urchin's Introduction

Urchin

A tool to perform batch analyses on spike-sorted multielectrode array data.

How to run

Since this is not a package at this time, the file urchin.py contains the Urchin class used to run analysis. Necessary dependencies include numpy, scipy, and h5py.

The urchin class takes three parameters:

Parameter	Description
path_to_sorted_data	Directory with spike-sorted data
path_to_raw_data	Path to raw .h5 data
path_to_json	Path to JSON stimulus config file

The analysis scheme happens as follows:

Each stimuli in the config file must be assigned to bounds dictated by the TTL pulses.
- This is run when Urchin is constructed, and stored in the object parameter stim_bounds, where each value in the list is a separate stimulus.
Any sub-stimuli timing features must now be found, which may or may not be present depending on the stimulus. For example, a MovingBar stimuli has multiple sub-stimuli for each direction of the bar.
- These sub-stimuli features can be determined by passing in the index in stim_bounds of the stimuli to analyze to the function split_into_sub_stimuli. The output of this function is used to perform batch analysis.
Using the output of split_into_sub_stimuli, the spike times for each 'good' cluster marked on Phy must be extracted within the timing bounds of the sub-stimuli.
- Pass this output into the function extract_spiketimes_within_bounds, which extracts spike times within the bounds of the sub-stimuli.
At this point, several spike count based analyses can be performed, such as the PSTH for a cluster, the receptive field of the cluster, the vector DSI of a cluster, or the average firing rate of a cluster over the sub-stimulus time period.