Drug discovery is becoming slower and more expensive over time, despite many advances in technology ... [+] and biology. Will the coronavirus pandemic shake pharma out of its inertia? Here are five learnings from synthetic biology R&D that just might change the game in pharma.

We have a love hate relationship with pharma. We hate them for taking our money, and now we pray to them to cure COVID.

Depending who you ask, it costs a pharma company $1 billion to $4 billion and 10-15 years to bring a new drug to market. Whats worse, fewer than 10% of drugs actually make it to market to help people.

(By comparison, SpaceXs flagship rocket system, Falcon 9, cost about $400 million and took five years.)

The observation that drug discovery is becoming slower and more expensive over time is known as Erooms law, which says the cost of bringing a new drug to market roughly doubles every nine years. This is despite advances at the intersection of technology and biology, including computational drug design, high-throughput automation, and the ability to read, write and edit DNA faster and cheaper every day.

Moores versus Erooms law: the power of computing per unit costs doubles every 18 months under ... [+] Moores Law, yet the unit cost of drug development has increased to the point where drug development has become nearly cost ineffective, following an inverse of Moores law (Erooms law).

Moores versus Erooms law: the power of computing per unit costs doubles every 18 months under Moores Law, yet the unit cost of drug development has increased to the point where drug development has become nearly cost ineffective, following an inverse of Moores law (Erooms law). Minie et al, Drug Discovery Today.

Why is it still so difficult to make new drugs? How are we ever going to bring down health care costs if each new treatment costs as much as putting a person on the moon, with patients ultimately bearing the costs?

There was another industry about 70 years ago that was mired in high costs and failing products: the auto industry.

The automobile industry performed admirably during the world wars, transitioning from making consumer vehicles to producing the trucks, tanks, airplanes, and even helmets, ammunition, bombs and torpedoes needed for the war effort, totalling one-fifth of the nations war production. But in the postwar era, quality engineering took a back seat to the wartime production mindset of build a lot of it now. By the mid-1960s, new-car buyers could expect an average of twenty-four defects per vehicle, many of them safety-related.

Around this time, a small number of nerdy rebels, led by W. Edwards Deming and Genichi Taguchi, offered a new approach to rethinking how we make cars. The more mature US auto industry wouldn't listen, but the fledgling Japanese auto industry emerging from WWII had nothing to lose. It was during this time that Japans well-built and functionally designed cars made that country the worlds leading automobile producera position it has never relinquished.

The product brochure for the 1975 Honda Civic pretty much says it all.

U.S. auto manufacturing eventually caught on and adopted a total quality management approach, which has brought an industry-wide culture and commitment to producing products that meet or exceed customer quality expectations. The result is that now every car you buy, no matter how cheap, works pretty much flawlessly.

Today, big pharma is afflicted by the same kind of widespread failure. Despite the staggering time and cost to make a new drug, nine out of ten drugs still fail in clinical trials. Its like throwing away nine out of ten cars at the end of the production line because they dont work!

How can any company that works like this survive? Could a similar transformation of the pharmaceutical industry repair this broken system?

Hopefully, that transformation may already be unfolding. In this case, the nerdy rebels are from the field of synthetic biology. If youve read my Forbes column before, you may know that synthetic biology combines advances in computation, automation, and our ability to read/write/edit genes to change the way we build things with biology.

Synthetic biology is rooted in academia, industrial biotechnology, and bioengineering, and its early practitioners dreamed that we could rationally engineer cells to solve society's needs more quickly, more effectively and more sustainably than conventional biological and chemical technologies. Perhaps most importantly, these dreamers brought a nave willingness to imagine the unimaginable, to try the impossible, to attempt the wild ideas that everyone else thought were crazy.

Some ideas, like a library of standard biological parts, have been difficult to achieve. Others, like engineering soil bacteria to increase the amount of nitrogen fertilizer delivered to plant roots, have the real potential to slash terrible nitrogen runoffs and help feed the next billion humans on Planet Earth. If we dont dream big, well never achieve many of our most important goals.

This applies to the pharmaceutical industry now more than ever.

One of my favorite nerdy rebels is my friend and colleague Tim Gardner, founder and CEO of Riffyn. As a graduate student at Boston University, he is renowned for having performed some of the early pioneering experiments in synthetic biology. After holding roles in academia and leading industrial R&D groups, Tim founded Riffyn with a mission of helping scientists spend less time sorting through mountains of data and more time asking important questions.

This problem is especially acute in pharma today. Theres more and more data, but a lack of interoperability of that data, leading to data fragmentation and lack of reproducibility which is one of the biggest issues facing scientific research today. In essence, Tim believes we need to re-think R&D altogether.

Talking with Tim really sparked the idea for this column, and many of the ideas I share here are his. So I asked him: What are the five lessons you have learned from your experience in biomanufacturing that pharma needs to hear? Heres what he said.

In the 1960s and 70s, the adoption of kanban and total quality methods helped the Japanese auto industry quadruple its productivity. Eventually, it did the same for US manufacturing. In the 2010s, Tim showed we could apply those same methods to the R&D practices in one of the first synthetic biology companies, doubling the productivity in the R&D organization overnight. More recently, the adoption of such methods and supporting digital technology helped one of the world's oldest biotech companies double its pace of development of a biofuels product with half the normal effort.

Quality methods work, and not just for manufacturing. They can transform R&D. The reason is very simple: when you have poor quality in your R&D, your results are buried in noisy data and you can't discern fact from fiction. Decisions become more like roulette than science. And like roulette, you lose a lot more often than you win. This wastes enormous amounts of time, money, and resources on dead-ends. In the auto industry, it meant throwing away a significant portion of industrial capacity on lemons. In the pharma industry, it contributes to the 90% failure rate of drug candidates.

