Tulane University researchers have developed an AI tool called MAFDA (Novel Machine-learning-based Automatic Fly-behavioral Detection and Annotation) that can analyze the behavior of fruit flies. By using cameras and specialized software, the system can track and identify complex interactive behaviors of individual flies within a group, enabling researchers to compare and contrast behaviors across different genetic backgrounds.
Fruit flies have been instrumental in scientific studies due to their simple genome and short lifespan, providing insights into inheritance, immunity, and other aspects of biology. The MAFDA system improves upon previous algorithms by accurately tracking individual flies within a group, making behavior studies more efficient.
Wenkan Liu, the graduate student who developed MAFDA, emphasized the significance of the platform, stating that it accelerates research, reduces human error, and offers detailed insights into behavior genetics. The tool enhances reproducibility and opens up possibilities for large-scale behavioral analysis.
MAFDA was initially developed for a study that explored the gene responsible for fruit flies’ perception and production of pheromones. The findings challenged the prevailing notion that separate genes control these processes, with potential applications in human behavioral evolution, metabolism, and sex dimorphism.
The researchers envision broader applications for MAFDA, including studying other insects, mice, fish, and examining the effects of drugs. The system’s capabilities can be further improved by providing more data, allowing for the identification of various behaviors such as courtship and feeding.
MAFDA is already being used in other research projects at Tulane, and efforts are underway to package the system for broader use within the scientific community. While the original goal was to identify the health status of flies, the researchers hope that the tool will find widespread use and potentially expand its applications in the future.
Source: Tulane University