Spike sorting pipeline for OpenEphys

[zivadinac]

Make a spike sorting pipeline for data recorded with OpenEphys GUI and acquisition board.

The pipeline has to do the following:

• Automatically sort spikes

• Generate cluster quality measures

• Enable manual curation (phy)

• Easily process data acquired with different drives

• Concatenate different recording sessions and process them together (whole recording day)

• Output data in the standard clu-res JC lab format

• Maybe: place fields, eeg(h), swr detection? (consider these, but it's not spike sorting pipeline)

[zivadinac]

Specification:

• The pipeline will be delivered as a a script with params: drive geometry file, detection threshold, output directory, input files The script will concatenate input files and process them together. Output of the script will be a report with cluster quality measures and clustered data ready for manual curation. Optionally, user should be able to choose clu-res as output format.
• There will be a separate script for generating cluster quality report, so that it can be used after manual curation
• Additionally, there will be one more script to convert manually curated data to clu-res format of JC lab. Optionally, user should be able to choose to do this automatically when invoking the main script.

[zivadinac]

Implementation:

• Use spikeinterface for implementation

• Specify different drive layouts (geometries) with config files (example); we will have at least two, for small and big drive

• Process each tetrode separately (example)

• Concatenate different recording sessions and process them together (example)

• Use mountainsort for automatic sorting

• https://github.com/zivadinac/open-ephys-spike-sorting/issues/2#issue-1163057732

• Generate cluster quality report with spikeinterface (example) and put it into separate script: https://github.com/zivadinac/open-ephys-spike-sorting/issues/3#issue-1163060055

• To convert output to (phy) format, adapt the existing script from the axona pipeline: https://github.com/zivadinac/open-ephys-spike-sorting/issues/4#issue-1163062577

• Convert output from (phy) format to clu-res format like in axona pipeline: https://github.com/zivadinac/open-ephys-spike-sorting/issues/5#issue-1163071528