Data for: Roles of the cerebellar vermis in predictive postural controls against external disturbances
Data files
Mar 25, 2024 version files 666.39 MB
-
data.zip
-
DeepLabCut.zip
-
programs.zip
-
README.md
Abstract
The central nervous system predictively controls posture against external disturbances; however, the detailed mechanisms remain unclear. We tested the hypothesis that the cerebellar vermis plays a substantial role in acquiring predictive postural control by using a standing task with floor disturbances in rats. The intact, lesioned, and sham groups of rats sequentially underwent 70 conditioned floor-tilting trials, and kinematics were recorded. Six days before these recordings, only the lesion group underwent focal suction surgery targeting vermal lobules Ⅳ-Ⅷ. In the naïve stage of the sequential trials, the upright postures and fluctuations due to the disturbance were mostly consistent among the groups. Although the pattern of decrease in postural fluctuation due to learning corresponded among the groups, the learning rate estimated from the lumbar displacement was significantly lower in the lesion group than in the intact and sham groups (approximately 35% of the sham group). These results suggest that the cerebellar vermis contributes to predictive postural control.
README: Title of Dataset: Roles of the cerebellar vermis in predictive postural controls against external disturbances
- Access this dataset on Dryad
- In this study, we tested the hypothesis that the cerebellar vermis plays a substantial role in acquiring predictive postural control by using a standing task with floor disturbances in rats. The intact, lesioned, and sham groups of rats sequentially underwent floor-tilting trials, and kinematics were recorded. Although the pattern of decrease in postural fluctuation due to learning corresponded among the groups, the learning rate was significantly lower in the lesion group than in the intact and sham groups.
File structure and overview
About kinematic data
- In this study, rats applied with black ink markers were recorded using high-speed cameras, 2D coordinates of the markers on the cameras were determined using DeepLabCut (DLC), and 3D coordinates of the markers were determined using DLT method. Data.zip contains the 3D marker coordinate data files in the form of csv.
- The data for each rat in intact and lesion groups ("RXXX") is in the following folders: "Data.zip/data/YYYYYY_RXXX"
- If there are any trials in which the data spans two files, the second files (results of manual digitization) are in the following subfolders: "Data.zip/data/YYYYYY_RXXX/RXXX_dipp"
- The data for each rat in sham group ("SXXX") is in the following folder: "Data.zip/data/220228_220307_S2101-6"
- The file name and folder for each trial are listed in the "fileName" and "folder" columns of "files3.xlsx". If the data spans two files, also listed in the "fileName2" and "folder2" columns.
- If the digitized data by DLC do not meet the standards ("likelihood") stated in the paper, if it is obviously digitized incorrectly, or if the time window is outside the scope of this study, the field is left blank or marked with "nan" or "*".
About executable program and its accompanying information
- Programs.zip contains executable program for outputting research results (including figures) from kinematics data (main.py), the collection of functions used there (programs.py), information on each trial (files3.xlsx) , marker and segment definitions (markerList.xlsx and segList210921.xlsx), and information on data excluded from the analyses as outliers (elimFromPaper2.xlsx and elimFromPaperCrest.xlsx).
About DLC files
- DLC.zip contains the DLC configuration file (config.yaml) and trained network folder (shamMar26-trainset95shuffle1).
Description of each file
About kinematic data files (csv)
- The rows and columns indicate frames (200 Hz, the first frame corresponds to conditioning stimulus) and 3D coordinates of markers, respectively.
- Markers are in the order listed in "markerList.xlsx" (described later).
- x: forward(+)/backward(-), y: left(+)/right(-), z: upward(+)/downward(-) .
main.py
- The Python program to derive research results from kinematics. This research was conducted using Python version 3.7.4.
- Definitions of parameters:
- "E", "E2": information on whether to execute each program in the file.
- "I": basic information necessary for the analyses, such as marker list and data excluded from the analyses.
