A new AI system can decode the behaviors of fruit flies. Because this is crucial for future human genetics research

The MAFDA system can track individual flies within a larger group, identify behaviors, and compare those behaviors to fruit fly genotypes. (Photo by Tulane University)

How to tell if a fruit fly is hungry? Ask a computer.

While it may sound like a bad dad joke, it’s a reality at Tulane University, where researchers have developed a new AI tool that can tell you if a fruit fly is hungry, sleepy, or singing (yes, fruit flies do sing). .

Dubbed MAFDA (for Novel Machine-learning-based Automatic Fly-behavioral Detection and Annotation), the system uses cameras and newly developed software to track and identify complex interactive behaviors of individual flies within a larger group. This allows researchers to compare and contrast the behaviors of fruit flies with different genetic backgrounds.

For more than a century, scientists have used the simple genome and short lifespan of fruit flies to decode the mysteries of heredity and immunity in humans with studies on Drosophila melanogaster, as fruit flies are known in the scientific world, winning six Nobel prizes. Fruit flies and humans share 60% of the same DNA.

Previous algorithms were less accurate at tracking individual flies within a group, but the MAFDA system makes it easier to study the tiny winged insects.

Fruit flies are like pioneers in discovering new things, from innate immunity to the chromosomal theory of inheritance, said corresponding author Wu-Min Deng, PhD., professor of biochemistry and molecular biology and Gerald & Flora Jo Mansfield Piltz Endowed Professor in Cancer Research at the Tulane School of Medicine. Being able to quantify fly behavior is truly a breakthrough in behavioral studies.

Wenkan Liu, a graduate student at the School of Medicine who developed the MAFDA system, said the importance of the platform is undeniable.

It speeds research, minimizes human error and provides complex insights into behavior genetics, Liu said. This tool is potentially critical as it improves reproducibility and paves the way for new explorations in large-scale behavioral analysis.

MAFDA was developed as part of a recent study, which found that the gene that causes flies to sense pheromones, chemicals produced by other fruit flies that are vital for attraction and other processes, is the same gene that check the production of pheromones first. These results, published in The progress of science, challenge the status quo view that separate genes control the production and perception of pheromones. The results have broad applications in the fields of human behavioral evolution, metabolism and sexual dimorphism.

Going forward, the researchers hope to see MAFDA used in a variety of applications. Jie Sun, lead author and postdoctoral fellow at the Tulane School of Medicine, said MAFDA could eventually be used to study insects other than mice and fish, and the system could be useful for studying drug effects in humans. .

The more information we give the machine, the better it can correctly identify different behaviors from courtship to feeding and so on, Sun said. This is a very important and significant tool.

MAFDA is already in use on other research projects in Tulane, and researchers are working to package the system so it can be used by more scientists both in Tulane and around the world.

That’s the goal, Deng said. The original idea was to be able to identify the health status of the flies. That may be asking too much right now, but we hope this is used more widely by the community and we hope we can go in that direction in the future.