A New IDEA: Pairing Trainees with Computational Experts During the Pandemic

The COVID-19 pandemic has presented researchers with a host of new challenges. For researchers in NC State’s College of Agriculture and Life Sciences, one way of overcoming those challenges was to expand their research into the computational arena.

This April, while research in the lab and the field was largely paused in order to reduce the spread of COVID-19, the CALS Research Committee (CRC) and North Carolina Agricultural Research Service (NCARS) sent out a survey to every department in the college asking for trainees who had data they needed additional expertise to analyze or model computationally. They called this project Integrating Data with Exploration and Analysis: Taking the Research Enterprise to the Nth Dimension, or IDEA-TREND.

Seven trainees filled out a survey and wrote a one-page project proposal outlining their project and how they would like to collaborate with a computational expert to advance their research.

Three CRC members, Michael Reiskind, Benjamin Reading and Ross Sozzani worked to connect these trainees with other researchers across the university — or even within their own department — who had the needed expertise in bioinformatics, computational modeling, machine learning or analyzing big datasets.

While a significant amount of research has resumed on campus and at university field labs and research stations, the CRC plans to send out another call for IDEA-TREND proposals soon. The CRC recognizes the importance of these connections and the value of computational research brings to many experimental biology projects.

“The future of science is collaborative, and both empirical scientists and theoretical or quantitative scientists will need each other,” Reiskind said. “The more we get out of our silos, especially as students and post-doctoral scientists, the better our science becomes. Finding matches has been very rewarding, and hopefully some superior science will come out of it, for both sides of each collaboration.”

Here’s how the collaboration is going for three trainees.

Ben Graham and Abhirav Kariya

Ben Graham is a postdoctoral research scholar in the Department of Crop and Soil Sciences. He works in the lab of Candace Haigler.

He primarily studies the development of cotton fibers at the cellular level. He wanted to work with a computer vision expert to try to speed up the process of getting shape data from microscopic images of growing cotton fiber cells.

A black and white microscopic image of a bunch of finger-like cotton fibers.
A microscopic image of young cotton fiber cells. Ben Graham is working with a computer vision expert to develop an algorithm that can recognize the important features of the cotton fibers and gather the necessary shape information from thousands of images. Image courtesy of Ben Graham.

Sozzani connected him with Abhirav Kariya, a graduate student in Edgar Lobaton’s lab in the College of Engineering.

“Right now, we’re working towards a way to automate gathering geometric data from growing fiber cells,” Graham said. “The project is still in the infancy stages where we’re trying to train the program to recognize the cell tips. Eventually we want to develop a computational model of cotton fiber development to show how the fiber cell grows based on all of those observations.”

Kariya is working on the computer vision algorithm that can recognize the important features of the cotton fiber cells, while Graham is gathering a set of microscopic images on which to train and test the machine learning algorithm.

The ultimate goal of Graham’s research is to understand how the cotton fiber cell develops and make the fiber from common upland cotton more like the high-quality fiber from Egyptian or Pima cotton. Those extra-long, thin fibers produce stronger, thinner and more durable fabrics.

“IDEA-TREND and this collaboration provided a great opportunity for us to explore a project that Candace Haigler and I have been talking about for a long time, but was never my top priority,” Graham said. “Working with computational researchers gives us a totally new dimension with which to explore our data. And of course, during a pandemic it provides a wonderful alternative to most of my work, which is done in front of a microscope or in a greenhouse, where access is limited due to COVID-19. If something really great comes of the collaboration, then it’ll be a wonderful side effect of such an unfortunate circumstance.”

Haigler was able to find funding to partially support Kariya this semester, which strengthened the partnership.

Haigler and Graham’s research is primarily supported by Cotton Incorporated, a not-for-profit organization funded by U.S. upland cotton producers and importers with its world headquarters in Cary, North Carolina.

Johanna Elsensohn and Aram Mikaelyan

Johanna Elsensohn is a sixth-year Ph.D. student in the Department of Entomology and Plant Pathology. She is co-advised by Coby Schal and Hannah Burrack.

The focus of her thesis work is the behavior and ecology of a tiny agricultural pest. This pest is a fruit fly that lays its eggs in small fruits, like raspberries, blackberries, blueberries and strawberries, rendering the fruit unsellable. She wanted additional expertise to help her analyze previously collected data on the community of microbes found on the surface of blackberry fruit.

Reiskind connected her with Aram Mikaelyan, an assistant professor in the same department, who studies insect-microbe interactions for wood-destroying pests such as termites.

“We hope to relate the change in the microbial community over the season, to the level of infestation in the fruit,” Elsensohn said. “I was really excited that he was willing to collaborate and show me the ropes. He is an expert at doing microbial analysis, which is not something I felt 100 percent comfortable doing on my own.”

During Elsensohn’s thesis research, she studied host-marking behavior, where after a female fruit fly laid an egg in a berry, she would leave a tell-tale marker to let other insects know that the berry was “occupied.” If this mark is from a type of bacteria or yeast, the microbial analysis should uncover it.

This summer, Elsensohn was focused on writing her dissertation. This fall, she and Mikaelyan will start collaborating on analyzing her microbial data.

“It’s nice to be able to collaborate with people that you think you might not have a lot in common with research-wise, but it turns out we do,” Elsensohn said.

Elsensohn’s research was supported by a National Science Foundation integrative graduate education and research traineeship and funding from the Southern Integrated Pest Management Center and the North American Bramble Growers Research Foundation.

Savannah Curtis and Rachel Wooliver

Savannah Curtis is a fourth year Ph.D. student in the Biomathematics Graduate Program. Curtis works with Dr. Cristina Lanzas in the College of Veterinary Medicine studying health-care associated infections in humans, specifically those caused by the Clostridioides difficile bacteria, also known as C. diff.

Reiskind connected her with Rachel Wooliver, a postdoctoral research scholar in Seema Sheth‘s lab in the Department of Plant and Microbial Biology.

Wooliver studies whether certain plants have evolved the ability to tolerate warmer temperatures in response to climate change or being introduce to a new area.

Together they are working to apply advanced statistical models to Wooliver’s data on the thermal tolerance of plants to uncover if certain seed families within plant populations have adapted to climate change better than other seed families.

“When I was approached about this project, I had just finished a course in Bayesian statistics, so it fit pretty nicely into my wheelhouse,” Curtis said. “It’s been really cool to take methods and statistical software I know how to use and apply them to a project in a research field outside of my own. And I would have never done anything with plants, had it not been for this collaboration.”

Curtis is currently testing out some statistical models on a subset of Wooliver’s data, but they are close to being able to analyze the complete dataset.

Collecting the data Curtis needs for her thesis research was halted due to the COVID-19 pandemic, but she has been able to apply her skills to study COVID-19 as well as collaborating with Wooliver.

“Interdisciplinary research is helpful beyond just research,” Curtis said. “It’s also been hugely important with learning how to talk to people. Figuring out how to translate my math jargon and talk to someone else with their own expertise and their own jargon? It’s not always easy but it’s one of the most rewarding things. Being able to get two or more parties talking and excited about the same thing is reinvigorating.”

Curtis is supported by a prestigious National Science Foundation graduate research fellowship.

CALS is working together to tackle the challenges of COVID-19

This post was originally published in College of Agriculture and Life Sciences News.