The Economist writes that in 1944, Mexico imported huge volumes of wheat: about 60% of the wheat consumed in the country was imported. In order to ensure the country an adequate supply of the staple product, the government decided to take advantage of financial support from the Rockefeller Foundation and create the Cooperative Wheat Research and Production Program, led by Norman Borlaug.
Borlaug's breeding programs and his team lasted for 20 years and included improving the disease resistance of Mexican wheat, as well as increasing yields by breeding shorter plants that were resistant to lodging when heavily fertilized. By 1963, around 95% of the wheat sown in Mexico was a result of their work, yields increased sixfold, and this experience was adopted by India and Pakistan to feed their rapidly growing populations. For his achievements, Borlaug was awarded the Nobel Peace Prize in 1970.
The selection of agricultural crops is a complex task. Geneticists must determine which traits to investigate, cross plants that seem to possess these traits, conduct field trials, and await results to find out if their new plants have improved. The interaction between plant genes and environmental factors such as weather, soil conditions, and other ecological conditions in which they grow is complex. Determining which genetics are suitable for specific conditions can take decades, as was the case with Borlaug in Mexico.
Newly formed company Heritable Agriculture, spun off from X lab of Google LLC, is developing an approach that uses artificial intelligence (AI) to predict genetic improvements in yield, taste, nutritional value, and photosynthetic capacity in specific environmental conditions. Their software has been trained based on data collected by the company's employees over the past six years.
These data describe the behavior of different gene combinations of plants in specific soil and weather conditions, gene expression, and the concentration of various metabolites during plant growth. Scientists processed data from approximately 14,000 samples taken from field trials conducted in Nebraska, Wisconsin, and California with seven different crops. After identifying the desired genetics for a specific environment, another model determines the shortest path of selection to achieve this genetics based on available plants.
Brad Zampft, one of the co-founders of Heritable, reports that the company's system can create a crop with the desired genetics to obtain the desired trait in just one year. He presented data on the success of Heritable's approach at the Plant and Animal Genomes Conference in San Diego on January 13. It was shown how the company's software can be used to quickly obtain corn with precise timing control for flowering. Heritable has already used its software to create plants with specific properties for clients (without disclosing information about them), including tastier leafy green vegetables. Zampft notes that "traditional selection of agricultural crops is too slow and expensive to achieve the great results that we synthetic biologists can achieve: nitrogen fixation, sustainable forestry, food as medicine, carbon capture."
Other biotechnology companies, including Inari from Massachusetts and Pivot from California, are also focused on improving yields through genetic editing and soil microbiome enhancement, respectively. Over the past ten years, agricultural biotechnology companies have attracted approximately $40 billion in venture capital.
Heritable's advantage over other companies lies in their ability to apply modern computational methods to accelerate the selection process, allowing them to search and explore more possible genetic combinations by reducing research costs that were previously unattainable.
Heritable also focuses on less industrial crops such as berries, green vegetables, some grains, and legumes. Their priority is to achieve success in breeding oats, barley, rye, chickpeas, bok choy, avocados, and grapes to introduce improvements similar to those achieved in corn over the past 100 years.
If selection becomes cheaper, it will be possible to adapt a greater number of plants to different environments. For example, farmers in developing countries with limited resources will be able to grow and use plants specifically created for their needs. This is particularly useful considering that staple crops are facing pressure from rapidly changing climatic conditions, too fast for traditional selection.
Heritable also plans to apply its computational capabilities to tree breeding and thus to forest management. By breeding local trees so that their wood is competitive compared to average timber, biodiversity can be increased. The company is already collaborating with ArborGen, a sapling supplier, to improve pine.
Heritable is part of Alphabet's strategy to develop companies created within its X division. Some of these companies have been acquired by Google, while others have become independent, but continue to operate under Alphabet, such as Waymo, a maker of self-driving cars. Companies like Heritable receive support from the X division, which incubates these companies and attracts venture investments from external sources, enabling them to operate independently. Heritable and their successes in "agrogueling" would have been valued by Norman Borlaug.