Study of mouse behavior in different gravity environments

Many experiments have analyzed the effect of the space environment on various organisms. However, except for the group-rearing of mice in space, there has been little information on the behavior of organisms in response to gravity changes. In this study, we developed a simple Active Inactive Separation (AIS) method to extract activity and inactivity in videos obtained from the habitat cage unit of a space experiment. This method yields an activity ratio as a ratio of ‘activity’ within the whole. Adaptation to different gravitational conditions from 1g to hypergravity (HG) and from microgravity (MG) to artificial 1g (AG) was analyzed based on the amount of activity to calculate the activity ratio and the active interval. The result for the activity ratios for the ground control experiment using AIS were close to previous studies, so the effectiveness of this method was indicated. In the case of changes in gravity from 1g to HG, the ratio was low at the start of centrifugation, recovered sharply in the first week, and entered a stable period in another week. The trend in the AG and HG was the same; adapting to different gravity environments takes time.

c) Histogram of log converted activity continuous (using data before low-pass filter) Supplementary Activity amount when sampled from time course data.
Time course data obtained log-converting the activity continuous data after low-pass filter is shown in Fig. S1_1 and a histogram of the data is shown in Fig. S1_2.
Figure S1_1 Time course obtained log-converting the activity continuous data after lowpass filter (horizontal scale is 1/30 s). Here, the region is divided into five areas (A< -12, -12≦B<-6, -6≦C<0.5, 0.5≦D<7, 7≦E) as shown in figure S1_2. A summary of the state of movement of mouse is shown in     Yellow is below the p-value upper limit. The activity was analyzed using the HomeCageScan (HCS) software and manual annotation. The software does not identify movements of less than 6 frames out of a 30frame movie. The identification accuracy of the software is over 80% on average. The estimated amount of activity from WT Rest of Fig. 2A control.

Supplementary
The vertical axis (Time) in Fig. 2A is shown in seconds/hour. Therefore, hour was converted into seconds and calculated as a ratio. Awake / Active (%) was calculated by subtracting averaged Rest/Immobility/Inactive(%) from 100%.

Setup
Start EthoVIsion and click "New Default Experiment".
Enter the required data on the Experiment Setting screen.
Click "Arena Settings" and determine the analysis area of video.
Click "Trial Control Setting" and use the default.
Enter the following items and click "save Changes" at the end.
-Method: gray scaling -Activity Settings: Activity Threshold 3 or 17 Background noise 1 Compression artifact filter on

Acquisition
Select "Open Acquisition" from "Acquisition" on the menu bar.
Set "Arena Settings", "Trial Control Settings", and Detection Settings" and press the green button on Acquisition Control screen to start data collection.

Analysis
Click "Data Profiles" and use the default.
Click "Analysis output" and press "Calculation" button.

Supplementary Information on AIS Activity analysis R script flow.
Output file name, threshold, inactive bins, active bins to a file.
Count first read data above antilog converted threshold as the Active, below as the Inactive.
Basically, the minimum point between peaks is calculated as X-coordinate of separation point (the threshold).
(It depends on the case. See R script in Supplementary R script.) Classified peak of probabilistic density by number of peak 4 peaks or more (198 combinations of peak positions and numbers) or 3 peaks (24 combinations of peak positions and numbers) or 2 peaks (1 combination of peak positions and numbers) Estimate probabilistic density using Kernel Density, instead of histogram.
Convert Data by -log 0.9 Read Data passed low-pass filter