August 12, 2019

By Michelle Pelletier Marshall, GAI Media

Imagine eliminating the wonder of knowing if your agricultural product is fresh before examining it, or having advance warning that your livestock may soon be sick, simply by the scent it gives off? Exclusive olfactory biosensor technology from Aromyx now enables these dreams to be a reality, making the science of olfaction measurable, actionable, and easy-to use to achieve business, product, and technical goals.

Founded in 2013, Mountain View, California-based Aromyx is a privately-owned startup whose team includes eight Ph.Ds with unique knowledge about using synthetic DNA and advanced robotics in building sensor systems based on olfactory receptors. Leading the charge as CEO is biotechnology entrepreneur Dr. Josh Silverman.

Dr. Silverman has decades of experience in new technology research, development, and commercialization, and has held positions at public biotech companies including Maxygen, Inc. and Amgen, Inc. He has raised over $300 million in venture funding in his career to date. He also is a scientist by trade – with a B.Sc. in Molecular Biology from UC San Diego and a Ph.D. in Biochemistry from Stanford University – and is a recognized author in scientific journals with over 90 approved and pending patent applications.

GAI News reached out to Dr. Silverman to get more info on the latest in Aromyx’s smell and taste technology achievements.

1.) Aromyx’s proprietary method is set to standardize and visualize the entire olfactory pattern of any given odorant substance. Please provide an overview of how the system works.

Our core platform is a biotechnology platform. So when we talk about taste and smell, the best example is a wine review. Each person examining a wine would come up with about 10 different adjectives about how they think the wine tastes or smells, but each person would come up with a completely different description. The problem is that the readout of our taste and smell is completely subjective.

What we are doing at Aromyx is focusing on the multiple receptors present in the nose and on the tongue for each of the odorants and measuring the output specifically. Similar to how any given color can be specified on the Red Green Blue (RGB) spectrum, we measure the biochemicals present during the act of tasting and smelling, and then represent these odorants with a number and calculation.

Right now the system of classifying an odor or taste as “good” or “bad” is very much trial and error, based on human input panels, where a “bad” sends the product back to the drawing board. The benefit with our system is that we can precisely describe, and therefore create, any given odor or flavor because we are able to assign a quantitative number to it and draw vectors in odor spaces that lead us from point A (the “bad” smell/odor) to point B (the “good” smell/odor). We have brought science into the equation to allow for rational design and optimization of flavors.

 2.) There are several industries where your technology has proven useful – with fermentation for pharmaceuticals, for fraud protection in fragrances, and for early detection in some cancers – what are the top applications for food and ag?

The applications are very broad. We use taste and smell everyday without even thinking about it; taking it for granted. The olfactory bulb is tied into one of the oldest parts of our brain, in that core memory/core emotion, good/bad feeling area. There is such a complex input that it is hard to make rational decisions, which is where the subjectivity comes in.

For something like fermentation – which is a very complex process where you are growing microorganisms in a complex environment with its own metabolism that secretes certain chemicals, and is used in pharmaceuticals, wine and beer, and even sourdough bread – we can help precisely identify the stages by assigning a number, rather than having to rely on the person who sniffs or smells the batch and can only use his/her words to describe the stage it’s at.

We have one person’s sense of smell and taste deciding on tens of thousands of liters of wine, which could be millions of dollars in revenue, and it’s hard to measure and recreate that product.

For food and ag, there are 400 receptors in the human nose and we can detect a wide range of different chemicals and inputs, but for most applications you don’t need all 400. For ag customers, things like degradation and spoilage are the easiest ones in which to apply our technology.

3.) What does the future hold for Agromyx?

Product development is absolutely where we want to go, where we can start to design juices with different flavor profiles that consumers have never tried before. We can measure flavor profiles that no one is producing now and give companies the recipe to make those. This design opportunity would unlock a whole lot of value for them.

We also expect to be able to identify the receptor preferences for a person of Chinese background versus a person of African background versus those of European descent so we can determine what each group would prefer in a beverage and start to design products and flavor profiles specifically for certain demographics, descent, or genetics. That is a level of detail that has been very difficult to identify as products go international.

For the ag future, we see the sensor helping agriculture in assessing spoilage, food safety, detecting the presence of mold, or grading the quality of produce and optimizing flavors (instead of one person grading a whole truckload based on one sample of the fruit). Companies are very interested in the “white space” between what varieties they provide and what their competitors do to identify a new, better-tasting product.