- "AS": an array representing trial numbers (rows in files3.xlsm) divided into the groups (intact, lesion, and sham).
- "AS2": of the above trial numbers ("AS"), 8 trials in the early and final stages of learning are extracted.
- "fAS": expansion of the AS or AS2 into a one-dimensional list.
- "dTS" (before line 530): collection of analysis results (time series). The keys "M", "S", and "CREST" represent the marker coordinates, segment angle (segments are described in "paperSegL" parameter), and horizontal coordinates of the iliac crest marker, respectively.
- "dSV": collection of analysis results (scalar). The keys "light", "ts", and "t60" represent the segment angles at the light off, start of floor tilts, and 60 frames after the tilt start. The keys "ts-light" and "t60-ts" represent the changes in segment angles between those time points.
- "a", "a2-10", "b0-7", "c0-10", "d", "d2"-"d10": transformations of dSV or d60 for statistical analyses and figure outputs.
- "dTS" (from line 530), "d60": time series of iliac crest marker coordinates and the coordinates 60 frames after the start of floor tilts.
- "tiltNum", "tiltNumPack": number of floor-tilting trials.
- "AB", "AB2", "AB_av": parameters and their transformation in the learning equation. AB[0] and AB[1] represent parameters A and B in equation (1) (see paper), respectively.
files3.xlsx
- This file contains information for each trial (row).
- Description of columns:
- "gyoAS": ID shared with other excel files (described later) to identify the trial number.
- "group": intact ("i"), lesion ("l"), or sham ("s") groups.
- "rat": rat ID.
- "fileType": indicates whether the kinematic data was digitized with DLC only ("dlc"), manually ("dipp"), or a mixture of those ("dlc-dipp").
- "folder" and "fileName": folder and file names of kinematic data.
- "folder2" and "fileName2": second folder and file names if the kinematic data spans two files.
- "markerList": list of markers included in kinematic data (specifically listed in "markerList.xlsx")
- "notIncHead", "outFootAgl", "outNose": "1" is entered if head segment, foot segment, and nose marker are excluded from the analyses, respectively.
- "tilt": "1" is entered for trials in which the floor tilted (trials in which the floor did not tilt due to malfunction were excluded from the analyses), respectively.
- "tiltNum": the number of tilting trials experienced by each rat.
- "4soku": "1" is entered if the rat is unable to keep standing with their hind limbs until the floor tilts.
markerList.xlsx
- This file contains lists of markers.
- Rows and columns are the names of markers and marker lists, respectively. The number in each cell represents the order of a marker in each marker list.
elimFromPaper2.xlsx
- A list of segment and time information that was excluded from the analyses due to outliers or missing. "1" is entered for the corresponding time.
- Description of columns:
- "fileName": file name (trial name) corresponding to "fileName" in files3.xlsx.
- "gyo": trial ID corresponding to "gyoAS" in files3.xlsx
- "segment": segment excluded from the analysis.
- "light", "light-ts", "ts", "ts-t60", "t60": timings and time windows mentioned above.
- "Biko": a short note explaining the reason for exclusion.
elimFromPaperCrest.xlsx
- Contains a list (rows) of trials in which the iliac crest was excluded from the analysis.
- Description of columns:
- "gyo": trial ID corresponding to "gyoAS" in files3.xlsx
- "note": a note explaining the reason for exclusion.
- "elimStat": "1" is entered if the frame after the floor tilt is excluded from the analysis.
- "start", "end": first and last frames in the excluded time window (the start of floor tilting is set to 0).
segList210921.xlsx
- Contains information such as markers that define each segment.
- Of the segments listed here, those used in the paper are listed in the "paperSegL" parameter in main.py.
- Description of columns:
- "t1", "t2": markers at both ends of the segment.
- "t1sub", "t2sub": the second marker if the end of the segment is defined by two markers.
- "target": initial values for calculating time series of segment angles.