Computational modeling reveals biological mechanisms underlying the whisker-flick EEG

This repository contains postprocessed data used in the paper "Computational modeling reveals biological mechanisms underlying
the whisker-flick EEG" by Tharayil et al.

The tar files `original.tar.gz`, `disconnected.tar.gz`, and `rewired.tar.gz` contain the simulation configuration files used to produce the data in that paper. The file `original.tar.gz` contains the simulations using the unmodified circuit. The file `disconnected.tar.gz` contrains the simulations using the same circuit, but which isolate the contributions of specific presynaptic populations by removing most internal connectivity and playing in a portion of the spikes from the original simulation, as described in the paper. The file `rewired.tar.gz` contains all of the simulations using the Schneider-Mizell connectome.

The tar file `spikesWithPopulationsRemoved.tar.gz` contains the spike files, excluding spikes from specififed populations, which are replayed into the disconnected simulations (using both the original and Schneider-Mizell circuits) to isolate the contributions of specific populations.

The simulations described in the previous paragraph can also be found on our Github repo, along with a fill description of the directory structure. However, unlike the Github repo, the tar files also contain spiking outputs (`spikes.h5` files in the `reporitng` subfolders) and postprocessed EEG signals. In each simulation folder, the `pkls/` subfolder contains several `.pkl` files, each of which stores the summed contribution to the EEG for all cells of a particular type. By cloning the github repo and copying the `spikes.h5` files and the `pkls/` subfolders into the corresponding locations in the Github repo, the figures in the paper can be reproduced without rerunning the simulations.

As described in the Github repo, running the simulations requires a "coefficients file", which the CoreNEURON simulation engine uses to calculate EEG signals at run-time. These files are generated using the BlueRecording package. The tar files `originalCoeffFiles.tar.gz` and `rewiredCoeffFiles.tar.gz` contain the coefficient files used to calculate the EEG signals from the original and the Schneider-Mizell connectomes, respectively.  By copying these files into the locations indicated in the ReadMe of the Github repo, the step entitled `Generating the EEG weights files` in the ReadMe can be skipped.

If you do wish to generate the coefficients files manually, instructions are available in the Github repo. The file `EEG_Lambda.h5` is used to generate the coefficients file for recordings with a reference electrode over the lambda point. The corresponding file to generate the coefficients for recordings with a reference electrode over the S1HL is the file `EEG.h5`, available here.