We also want to create a marketplace where people can actually buy and sell raw materials based on taste and flavor. Our goal is to have systems for use in the field because it doesn’t require electronics, it’s a measurement that is above or below a threshold for a yes or no answer, much like how a pregnancy test works. This requires zero electronics, zero training, and doesn’t even require literacy, so this sensor can be provided at farms in the developing world and people can perform onsite tests to see if something is safe or not, spoiled or not.

We are designing products that can be used at point of sale, point of transfer, point of delivery, and other off-site uses.

4.) Aromyx technology offers advantages over comparable technologies from a cost, time saving and sensitivity perspective. Can you provide examples of this?

We just worked with a large beverage company on a fruit juice. Their only way to tell if their new batch of juice was good or bad to use in their drink was to bring a group of people into a room and have them taste it and grade it. This process is very noisy, subjective, and imprecise, and the company is making significant buying decisions based upon the results.

So what they did was bring us a “good” juice and three “degraded” juices and we examined this in our lab against our library of receptors. We identified the human receptors that could differentiate the good juice from the bad one, and identified the ones that indicated different degradations. Now we can build them a sensor that can detect the degradations, and they can quickly, easily, and cheaply test each batch of juice and get a quantitative number that indicates a good or bad juice, and the degradation in each, without the subjectivity of relying on the human group perceptions. This helps them put a numeric value on if a batch is good or bad. In future batches, they can use the numbers to test current juice, or farther down the road, use the numerical grade to create new juice products.

5.) Is there any comparable technology to yours, or main competitors?

Our main competitors right now are people – the 10-person tasting panel. And as imperfect as that is, it clearly works. Food companies make money. They are able to get a reasonably reproducible product out into the market, and people buy it. Even though we think there is a much better way of doing it, these companies have built the systems and they work. They may not work optimally, but they work. These are what we are trying to displace and bring more standardization and science into that process.

And there are no numbers to compare our method to the 10-person panel, as that is so subjective. It depends on who you talk to. Some companies say these human panels are very expensive and they want to move away from them, but others say it’s not too difficult to get info from this group. Our view is the quality of the data is very different. The idea of 10 people in a room creates so much noise and one company can’t compare their data to another’s. For the companies who are doing human panels very rigorously, they see a big cost and time advantage with our technology. For the others, it’s less a cost savings, but a way to get better, quantitative data so they can make better decisions.

We don’t want to be a sensor company. We want to be selling the data insight and the value in that. Helping with product design, reducing inefficiencies in the supply chain, etc., and more has much more value, and that’s where we are heading. 

6.) Aromyx closed on a $3 million Series Seed equity financing round just last month, with the round led by Ulu Ventures with participation from Rationalwave Capital Partners, Merus Capital, CE Venture Capital, Stanford University, and Radicle Growth. What will this funding be used for, and what’s on the horizon for future investment opportunities?

Right now, we have a lot of customers, though we can’t publicly discuss them, and we have been pleasantly surprised at how strong the interest has been since we brought this technology forward. Because of this interest, we expect to raise more money towards the end of this year to expand our capacity, raise our throughput, and address more of the customers who want to use our technology.

The current round of money is going towards expanding our science team. We also recently moved to a new location and expanded our lab, and have been able to engage with new customers and do more pilot projects. As I noted we are currently “oversubscribed”, so we expect that the next funding round will be much larger, upwards of $10 million.

ABOUT JOSH SILVERMAN

Dr. Josh Silverman has 20 years of experience in new biotechnology research, development, and commercialization. Before joining Aromyx, he was founder and chief technology officer of Calysta, Inc., where he led development of technology converting methane gas to protein using bacteria. Previously, he was vice president of Drug Development at Amunix, Inc. Dr. Silverman co-founded Versartis, Inc. in 2008 and participated in the founding and financing of four startup biotechnology companies (Avidia, Diartis, Siluria, and Calysta). He has held positions at two public biotechnology companies, Maxygen, Inc. and Amgen, Inc.

Dr. Silverman earned his B.Sc. in Molecular Biology from UC San Diego and his Ph.D. in Biochemistry from Stanford University.

-Michelle Pelletier Marshall is managing editor for Global AgInvesting’s quarterly GAI Gazette magazine and a regular contributor to GAI News. She can be reached at mmarshall@globalaginvesting.com.