I agree with Andrew's response. A lot of the data analysed in the lab where I currently do my PhD (and I guess most of the e'phys labs out there) is analysed in custom written code (Matlab or python). But there are also "out of the box" solutions like OpenElectrophy (http://neuralensemble.org/OpenElectrophy/, you can do spike sorting, LFP analyses, data visualization and so on..). OpenElectrophy is actually one out of many open source softwares for e'phys hosted by the NeuralEnsemble project (http://neuralensemble.org/).

Hope it helps!

I agree with Andrew's response. A lot of the data analysed in the lab where I currently do my PhD (and I guess most of the e'phys labs out there) is analysed in custom written code (Matlab or python). But there are also "out of the box" solutions like OpenElectrophy (http://neuralensemble.org/OpenElectrophy/, you can do spike sorting, LFP analyses, data visualization and so on..). OpenElectrophy is actually one out of many open source softwares for e'phys hosted by the NeuralEnsemble project (http://neuralensemble.org/).

Hope it helps!

Spike 2 is nice as a start, but a lot of individual groups have developed their own algorithms. Using Matlab is my prefered option.

I usually run the preliminary analysis in Offline Sorter and NeuroExplorer and then export the data into MATLAB for any further processing. I have seen other colleagues use Igor, but I never had a chance to work with it.

Thank you for you answers :)

So far I had the chance to work with two interface based on the same clustering method: 1) MClust: 100% matlab based and some compile C and 2) spike sort 3D. Both uses KlustaKwik for clustering (which is definitively not the best but does the work + a lot of manual clustering behind). I was impressed by spike2 sorting method (especially good for sorting burst of spike with variation in waveforms). Plexon is also very recommandable but i haven't had the chance to try to directly.

I think this discussion is very important for the neuroscience community, as there are no clear unified method to sort and attribute spikes to a putative neurons (for multi channel extracellular of course), and the need for better, more transparent sorting is definitively needed in order to align all those publications out there.

I would use UltraMegaSort2000. Peter

We just created a new sorting software, based on template matching algorithms. It is very fast, accurate and semi automatic. The code can sort with few percent errors cells in vitro and in vivo (this has been validated with juxta recordings), handling up to thousands of channels in your recording. The documentation can be found here http://spyking-circus.rtfd.org an the preprint is located there http://biorxiv.org/content/early/2016/08/04/067843

Now I know what I'm looking for. An open online review site (like Amazon reviews or TripAdvisor reviews) for spike sorting packages! I'd go with the one with 5 stars and the most reviewers and detailed comments including screenshots etc. Now that there are so many out there, I can't possibly try them all out by myself, can I?

I'm using Spike2 by the way. Interested in those new tools, but can't really decide. The most of them seem hard to learn (I may be wrong).

In the past, we contributed a spiking analysis toolbox to FieldTrip, it can be combined easily with other types of continuous data (no spike sorting though, and it's Matlab, which is not ideal).