Skip to content

Latest commit

 

History

History
109 lines (85 loc) · 4.36 KB

readme.rst

File metadata and controls

109 lines (85 loc) · 4.36 KB
Raw

pixels

Data Organisation

New raw data should be moved into your raw folder, inside a folder whose name starts with the format "YYMMDD_mouseID" for that recording session. These files are automatically compressed for storage so data processing copies them to the interim folder if they are not already present there, uncompressing if required. The data is processed from interim and the results saved in processed, where they can be accessed for analyses. Note that everything in the interim folder can be regenerated, so the entire interim folder can be deleted without losing anything important.

data
├── interim
│   └── YYMMDD_mouseID_extrainfo
│       ├── YYMMDD_mouseID_gN_t0.imec0.ap.bin
│       ├── YYMMDD_mouseID_gN_t0.imec0.ap.meta
│       ├── YYMMDD_mouseID_gN_t0.imec0.lf.bin
│       ├── YYMMDD_mouseID_gN_t0.imec0.lf.meta
│       ├── USB_Camera.tdms
│       ├── USB_Camerameta.tdms
│       ├── NeuropixelBehaviour(0).tdms
│       └── cache
│           └── * see below
├── processed
│   └── YYMMDD_mouseID_extrainfo
│       ├── YYMMDD_mouseID_gN_t0.imec0.ap_processed.h5
│       ├── YYMMDD_mouseID_gN_t0.imec0.lf_processed.h5
│       ├── NeuropixelBehaviour(0)_processed.h5
│       ├── action_labels_0.npy
│       ├── sync_0.png
│       └── lag.json
└── raw
    └── YYMMDD_mouseID_extrainfo
        ├── YYMMDD_mouseID_gN_t0.imec0.ap.bin.tar.gz
        ├── YYMMDD_mouseID_gN_t0.imec0.ap.meta.tar.gz
        ├── YYMMDD_mouseID_gN_t0.imec0.lf.bin.tar.gz
        ├── YYMMDD_mouseID_gN_t0.imec0.lf.meta.tar.gz
        ├── USB_Camera.tdms.tar.gz
        ├── USB_Camerameta.tdms.tar.gz
        ├── NeuropixelBehaviour(0).tdms.tar.gz
        └── extra
            └── ** see below

* Some basic analyses will save the result of their calculations into this cache folder.

** Any files collected on the recording day that should be ignored by the pipeline should be put inside a subfolder(s) within the session's folder. The name of the folder(s) are not important.

Pipeline

[raw] Compressed raw data

[interim]    raw spike data             behavioural tdms          LFP data          camera
                   ┃                           ┃                     ┃                 ┃
                   v                           ┃                     ┃                 ┃
              spike sorting                    v                     v                 v
           downsample to 1kHz      create action labels 1kHz      resample            DLC +
           phy manual curation                 ┃                     ┃              resample
                   ┃                           ┃                     ┃                 ┃
                   v                           v                     v                 v
[processed]   spike data                 action labels           1kHz LFP      1kHz DLC coordinates

The first processing step run will align the probe recordings and the behavioural data and save the lag - that is, the number of points of overhang at the start and end of the behavioural data - into lag.json in the processed folder. The sync step will also save a figure to sync_0.png, which should be checked to visually confirm that the syncing went well.

Conda

These commands can be used to create a conda environment with all libraries used by the pipeline:

conda create -n pixels numpy pandas nptdms scipy matplotlib opencv -c conda-forge
conda activate pixels
pip install ffmpeg-python spikeinterface probeinterface

This does not include deeplabcut and it's dependencies - see the deeplabcut docs for how to install.

Resources