Interview of Synthetic Biologist Andrew Hessel by Sander Olson

Here is the Andrew Hessel interview. Mr. Hessel is a synthetic biologist who is the founder of the Pink Army Cooperative, which aims to leverage synthetic biology to create new individualized treatments for breast cancer. He is also a faculty chair at Singularity University. Mr. Hessel advocates using an open source approach for creating DNA code.

Question: What is the pink army cooperative?
Answer: The Pink army initiative launched in 2009, with the purpose of creating a new business model for drug development — one that allows personalized drugs to be created safely, quickly, and affordably, perhaps even free. The main goal of the initiative is to expedite drug development by utilizing the new field of synthetic biology. Synthetic biology will largely remove the economic and technical barriers that substantially delay drug development, through individual personalization of medicines, which reduces the complexity of clinical trials.

Question: How exactly will these barriers be removed?
Answer: Drug development is currently hobbled by regulatory barriers. The clinical trials generally take a decade or more to commercialize. Moreover, the drugs are developed using a general purpose, one size fits all approach. By using synthetic biology we can tailor specific treatments for specific cancers and greatly reduce both the time and the cost of drug development.

Question: What is the main purpose of the singularity university?
Answer: I am the co-chair of bioninformatics and biotechnology at the Singularity University. Singularity has a very unique business model when it comes to education. We focus on training students in technologies that can go exponential. I emphasize to the students that the reading and writing of genetics code is an exponential technology in the same way for biogenetics that Moore’s law is for computing. We are hitting the knee of the curve of that technology.

Question: Some are worrying about Moore’s law dying in the next five years. If it does, can the biotech revolution still happen?
Answer: People have been talking about Moore’s law dying for decades – I will believe it when I see it. I don’t see it ending anytime soon. In biotech, we are actually generating genetic code at a rate that is several times faster than Moore’s law. Eventually, just storage of the data will be a challenge given the real costs of our present computer architectures. We’re still waiting for the right application or output that connects genetic information to the mass market and generates scaleable revenues.

Question: Craig Venter’s team recently created a new prokaryotic life form. How important is this accomplishment?
Answer: I believe Venter’s accomplishment will have profound ramifications. I see DNA as a programming language, and like any language there are three components – reading, writing, and comprehension. Craig has been at the forefront of reading and understanding DNA code, and today he is a leader in synthetic technologies. The creation of an artificial prokaryotic life form only scratches the surface of the greater potential of synthetic biology. The ability to easily engineer living organisms is perhaps the most powerful technology humans have made to date.

Question: Robots are starting to emerge in sequencing labs. To what extent can this field be roboticized?
Answer: Modern sequencing labs are highly automated. Synthetic biology has at this point been only minimally roboticized. The software tools for metabolic engineering or designing synthetic genomes are still crude. There is still considerable room for improvement and automation. Genome design, synthesis, and testing will becoming increasingly computer-assisted and automated over time, as has other areas of engineering.

Question: Will the emphasis in coming years shift from genomics to proteomics?
Answer: Proteomics explore the structural and functional materials that make up living cells. However, it’s important to understand that all the “omics” fields, including proteomics, metabolomics, or epigenomics, are encapsulated the genetic code itself. So every omic field will in the end serve to supplement and enhance the field of genomics.

Question: You have argued that a tipping point will occur when 10 million base pairs can be sequenced for $100. What will happen when that threshold is crossed?
Answer: The reading of DNA is four or five orders of magnitude less expensive than writing DNA, which is presently hovering at about 50 cents per base. The cost of synthesizing 10 million base pairs should fall to $100 within five years of the synthesis being made technical priority by the scientific community. Without a concerted push or incentives, this price point will just take a little longer, perhaps a decade. I believe it’s worth getting the cost of synthesis down sooner than later, as it will speed the development of new applications, many worth billions of dollars.

Question: How long until genome sequencing becomes available on an iphone?
Answer: The new Ion Torrent sequencer uses semiconductor chips to read DNA code and costs $50,000. That is a significant drop from the previous generation machines, which cost about $300,000, and were more complicated. It already has an iPod dock to upload data. This technology will definitely shrink as it is refined. As DNA sequencing moves to systems-on-a-chip, or incorporates fourth-generation technology based on nanopores, the sequencer could shrink to the size of a SIM card. With sufficient demand for sequencing DNA, and doing it in real-time, this miniaturization will happen fairly quickly.

Question: What will be the first mainstream application to be introduced that is dependent on synthetic biology?
Answer: That is the billion dollar question. If I had an answer to this question, I would be locked in a lab developing it. There are some clues, though. Historically, biotech has focused on treating illnesses, but the average consumer isn’t sick. This limits the marketplace. In health, people spend money on things that gives them tangible value in their everyday lives, at affordable prices. This means energy, building materials, household products, cosmetics, foods, pets, sensor and diagnostic technologies, and perhaps even smart drugs and intoxicants, like beer or wine. Whoever successfully brings biotechnology innovation to the masses will generate a fortune that rivals Google.

Question: When will the first human organs be created using synthetic biology?
Answer: Human organ cloning is actually more about stem cell engineering than synthetic biology per se. I predict rapid advancements in that area, due to rapid 3D printing technologies. The leader in that space is a company called Organovo, which has just announced the first commercial cell printer. Once we can print cell-based structures, we can produce everything from synthetic foods to organs. Given an aging population, it’s only a matter of time before fully functional cloned human organs become commercially available.

Question: How much progress can be expected in the field of synthetic biology by 2020?
Answer: This technology is in the knee of the s-curve and will grow exponentially for decades. By 2020 we will be able to engineer simple living systems routinely, whether it be a single protein, a metabolic pathway, or simple multicellular creatures. Eventually, testing and measurement of what we’re programming will become the limiting factor. Overall, the potential of synthetic biology is comparable to the potential of computers. We’re going to see it broadly applied in human endeavors.

If you liked this article, please give it a quick review on Reddit, or StumbleUpon. Thanks