Hi all, my datafile (.plx) has already had LFP and spikes, and the spikes of each channel have been pre-sorted by Plexon system. Is it possible and convenient to use wave_clus to re-sort each signal channel?
Hi @Karlie.
As far as I know, the Plexon systems initially sample the spikes at 40k. These are the presorted spikes that your .plx files have incorporated.
However, the LFPs that are stored in the .plx files (at least those that I worked with) are downsampled at 10k.
The spikesorters on Brainstorm are supposed to be used on raw signals (in your case, Brainstorm would think that raw is 10k). If you run the spike sorters through Brainstorm at a 10k signal, you'll probably get the spikes that have relatively wide waveforms and high amplitudes, but I'm not sure if it would be an accurate representation of your entire spiking activity.
Are your LFPs sampled at higher frequencies or you also have them at 10k?
If you are at 10k and you think you absolutely need to run the spikesorter again, I suggest you run Waveclus and compare the results from WaveClus and from the pre-spike-sorted .plx.
The other workaround you can do, is run the Plexon offline spike-sorter (you need a license for that - I assume your lab has one if you have those files). The new clusters you form on this spike-sorter are automatically updating the .plx file, therefore you will be able to import the updated .plx file in Brainstorm and completely skip the spike-sorting step within Brainstorm.
Hope that helps,
Konstantinos
1 Like