tech will quietly end up being a boon for synthetic biology.
Elliot Roth, cofounder and CEO of Spira, listed education silos as one of the key challenges standing in between synthetic biology and winning at a global scale.
The strength of an industry is predicated on the strength of the workforce. The software industry has "won" in the world due to the speed and adaptability of the workforce. One can code anywhere in the world with a laptop and a Wi-Fi signal. Scientific work right now is horribly disfigured and crippled by ancient hierarchical educational models that predicate a five-year PhD, earned by learning outdated concepts using PDFs that lock up scientific information and keep ideas trapped in the ivy-covered towers of academic institutions. A revisiting of the educational system is needed.
The fluidity and adaptation of the workforce, as well as the openness of scientific information, can and will lead to an explosion of scientific creativity. The dream is that cloud labs, auto-pipetting robots, fluidics, countertop DNA printers, nanopore sequencing, and a plethora of other technologies will see lower costs from an open hardware and software movement, and also lead to more and more companies that come from a garage, DIYbio lab, and remotely to win out of left field.
He offered a Mark Twain quote to summarize this point: "I never let schooling interfere with my education."
Although educational models and accessibility introduce friction for building a competent and cohesive workforce, the challenges are more nuanced. Making biology easier to engineer won't help much if biology is still expensive to engineer. The cost of an ambr250 minibioreactor system, or Tecan liquid handling robot, or many reagents remain out of reach for many academic labs, let alone a new startup. There's a market for these tools, but it's mostly selling to drug developers and reshuffling venture capital among startups selling to one another.
Will Canine, cofounder and former executive of Opentrons, says one way to win is to democratize existing technology. "For example, take a tool and make it 10x cheaper and easier, rather than focus so much on creating some expensive, novel, never-been-seen-before thing."
One of the field's brightest minds agrees. Keoni Gandall, biohacker and cofounder of Trilobio, argues we need to tackle societal and technical obstacles. On the societal side, we need to create a belief that the cost of biotech should be dropping. On the technical side, we need affordable hardware that can execute truly generic protocols, much the same way software can boot up on one robot or 100 robots, without humans.
If the Human Genome Project could be summed up with "Bought the book; hard to read," synthetic biology could be summed up with "Easy to write; nothing to say." You can put together words, but not sentences.
For me, winning would be increasing humanity's ability to create with biology. While both synthetic biology and biohacking have made a splash with their branding, neither have dramatically impacted how bioengineering is done. We simply can't do enough experiments right now to make synthetic biology cost effective for anyone to try out crazy ideas. Those experiments cost too much.
Anyone should be able to access and use the robots with no knowledge of how to physically do biology, and protocols should be able to be run locally or on a cluster of thousands of robots with identical results.
This might sound confusing to the uninitiated, but Keoni is right. (As a general rule, when it doubt, heed the prophetic words of the red polo.)
The field of synthetic biology today isn't offering much new aside from high-throughput experimentation for those who already know what they're doing. Much of what gets called synthetic biology is really just applying molecular biology techniques at larger scales and faster speeds. Scales are larger and speeds are faster relative to moving tiny volumes of liquids around by hand, but they're not enough to engineer new capabilities into biology – we're just better able to see what capabilities are already programmed in.
This is admitted by the leading technology platforms. The following is taken from a 2016 letter between the San Francisco Planning Department and Zymergen as the two tried to determine the proper classification for its robotic foundries [quotation original, emphasis mine]:
… while Zymergen uses propriety software and advanced robotics systems, it does not use advanced biological techniques. The biological techniques employed are "basic protocols developed in the 1970s and 1980s and applied broadly in the fields of fermentation and analytical chemistry" and are conducted in a laboratory with Biosafety level 1 controls.
That's another way of saying it's still early, but Andrew Hessel says not to be discouraged.
The founder of Humane Genomics and catalyst behind the Genome-Project-write (GP-write) said it's important to acknowledge the challenges of the present while not losing sight of the possibilities of the future. "Programming biology will one day be as commonplace as programming a computer is today. We're still in the earliest developmental stages of this technology and have only scratched the surface of what can be made. It will be overhyped in the short term and dramatically change capabilities in the longer term."
How to Turn Startup Success Into Commercial Success
"You own businesses, not technologies." – Ol' Maxxie
There was a clear distinction in responses between individuals at emerging startups and those who are crossing or have crossed the chasm into commercial-stage operations. There's no single reason industrial biotech and synthetic biology companies have failed as of this writing in December 2022, but overlooking commercial realities has played a dominant role.
A good place to start is acknowledging the most important omics is economics. Stanford University professor of bioengineering Drew Endy, cofounder of the iGEM Competition and one of the founding fathers of synthetic biology, reminds entrepreneurs, "Make a product that actually makes money. There's no hiding from the nuts and bolts of the cost of goods sold. Get it right. Don't start with bulk commodity products unless you are absolutely convinced the margins work at the scale you can realize with the capital that's actually available to you."
Many starry-eyed founders have conviction in their economic modeling and ideas, but that may blind them to what really matters. BioAmber had an awesome industrial-scale bioprocess for manufacturing succinic acid, but no one wanted succinic acid because it must be converted into more useful materials. The overall economics weren't right and there weren't enough customers.
It's striking that every response from an individual at a commercial-stage company mentioned the word "customers" – the only participants to mention that word at all.
Michael Saltzberg, CEO of Covation Biomaterials (formerly Dupont's industrial biotech arm), has long been one of the most pragmatic voices in industrial biotech. The company he helms recently celebrated the 22nd anniversary of the first commercial fermentation run for 1,3-propanediol (PDO), which is used in the Sorona brand of polymers and fibers.
