chronic-neuropixels-protocol

February 9, 2021 · View on GitHub

(Chronic Neuropixels recordings in mice and rats; Rik van Daal , Cagatay Aydin , Frédéric Michon , Arno Aarts , Michael Kraft , Fabian Kloosterman, Nature Protocols; NP-P200042C, 2021 )

alt text

Neuropixels chronic fixtures

Parts can be printed at Materialise

CAD files for 3D parts

Data and Code

System requirements

Operating system: Windows Server 2012 R2 Standard

Versions of the software:MATLAB (R2020b), Python (3.7.7)

Overview of data analysis steps

Raw data is filtered and sorted using adapted version of ecephys_spike_sorting pipeline

Filtering: CatGT Version 1.2.6 is used with parameters given below

catGT_cmd_string = '-prb_fld -out_prb_fld -aphipass=300 -aplopass=9000 -gbldmx -gfix=0.4,0.10,0.02'

Sorting: Kilosort2 is used with default parameters

Number of good units are extracted by using custom written update_cluster_group.m function located in code

% amplitude_cutoff, presence_ratio, isi_viol are generated by using
% https://allensdk.readthedocs.io/en/latest/_static/examples/nb/ecephys_quality_metrics.html
idx = Tmetric.amplitude_cutoff<0.1&... 
    Tmetric.presence_ratio>0.95&...
    Tmetric.isi_viol<0.5&...
    ig; % index for Kilosort2 'good' label after noise removal

Spike statistics:

Spike count, amplitude, RMS values are extracted from filtered data by using neuropixels-evaluation-tools

Instructions for custom written code

  1. Download and extract the meta data

  2. Add code and data folder into the MATLAB path (using addpath function)

  3. Change filtered_folder as your data directory where data is extracted and change figure_folder as figure directory where one can have pdf printed figure panels in plot_event_rate_unit_yield.m script. This scipt reads meta data as explained below;

%% read data files
% hparams.txt - user defined meta file (contains; recording date, probe...)
% XX.meta - meta file generated via spikeglx during the data acqusition
% files indicated below generated via neuropixels evaluations tools
% XX.ap_sum.txt - summary metrics 
% XX.ap_chan.txt - metrics generated for each channel across the shank
% XX.ap_hist.txt - spike amplitudes histogram 
% cluster_group.tsv - summary file indicating unit yield generated via KS2

  1. Run plot_event_rate_unit_yield.m script.

  2. Generated figure panels can be found in figures