Skip to content

Latest commit

 

History

History
108 lines (67 loc) · 6.6 KB

README.md

File metadata and controls

108 lines (67 loc) · 6.6 KB
Raw

Locomotion data analysis and presentation

(start : Oct 2017)

The scripts and classes are used to extract/analyze/display locomotion, imaging and electrophysiological data recorded in the Cerebellum of mice walking on a runged treadmill. The recordings are performed in the Isabel Llano's lab starting Autum 2017.

The raw data is been recorded with ACQ4 as well as ScanImage and is stored on lilith and spetses, the lab backup servers.

Information about the experiments are furthermore stored in an google document Experiments Cerebellum awake : April 2017 onwards. This google spreadsheet is read by the analysis scripts and used to access data of specific mice.

Content of the documentation

Overview of the data pre-processing pipline

Analysis pipline

Calcium imaging experiments

The analysis below is applied to experiments involving caliucm imaging with ScanImage. ScanImage saves acquires frame stacks as tif files with are not directly linked to the associated recording in ACQ4. However, the time stamp of the ScanImage file and the ACQ4 recording are identical (up to a second precision).

File What it does
(locorungs) getCalciumTracesDetermineRois.py The script read the ScanImage tif files and feeds them into the Suite2p analysis pipeline. Suite2p parameters are adapted to our recordings. However, the script requires interaction if more than the standard 5 tif files exist in the raw data folder. The average image is generated and saved in the suite2p analysis folder.

Typical work-flow to analyze calcium imaging data

  1. run (locorungs) getCalciumTracesDetermineRois.py to run Suite2p analysis in recorded tif files.
  2. run (suite2p)$ python -m suite2p launch suite2p and visually sort cells vs. no-cells.

Experiments with walking behavior recordings

The analysis below concerns experiments during which animal behavior has been recorded through the rotary recording on the treadmill and/or the animal was filmed with the high-speed camera (GigE). Global movement parameter is the overall walking progress of the animal during the recording.

File What it does
getWalkingActivity.py The script extracts the Rotary encoder data and saves it to hdf5 file.
plotRecordingOverviewPerAnimal.py The script uses the extracted Rotary encoder data and generates an overview figures.
extractBehaviorCameraTiming.py Sets the ROI location for the LED used for synchronization and determines the exact timing of individual frames.
If 3rd LED is not blinking : A modified version of this script exists in case the 3rd LED is not blinking (due to a loose connection). The modified version is part of a different git branch called LED3broken. This branch can be activated by switching from the main branch to LED3broken by typing git checkout LED3broken in the command line. The script is launched as usual through (locorungs37) python extractBehaviorCameraTiming.py. Switching back to the main (default) branch can be done with git checkout main.
getRawBehaviorImagesSaveVideo.py The script extracts images recorded with the high-speed camera. The extracted images are saved as avi video for the tracking procedures.
analyzePawMovement.py The script uses openCV to analyze the behavioral videos. Paw and rung locations are extracted.

Typical work-flow to analyze global movement parameters during walking experiments

  1. run getWalkingActivity.py to read and save rotary encoder data.
  2. run plotRecordingOverviewPerAnimal.py to generate an overview figure.

Typical work-flow to analyze detailed paw movement during walking experiments

  1. run extractBehaviorCameraTiming.py sets the ROI location on the LED, extracts the luminosity trace and determines frame times.
  2. run getRawBehaviorImagesSaveVideo.py turn behavior camera recordings into movie.
  3. run analyzePawMovement.py to track paw and rungs.

Experiments to access motion artefacts

getPixelflyImageStack.py - The script extracts images recorded with the Pixelfly camera and saves them to a stacked tif file.

plotMotionArtefacts.py - The script uses the motion correction coordinates from the ImageJ Moco plugin. A figure is generated with comprises several recordings.

Typical work-flow to analyze experiments with walking behavior

  1. run getPixelflyImageStack.py extract Pixelfly images and save them as tif
  2. run ImageJ with the MocoMacro.ijm macro : generates x, y coordinates for motion correction
  3. run plotMotionArtefacts.py generate summary figure over several recordings

Work-flow to generate overview figure

The experiment overview figure contains information from the rotary enconder-, the video- and the calcium imaging recordings. It serves to have quick overview of the recordings, their implementation and the number of days an animal has been recorded. The overview figure can be generated with the plotRecordingOverviewPerAnimal.py script. Before this script can be run, data has to be extracted and some pre-analysis is required. In particular, the following scripts need to be run before generating the overiew :

  • getWalkingActivity.py : extract data from the rotary encoder recording and calculates speed of the wheel.
  • getRawBehaviorImagesSaveVideo.py : extract data from the high-speed video recordings, generates videos and saves timing information of the video recording.
  • getCalciumTracesDetermineRois.py : runs Suite2p on the calcium-imaging recordings obtained with ScanImage.

To complete the initial analysis, some more scripts need to be run on the data.

  • extractRungLocation.py : extracts the location of the rungs from the video generated with getRawBehaviorImagesSaveVideo.py
  • extractPawTrackingOutliers.py : Uses the paw tracking data generated with DeepLabCut and remove mis-tacked paw positions based on the paw displacement between frames. Large, unrealistic displacements are removed.
  • extractSwingStancePhase.py : uses information from the wheel speed, paw position and rung position to separate paw trajectoris into swing and stance phases