When it comes to understanding the inner workings of fruit flies, researchers at Tulane University have found an unlikely ally in the form of artificial intelligence (AI).
Their newly developed AI tool, MAFDA (Novel Machine-learning-based Automatic Fly-behavioral Detection and Annotation), is taking fruit fly research to a whole new level.
Tracking and Identifying Fruit Fly Behavior
Equipped with cameras and advanced software, MAFDA can track and identify the intricate behaviors of individual flies within a group. This breakthrough allows scientists to delve into the behaviors of fruit flies with different genetic backgrounds, uncovering insights into their habits and tendencies.
For over a century, fruit flies, scientifically known as Drosophila melanogaster, have been instrumental in deciphering the mysteries of inheritance and immunity. These tiny creatures, sharing 60% of their DNA with humans, have earned fruit fly researchers six Nobel Prizes.
While previous algorithms struggled to accurately track individual flies within a group, MAFDA revolutionizes the process, making fruit fly studies more accessible and precise than ever before.
"Fruit flies are like pioneers in the discovery of new things, from the chromosome theory of inheritance to innate immunity," explained Wu-Min Deng, Ph.D., a professor of biochemistry and molecular biology at Tulane School of Medicine.
"To be able to quantify the flies' behavior is really a step forward in behavior studies."
Wenkan Liu, the mind behind MAFDA and a graduate student at the School of Medicine, highlights the significance of the platform.
By accelerating research, minimizing human error, and providing intricate insights into behavior genetics, MAFDA holds the potential for large-scale behavioral analysis and opens doors to new explorations.
The researchers discovered that the gene responsible for fruit flies' perception of pheromones is also involved in pheromone production. This finding shatters the notion that separate genes control these processes and has far-reaching implications for the fields of human behavioral evolution, metabolism, and sex dimorphism.
Wide Array of Applications
Looking ahead, the researchers envision a wide array of applications for MAFDA. Lead author Jie Sun, a postdoctoral fellow at Tulane School of Medicine, envisions the tool being used to study not only other insects but also mice, fish, and even the effects of drugs.
With each piece of information fed to the machine, its ability to accurately identify various behaviors, from courtship to feeding, could improve over time.
MAFDA is already making waves in ongoing research projects at Tulane. The team is working on packaging the system for wider usage, aiming to make it available to scientists both within and beyond their institution.
"That's the goal," Deng said in a statement. "The original idea was to be able to identify the health status of flies. That may be too much to ask right now, but we're hoping this will be more broadly used by the community and hopefully in the future, we can go in that direction."
With MAFDA's insights into fruit fly behavior, researchers are embarking on a whimsical journey that promises to unlock further mysteries and push the boundaries of scientific exploration.
The study's findings were published in the journal Science Advances.
