collaboration project with the Shew lab
A mouse catches roaches. It gets 6 roaches per day. You can see the videos in prey_capture_videos
If a mouse catch the roach quick, we give them 1 roach every 10 minutes (a day of experiment can be done in 1hr)
If a mouse doesn't catch the roach, we give them 30-minute upper limit to catch a roach, then we swap a new roach in (that means a 3-hr experiment)
The updated time where the mouse get the roach (start_hunt) and caught the roach (end_hunt) are in this google sheet. Less updated, but stable record of prey capture information are in the repo folder prey-capture-record
The start_hunt and end_hunt time are based on the timestamp in the 1-hour videos in prey_capture_videos
Before the formal 6 roach * 4 days experiment, the mouse also get a pre-trial, which happens 2 days before the first formal trial (mouse was not food deprived before pre-trial).
Mouse get 1 roach at 5pm on the pre-trial day. The mouse have unlimited time to catch that roach. We take out the pre-trial roach next morning (10am-ish) if the mouse did not hunt for the entire night.
The file PreyCapture_D2.ipynb shows all analysis that I've done on this prey capture dataset.
I did it on the Hengen lab server, so the filepaths are a little different.
In Hengen Lab, each neural recording is put into a folder like this:
RCC00036_L1_W2_2025-05-19_11-28-08, where 2025-05-19 11:28:08 is the time we start the recording.
Each recording session usually last 1-3 days
The ephys data from the recording is sorted in 4hr blocks, so you will see files that looks like:
H_2025-05-20_15-28-11_2025-05-20_19-23-11_neurons_group0.npy
This filename means that it cotains the neural data from 2025-05-20 15:28:11 to 19:28:11
(each raw data file is 5 minutes, so the raw data file from 2025-05-20_19-23-11 ends at 19:28:11)
These 4 hour sorting blocks keeps going. A hunt session might contain 1 or 2 of these sorting blocks.
This google sheet records all neural data for this project, and also keeps track of the data clean process.
Video recordings always starts after the neural recordings.
Video files are formated like: [animal]_[cage-type]_[camera-name]_[time-range]
(e.g. RCC00036_bucket_e3v810a-20250522T083736-093737.mp4)
In rcc_prey_capture_record.xlsx, you will see a row like this:
trial start_hunt end_hunt capture_time Video_File 1 0:40:28 0:53:09 0:12:41 RCC00036_bucket_e3v810a-20250522T083736-093737.mp4
Given that the video starts at 16:37:06, the actual time of when the hunt starts is 08:37:36 + 0:40:28 = 09:38:04
Continue with the hunt started at 2025-05-22 09:38:04. The alignment is done completely on wall time.
To see the video of the hunt: watch between 00:40:28 and 00:53:09 in the 1hr video [RCC00036_bucket_e3v810a-20250522T083736-093737.mp4]
To get neural data of the hunt: get spiketime between 02:09:52 and 02:22:33 in the 4hr sorting block [H_2025-05-22_07-28-12_2025-05-22_11-23-12_neurons_group0.npy]
The spiketime start (02:09:52) is from 09:38:04 (hunt time, calculated in last secion) - 07:28:12 (sorting block start time)
The spiketime end (02:22:33) is from 08:37:36 (video start time) + 0:53:09 (video offset) - 07:28:12 (sorting block start time)