The biggest issue for emerging companies is most go around asking, "who wants to see my shiny hammer?" In the markets we're involved with, no one wants to see your shiny hammer – sometimes they might even need a wrench. No one cares about synthetic biology. They care about, "how can you make my product more sustainable?" and "how can you help me make more money?"
Your first priority is the technical performance of your material and how it can be validated by customers in their products, followed closely by cost and sustainability. Many people grossly underestimate the effort to bring a product to market, especially outside of commodity chemicals.
At least half of the time and expertise developing a new product at Covation Biomaterials is dedicated to downstream processing steps such as purification. Dr. Saltzberg said the business employs just as many chemical engineers as bioprocess engineers, which helps to explain the company's durable success.
As we saw earlier in this article series, one reason startup success has most often led to commercial failure is the translational gap that exists in scaling up lab experiments to commercial scale processes. Fermentation runs in ambr250s don't perfectly match the conditions of your commercial scale bioreactor. Similarly, scaling down processes in the opposite direction isn't always feasible, which hinders visibility into downstream processing steps that occur after fermentation. Cell engineering is sexier than centrifugation and distillation techniques, but commercial success often comes down to the boring details like economics, purification steps, and customers.
Dan Widmaier, cofounder and CEO of Bolt Threads, echoed these sentiments. "My short answer is really about building businesses that win in synthetic biology. Not just the technology, which is the interesting aspect of your question because it could be sliced in a number of ways."
He continued, "My answer for a synthetic biology business is to focus overwhelmingly on customer pain points they are willing to pay for today. This means providing a wide process window for the sophistication of the biology technology used in the business and really focus on the products and services that win in the market today."
I know, I know. You want a quote with hacking stuff and hockey-stick curves. Eben Bayer, cofounder and CEO of Ecovative, delivered when sharing his perspective:
For me, winning in synthetic biology means marrying the classic S-curve of a new disruptive technology with the positive externalities that can come from using higher order biological systems to grow things. The goal is to hack capitalism by giving it what it wants: ever expanding fiscal returns on a resource limited planet by using a technology (biology) that is multiple orders of magnitude more complex than what our prior industrial revolutions have created.
Okay, just kidding. Eben concluded with more boring details about economics and customers:
Focus on finding product-market fit with biology that does something amazing your customers will love and pay for. Prioritize products with outstanding returns and return on invested capital profiles.
The Ultimate Measure of Success: Language
Time for a confession: Synthetic biology isn't an industry.
Synthetic biology is an approach. It's a collection of tools and how they're applied. It's about applying engineering principles to make biology reproducible across applications and industries.
The media and conference platforms might call it an industry, but that's to tickle algorithms and sell sponsorships. This is similar to the rise of the phrase "net stocks" during the Dot Com Bubble of the 1990s or "genomics stocks" during the liquidity bubble that's now deflating. It's marketing, but not something that helps you think more clearly.
Words and language change over time. Like literally (eh?). No one talks about the "computer industry." We just call it "information technology" or simply "the tech industry" and keep the conversation moving. It means everything because it is everywhere.
A decade ago, there was a fierce debate about what to call synthetic biology. It made sense at the time (or we thought it did). After all, how could you ask governments to fund something if the people building it didn't even know what to call it? In reality, it doesn't matter what it's named, but what it's used to build.
This is what makes language an important marker for measuring success. As Antonio Regalado of MIT Technology Review told me:
Synthetic biology is an aspirational brand and success is when you don't have to call yourself synthetic biology anymore.
Someone said that "synthetic biology" is the phrase used for things that don't work or aren't good for anything. If they did work or were good for something, you wouldn't call them synthetic biology, you'd just call it what it is. The best-selling drugs in the world are antibodies made from a synthetic gene or coronavirus vaccines from synthetic genes. No one refers to them as synthetic biology.
Indeed, if something works, you may not talk about it at all.
Perhaps the ultimate marker of success is when we stop talking about synthetic biology and drop the prefix "bio-" from what we're building. It's easy to forget now, but ethanol used to be a small byproduct by volume from petroleum refining. When we first started manufacturing large quantities of it through fermentation, it was referred to as "bioethanol" by regulators and industry.
Today, the United States sports 21 billion gallons per year of "bioethanol" manufacturing capacity spread across 275 facilities. We just call it ethanol again, because how the hell else would we make it?
"Same with robots," continued Antonio. "If it's a big metal man, it's a robot. Once it works and is in use, it's not unfamiliar or strange, no one calls it a robot. Think of your dishwasher at home."
This is why I've simply started referring to all this as "living technology" and how Solt DB got its name. It was founded as BioMap, but we couldn't figure out how to monetize the database platform and therefore didn't launch (a good thing to realize before you launch). Before I put it back on the shelf in 2018, I wanted to change the name and branding. As anyone who's named a company before knows, at some point you inevitably look up Latin and Greek words, because that's cool. But it seemed so cliché to put "bio" in the company name. Everyone does that. That's not cool.
Latin and Greek did save the day though. Biology comes from the words "bios" meaning life and "-ology" meaning branch of study. If biology is the study of living things, then Solt DB would be the study of living tech.
Most important, calling it living tech deprioritizes the label. Similar to "information tech," we can all keep the conversation moving and focus on the important things. You know, like how to build something that's truly successful.
Note: Astute readers will notice the above responses are all from dudes. I reached out to a half dozen women, including some of the smartest executives in the field and someone who I view as the absolute best analyst in living tech. Unfortunately, much like my dating life, they all declined to participate or didn't respond in time. I did try.