Discussing the Ephys QC Pipeline

[Sarruedi]

Hi everyone

Yuheng (@yhcheah804) and me would like to start working on the panels for the QC for Ecphys data.

It would be great to get some input on the key metrics and panels to ensure we generate useful figures for the data release paper

We mentioned SNR a few times in our meetings and it would be good to know what metrics we should extract to show this. Also, should we show some ISI violation ratio, presence ratio, amplitude cutoff, firing rate distributions, and a drift/stability summaries?

Raw data QC: example traces and probe/channel map, or more? @severine2305

Does this cover what you need for the panels, or are there metrics you'd add or drop? It would also help us to know what the generally preferred format (per-unit vs. summary plots, single panels vs. grids, and file type/resolution) so the outputs slot straight into the figure.

@rcpeene We tried to find your notebooks, but couldn't locate them. Can you please send us the link?

[rcpeene]

Probably the most comprehensive notebook is this one:
https://alleninstitute.github.io/openscope_databook/predictive-processing-ephys/

for future reference, the other notebooks can be found at the website (linked on the main page of this repo) under the 'analysis' tab
https://allenneuraldynamics.github.io/openscope-community-predictive-processing/notebooks/intro_to_ephys_nwbs/