1.8.2 and 1.8.3 Practical: Spike sort raw data and inspect results with viewephys

Introduction

The full description of spike sorting at IBL is in our white paper !

This tutorial will show you how to:

Inspect a raw data file prior spike sorting

A_viewephys_on_raw_data.md

Open a spikeglx recording, during acquisition or after the fact and look at the raw data quality
Look at the bad channel detection and adjust thresholds if necessary

Launch a spike sorting and look at intermediate results

B_run_iblsorter.md

launch the spike sorter on the dataset (Optional)
look at and interpret the intermediate outputs of the spike sorter

Overlay sorter result over raw-data

C_Overlay_sorter_results_on_raw_data.ipynb

Load a snippet of raw data, and destripe it
Load the corresponding spike sorting data occurring during this raw data snippet
Place the information into viewephys, and navigate the user interface

You will investigate:

What signals are visible onto the raw data, giving you a sense of why we want to perform destriping
Whether the quality of the raw data is subjectively good
Whether many channels are flagged as "bad" and why
How well conditionned the cross
Whether the spike detection is subjectively good
Whether there is information contained in the spikes rejected as part of bad units

Pre-requisite

Python

Install ibl compatible environment
Install viewephys in this environment using pip

Data

Download sample data, following the data download instructions from the course

Additional resources

Viewephys github repo