chronic-neuropixels-protocol

KO | EN

gitlite — search

#javascript #python #hacktoberfest #react #ai #typescript #llm #go #golang #android #machine-learning #rust #deep-learning #linux

chronic-neuropixels-protocol

★ 11 Open GitHub ↗

Repo requested by Cagatay Aydin (for a paper)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# chronic-neuropixels-protocol
(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](https://raw.githubusercontent.com/nerf-common/chronic-neuropixels-protocol/main/fixtures_overview.jpg)

## Neuropixels chronic fixtures

Parts can be printed at [Materialise](https://www.materialise.com/en/manufacturing?gclid=Cj0KCQiA3smABhCjARIsAKtrg6KI-4CloUFmDMtG961YggM_I_BZ4re97FsboS6jPqCWgjePmS5XPqQaAv8xEALw_wcB)

### CAD files for 3D parts

- [Single probe fixture](https://github.com/nerf-common/chronic-neuropixels-protocol/blob/master/CAD_files/single_probe_fixture)
- [Dual probe fixture](https://github.com/nerf-common/chronic-neuropixels-protocol/blob/master/CAD_files/dual_probe_fixture)
- [Movable probe fixture](https://github.com/nerf-common/chronic-neuropixels-protocol/blob/master/CAD_files/movable_probe_fixture)
- [Adaptors for stereotaxic frame](https://github.com/nerf-common/chronic-neuropixels-protocol/blob/master/CAD_files/adaptor_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](https://github.com/jenniferColonell/ecephys_spike_sorting)

__Filtering__: [CatGT](https://billkarsh.github.io/SpikeGLX/#catgt) Version 1.2.6 is used with parameters given below

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

__Sorting__: [Kilosort2](https://github.com/MouseLand/Kilosort) is used with default parameters

Number of good units are extracted by using custom written `update_cluster_group.m` function located in [code](https://github.com/nerf-common/chronic-neuropixels-protocol/blob/master/code)

```matlab
% 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](https://github.com/jenniferColonell/Neuropixels_evaluation_tools)

## Instructions for custom written code

1. Download and extract the [meta data](https://github.com/nerf-common/chronic-neuropixels-protocol/blob/master/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;

```matlab
%% 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
```		
4. Run `plot_event_rate_unit_yield.m` script.

5. Generated figure panels can be found in [figures](https://github.com/nerf-common/chronic-neuropixels-protocol/blob/master/figures)

chronic-neuropixels-protocol

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README