Anne Jaap, Friedrich Schüßler, and Paul Mieske (Science of Intelligence): “Big mouse data: Characterizing mouse behavior through temporal statistics”

The study of animal behavior is rapidly changing due to recent advances in long-term recording and automated analysis. Here we use these new developments to characterize mouse behavior via their temporal statistics. We analyzed positional data (RFID detections) of groups of mice housed in complex environments over many months. We found that behavior spanning seconds to hours can be separated into three distinct temporal ranges or states: short states of up to 2 min, that correspond mostly to explorative and interactive behaviors; intermediate states between 2-20 min, consisting mostly of feeding and grooming; and long states beyond 20 min corresponding to sleep. Each state has a simple statistical description that allows for a simple model to recapture the broad aspects of the data. We further characterized these states across individuals and age and showed that the amount spent in each state is homeostatically controlled. Taken together, we uncovered a surprisingly simple and consistent description of the temporal statistics of behavior in mice. Our results open up new questions about the underlying mechanisms as well as similar characterizations in other species.

Photo courtesy of SCIoI Project 40.