You are here
A Venture Capitalist Looks at Genomics in Agriculture
While interest in genetics is on the rise in various industries, many agriculture industry pundits, entrepreneurs, and investors wrongfully assume genomics isn’t investible in agriculture. Sadly, only a handful of genomics companies (e.g., Hi Fidelity Genetics, a Finistere Ventures portfolio company; NRGene; RAPiD Genomics; and Spiral Genetics) are actively pursuing innovations in agriculture. This is a monumental mistake — a mistake that could prevent countless brilliant geneticists and technology innovators from exploring the emerging opportunities in agriculture.
The Evolution of Agriculture and the Evolution of Genetics
Agricultural genomics has the potential to feed the world. The opportunity cannot be ignored. As a geneticist and agriculture investor, I think it is vital to dispel this myth. In addition, it is important to fully understand current agriculture challenges and practical considerations when using genomics to solve these problems.
So why is the evolution of agriculture tied to the evolution of genetics? The power of genetics depends upon the ability to direct sexual exchange, and that is much easier done in crops than in animals or humans. From Mendel’s peas to population-level analysis of genomes, each new genetics breakthrough first achieved commercial deployment in agriculture. Recombinant DNA technology (molecular mapping in breeding and genetically modified crops), RNAi (first described in plants) and used for trait discovery and new chemicals, gene editing, and genomics (next-generation breeding named genomic selection) … the adoption of genetic technologies in agriculture has been remarkable.
The opportunity for genomics in agriculture is boundless. Agriculture is the only commercial market where every genetic advancement has been directly or indirectly commercialized in a global, multibillion dollar market. Moreover, it is the only market where genomes are actively bought and sold in mass scale. In fact, agriculture is currently the only market selling trillions of complex genomes annually. Not sure where? All you have to do is take a look at the global crop seed market.
The Value of Genomics in Agriculture
Based on Finistere Ventures research, total genome sales have now topped $50 billion per annum, a figure that would be much higher if it also accounted for breeding animals. However, the true value is based on productivity. For example, the U.S. corn market has surged thanks to breeding advances like hybrid corn in the 1940s. The best genomes now represent the largest input cost to growers and up to 80% of the gross margin products of the technology companies that produce them.
However, even with all of the advances to date, much more innovation is needed. Why? Agriculture is still fraught with volatility, which is exacerbated by temperature and water fluctuations. More research into breeding plants to withstand climate change is essential as genetic selection can mitigate drought and other environmental challenges. In addition to climate volatility, increasing productivity and yield is essential to feed the world’s growing population and meet rising protein needs. However, yields now need to increase at double the yield improvement rate, about 2.4% to 3% per year, in order to meet needs in 2050. Unfortunately, yield improvement has been decaying for decades and is now close to zero.
Putting Predictive Genomics Into Practice
Consumer demand is growing, and current technology approaches are not solving the yield pain points. Most breeding currently relies on an outdated, analog approach that limits potential value. However, there are new tools that can re-energize the paradigm of breeding.
Thanks in large part to rapid decreases in computing power cost and DNA sequencing, predictive approaches to genomics are emerging. A quiet revolution called genomic selection (GS) uses sequence data to select genomes. Predictive and relying on machine learning and advanced data science techniques, GS can optimize each contributor to yield (even those that were previously considered insignificant and ignored) within a population. GS has the potential to determine the best code at each position in the entire genome and then recapitulate it in the field. This approach could completely disrupt the agriculture game.
However, to fully optimize the value driven by the next wave of predictive genomics in agriculture, two core issues need to be answered within the breeding process flow.
1. Predicting Hybrid Vigor: The Breeding Bottleneck
A second need underlying predictive breeding is the ability to understand, or at least predict hybrid vigor or heterosis. It results when two inbred parents are crossed, and the resulting hybrid offspring is more vigorous than either of the parents. This has a clear economic value as these plants tend to be more robust and also deliver higher yields.
While hybrid vigor exists in many crop species, it is most notable in corn. In fact, Pioneer Hi-Bred was built on exploiting this phenomenon. However common, hybrid vigor still creates significant roadblocks. Currently, it is impossible to predict which parents will give the best hybrid vigor response. Consequently, breeders do multiple experimental crosses between potential parents to maximize the hybrid vigor outcome before settling on that season’s commercial hybrid. This is both time-consuming and expensive – in some cases, up to a third of breeding cost is spent in testing hybrid combinations and running multiseason trials.
Also embedded within this approach is the reticence to breed aggressively in both parents for fear of losing hold on the combinability within the pedigree. Ultimately, this exacerbates the length of time needed to develop new products that combine an optimized set of new traits. This means that to really change the dynamic of breeding, at least in some crops, the breeder needs excellent data and the ability to predict hybrid vigor. While this has been recognized for decades as a holy grail in breeding, a commercial approach to solve this challenge has yet to be developed.
2. Data in = Data out
Genomics and prediction is a game of statistical power to resolve the signal from the noise. However, the ability to do this is reliant on the data upstream of the predictive algorithm. Consequently, as the world moves rapidly toward genomic breeding technologies, a huge opportunity for creating novel and accurate data sets exists. To realize this, objective measurements and new methods of imaging and measuring plant characteristics will be needed. When these elements are in place, the industry will move away from the traditional 0-9 breeder scales.
To achieve this mind shift, new sensors, robotics, and imaging platforms will need to be deployed in field throughout a growing season. Consequently, these platforms will need to be robust, scalable, and affordable. Entrepreneurs have started to recognize these specific AgTech adoption challenges and have begun creating solutions. Venture-backed companies such as Blue River Technologies are building imaging platforms and sensor suites that can measure plant attributes in real time with great accuracy. Similarly, traditional lab- and greenhouse-based platforms, such as Lemnatec, are moving these systems in different formats to the field. Lemnatec is currently building out what it calls the Scanalyzer Field, which relies on a stationary robotic arm to analyze a fixed field area multiple times a day. Others such as Rowbot are focused on autonomous vehicles that can drive between rows of corn while imaging growing corn plants.
As Silicon Valley continues to transfer technology and skills, especially in imagery, to the AgTech space, high-throughput analysis of plant height, nitrogen levels, water stress, leaf angle, stem thickness, etc., will be readily accessible to the genomic breeder. However, despite these exciting advances, one large whitespace exists: analyzing the rhizosphere in field. Since breeding has largely occurred by observation, roots to a large extent have been out of sight and out of mind. This is changing with the realization that optimized roots present the most effective solution to an increasingly volatile climate. Crop germination, nutrient acquisition, and water utilization all depend on root systems. Consequently, optimizing them will yield significant gains for growers. Indeed, Pioneer’s AquaMax has shown how valuable root growth traits are in corn. To harness this power, the next-generation breeder will need to have tools to objectively measure biology below the ground in much the same way as it is measured above the ground. This new territory is a massive opportunity for plant analysis companies.
Myth Status – Busted
It is time to reinvigorate the focus on agriculture. Genomics companies and innovators alike need to understand the huge opportunities that exist for genetics in agriculture. It is also important to note that ancillary technology companies that can generate helpful datasets and insights by building predictive platforms and new devices will underpin this market. With enormous, high-margin, global markets on the table and an easier regulatory environment than many other industries, now is the time to band together and bust the genomics in agriculture myth.
This article is written by Spencer Maughan, Finistere Ventures, Palo Alto, California.