AI boosts crop improvement research

What is the role of artificial intelligence in crop improvement? Questions about artificial intelligence are becoming increasingly pressing in all fields. According to Jianming Yu, one of the world's leading scientists in the fields of quantitative genetics and plant breeding, AI provides a new lens to bridge science and practice in crop improvement.

“People have a lot of questions about how to start actively using AI in crop improvement,” said Yu, Pioneer Special Chair in Corn Breeding and Director of the Raymond Institute. But knowing how to best leverage the tools of AI is not easy.” His F. Baker Plant Breeding Center at the Iowa State University College of Agriculture. “There are many concrete examples of how these tools can be used constructively, but it hasn’t happened yet at scale.”

Yu's mission is to help his colleagues, students, and the general public learn more about the rapidly evolving field of AI. To this end, he and other co-authors, including Karlene Negus, a genetics doctoral student with whom he collaborates, have published an overview of the role of artificial intelligence in crop improvement in an academic compilation. Announced. advances in agriculture.

“Many scientists, even those with relevant backgrounds, don't always know where to start,” Yu says. “We have received feedback that the new paper is very timely and useful.”

Recently, the Iowa State College of Agriculture and Life Sciences asked Yu and Negus to review highlights of new publications and reflect on the use and impact of AI tools in their field.

Yu: One of the things we do in this paper is briefly sketch the historical background of AI. It has been in development since the 1940s, and what could be considered the third Summer of AI is underway. Deep learning systems defined the beginning of this era.

In crop improvement, AI has primarily been deployed to help process and understand very large, high-throughput datasets. Large-scale data poses new challenges in agricultural research and many other scientific fields, and AI tools are already providing a variety of solutions.

Mr. Negus: The field of AI has been changing rapidly in recent years. It can be difficult to know which methods are relevant for a particular use. To streamline the learning process in fields related to crop improvement, we describe more than 15 types and subtypes of AI and provide insight into how AI is being used in these fields. Although these methods are not exhaustive, we believe they are a good introduction to what currently exists and what constitutes tools that we expect to develop in the near future.

While today's newsworthy AI is most often highly sophisticated neural networks, other AI examples are repetitive and highly variable with sufficient variability to prevent the use of standard process automation. This extends to relatively simple robotic process automation, which uses AI “agents” that can run processes. , to relatively complex expert systems, fuzzy systems, and other types of advanced machine learning that seek to replicate the problem-solving abilities of human experts.

Machine learning (ML) is a type of AI that uses large datasets to improve through experience and learning, and then uses the results to solve problems and make predictions. ML is widely used in the field of crop improvement. ML techniques using genomic, environmental, phenomic, and other multi-omic approaches can help researchers capture environmental and genetic variation and better understand its impact on crop breeding and management.

Yu: Together, these applications are rapidly revolutionizing agricultural practices in the laboratory, greenhouse, and field.

When crop improvement researchers implement AI methods, it is desirable to know the potential benefits of AI methods over traditional methods. For breeders, the improved ability to monitor and predict crop growth and health under different combinations of genetics, environment, and management can greatly facilitate crop selection decisions. Producers will want to leverage AI to improve on-farm production management to improve sustainability and resilience.

For those involved in crop improvement, catching up is a familiar challenge. For the past century, that challenge has centered on meeting the demands of a growing world population, and this continues to be a major concern. Now, the changing climate is making the task even more complex. AI has great potential to solve these challenges, but there is much work to do to fully exploit this potential, and training and skills in these areas will need to grow rapidly.

Still, if the successes achieved so far by leveraging innovative technologies for crop improvement are any indication, the future of AI-assisted crop improvement is bright.