Building Bridges for Better Breeding
Since Gregor Mendel’s meticulous 19th-century pea plant experiments, plant and livestock breeders have relied heavily on data. Today, they have more tools and technology than ever for collecting, analyzing and using the growing amounts of data they need to make their work less hit or miss.
For Amanda Hulse-Kemp, helping breeding programs integrate the best of these technologies to speed their work and make it more efficient is not just a job, it’s a passion.
Since 2016, she’s served as a computational biologist with the Genomics and Bioinformatics Research Unit of the USDA’s Agricultural Research Service, or USDA-ARS.
Based at North Carolina State University, Hulse-Kemp is affiliated with the Department of Crop and Soil Sciences and the N.C. Plant Sciences Initiative and frequently collaborates with university scientists and engineers working to make agriculture more economically and environmentally sustainable.
Our ultimate goal is to get better traits faster and to get those plants into the growers’ fields faster … and it’s not just plants, it’s also fish and other animals.
“It’s a great partnership between NC State and USDA-ARS,” Hulse-Kemp says. “We start off with basic research at NC State and readily move on to the applied research through ARS, and I get to see much of the work translated to the field, which is very rewarding.
“Our ultimate goal is to get better traits faster and to get those plants into the growers’ fields faster, so everybody benefits. And it’s not just plants, it’s also fish and other animals. We work in whatever system is relevant and ag based.”
A Record of Award-Winning Progress
A string of awards suggest that Hulse-Kemp has made significant progress toward that goal since earning her Ph.D. in genetics from Texas A&M University in 2015. While a student, she was involved in sequencing the first upland cotton genome and in developing a device capable of analyzing the genetic makeup of many cotton samples at once.
In 2021, she received the USDA-ARS Herbert L. Rothbart Early Career Research Scientist award for using high-throughput techniques to rapidly advance breeding programs not just for cotton but also for spinach, tomatoes, coffee, peppers and other crops. Using these techniques instead of earlier methods, researchers could get usable genetic data in three days, as opposed to the weeks or months.
Hulse-Kemp went on to win the National Association for Plant Breeding’s 2023 Early Career Scientist Award, and in January, she was one of 400 scientists nationwide to receive the Presidential Early Career Award for Scientists and Engineers.
She is also part of the Breeding Insight and BI OnRamp programs that recently won the USDA Secretary Honor Award in the category of providing all Americans safe, nutritious food.
Integrating Tech Tools for Better Breeding Outcomes
As the director of BI OnRamp, Hulse-Kemp helps commodity-specific USDA-ARS breeding programs incorporate bioinformaticians into their teams to use advanced technologies for collecting and analyzing two types of data: phenotypic data, which relate to an organism’s observable traits, and genomic data, which provide information about the genetic makeup that influences those traits.
The goal, Hulse-Kemp explains, is to help the teams gain insights that enable them to better predict which breeding decisions will get them to their goals – whether that’s increased yields, better quality or pest resistance – faster and more efficiently.
“The bioinformaticians, or people like me, assess the questions that the breeding groups are working on, the status of technology they currently have available to them and the repertoire of technologies that we have both on the genomic side as well as on the phenotyping side to help them. Then we develop a plan and move forward with some of these technologies,” she says.
“The other essential aspect of the OnRamp program has been migrating a lot of the captured data that the team already has into a digital ecosystem or database in an organized fashion,” she adds. “Some of these breeding programs are over a century old, so that can be a complicated process.”
Addressing the Need for More Phenotyping Data
Going forward, breeders need better tools to gather and analyze data from their field trials, Hulse-Kemp says.
“In order to do predictive breeding, we have to have both the genomics and the phenomics,” she says. “I feel like we’re in a known space on the genetic side of things: We’re able to develop tools, markers and genomes relatively easily now if you have the money to do it. But we really need help on the phenomic side.”
To meet that need, Hulse-Kemp is working Steven Mirsky, of USDA-ARS, and Chris Reberg-Horton, of NC State’s Department of Crop and Soil Sciences and the N.C. Plant Sciences initiative, on a project called DASH — short for Digital Agricultural Systems Hub.
The project’s goal is to accelerate the use of artificial intelligence and machine learning to develop and deploy phenotyping tools in research plots and farm fields.
We’re in a known space on the genetic side of things. … But we really need help on the phenomic side.
Hulse-Kemp says that the data-based insights gained through DASH will not only aid breeders, they’ll also help scientists and farmers as they strive to address myriad agricultural production challenges.
When thinking about DASH and its potential for moving plant breeding and agricultural production forward, Hulse-Kemp recalls her days at Texas A&M working on cotton genetics and genomics.
In the lab and on her computer, she was using and helping develop leading-edge genomic tools, but her field world was not so high-tech.
One of her tasks involved collecting and counting cotton bolls to be tested for fiber quality.
“I struggled continually counting to 25, and if I have to count to a hundred or some high number, I know that I will not be a very accurate counter,” she says.
Finding ways to use technology to overcome such human limitations could yield multiple benefits, Hulse-Kemp says. More precise data could lead to more robust research and better farm management decisions, and freed from the time-consuming, monotonous work of data gathering, scientists and farmers would have more time to focus on the things they’re best at.
“In the end,” she adds, “everyone wins — scientists, producers and consumers.”
This post was originally published in Plant Sciences Initiative.