No one really likes change. Change is unknown, uncomfortable, and arduous. It puts you in awkward and vulnerable positions where you don't always have good answers. It puts you at risk. Oftentimes fundamental change emerges from the young, because they are too nave to realize the risks they face, or have no prior reputation to protect from awkwardness.

When Tim was a 23-year-old graduate student, he stood up at the Office of Naval Research to present his research on engineered bacterial circuits. Sydney Brenner, a soon-to-be Nobel Prize winner, was in the audience. Brenner stood up, walked on stage, pointed at Tims work, and declared thats all wrong." One month later, everything Tim presented was working just as he had described. That work was published a year later and became the founding work for the field of synthetic biology.

Tim still deeply admires the late Brenner (see lesson 5), but we all can be blinded by our confidence and comfort in past beliefs and habits. We have to take care not to believe everything we think when we hear some of the "silly" work that young innovators propose. We need to check our biases and automatic impulses to squash the new.

As scientists, we are obsessed with false positive resultsthe possibility that our discovery isn't true. And that's a good thing. It's what saves us from the tyranny of mystical thinking. It's what separates science from philosophy.

But false positives are not the whole story, and a lopsided obsession with them can be destructive. We see evidence of this from patient groups demanding faster drug development and greater access to experimental medicines, even if it poses risks of failure or even potential harms to health.

The other oft-forgotten side of the false positive story is the false negative. Its the "yin" to the false-positive "yang. A false negative is a missed discovery, a result that might be the next life-saving drug but one that we just don't detect. An R&D engine that recognizes false negatives looks very different than one that only considers false positives. It leads to multi-stage (tiered) statistical testing designs, to statistical design of experiments, to high-powered data analytics, and to implementation of quality practices as early as possible in R&D (see lesson 1).

Attention to false negatives helped Tim and others cut the time to market for new biotech products by more than a year and shave double-digit percentages off the costs of multi-million dollar production runs.

Scientific R&D has a broken relationship with data and software. Both are treated as an afterthought. There is the feeling in the mind of a product scientist that you are creating a new thinga new drug, a new enzyme, a new material.

That is true in theory, but in practice your entire day is spent creating data about a potential new thing, not the thing itself. Why does this matter? Because it's not good enough to make that thing just once. You have to create the designs, methodology, and specifications that allow anyone else to make that thing over and over. That means collecting vast amounts of high-quality data, as if you are manufacturing data itself. Many synthetic biology companies (Zymergen and Ginkgo Bioworks come to mind) have deeply internalized this idea, building their entire R&D architecture around the automated collection and integration of data for machine learning.

When you treat R&D as if you are manufacturing data itself, you get all the good behaviors that come with it: quality, consistency, efficiency, reusability, reliability, continuous improvement, and trust. This would be a welcome antidote to the present troubles of irreproducible scientific research and high failure rates in drug discovery.

Sydney Brenner said this in 2004 in reaction to the explosion of genomics data, most of it of low quality. The lesson absolutely applies today. But what is often misunderstood is that CAP data doesn't start with information technology. It starts by reengineering the processes that generate it. That means a fundamental transformation of how we apply the scientific methodone that shifts from the observational roots of yesterday, where the lone experimenter captures notes in a notebook, to a full-on adoption of an industrialized approach to science. This approach recognizes that science is a set of processes that can be designed, executed, and improved just like any other endeavor of human creativity. Tim describes this is greater detail on Riffyns blog.

The five lessons above add-up to a new pathway to scientific discoverya process-driven, data intensive, quality-oriented, industrialized kind of scientific undertaking. The shift is already underway, and it was driven in part by the nave spirit of creativity and innovation of synthetic biology. When we make this shift to an industrialized science, it will deliver leaps in scientific output akin to the industrial revolution of England in the 1700s. And with those leaps, the 90% failure rates in drug discovery will become a thing of the past. Ultimately, this shift in how we do science may be the greatest contribution of synthetic biology to our society of all.

I have previously written how big pharma has been slow to innovate and adopt the latest synthetic biology tools, which could vastly speed the creation of new treatments and vaccines. At pharma networking events Ive been to, most everyone is secretive and nobody shares what theyre really working on.

I had recently started changing my mind about pharma, with companies like Codexis becoming an engineering powerhouse to the pharma industry. These days, if you can imagine a drug, someone can make it.

And COVID has changed my thoughts about pharma even more. The old barriers have dissolved a little. People and companies are collaborating. I recently interviewed Moderna CEO Stephane Bancel and he agrees: people are collaborating really fast and well right now because they have a common enemy: COVID-19.

One of my colleagues, Molecular Assemblies CEO Mike Kamdar, recently summed it up this way: When youre in the pharma world, there is a lot of secrecy. But here, while we compete, there is a sense of collegiality. Thats just how the synthetic biology industry is.

The story of how the pharma industry responds to COVIDand synthetic biologys role in that responseisnt done being written. Lets hope its the first chapter of a bright new future where the tools and technologies available to us are used to bring people medicines better, faster, and cheaper.

Follow me on twitter at @johncumbers and @synbiobeta. Subscribe to my weekly newsletters in synthetic biology Thank you to Tim Gardner inspiring this article and sharing many of the ideas in it. Thank you to Kevin Costa for additional research and reporting in this article. Im the founder of SynBioBeta, and some of the companies that I write aboutincluding Riffyn, Ginkgo Bioworks, and Codexisare sponsors of the SynBioBeta conference and weekly digest heres the full list of SynBioBeta sponsors.

Full coverage and live updates on the Coronavirus

Read the rest here:

Five Things Big PharmaAnd Its InvestorsCould Learn From Synthetic Biology - Forbes