Python Scripts and Results

Description

This repository contains scripts that demonstrates presence of high-order cell assemblies from wide-field calcium imaging data. This is a collabo.pngrative project with Prof. Spencer Smith and Dr. Yiyi Yu from University of North Carolina, Chapel Hill. Part of the results has been presented at the 3rd International Conference on Mathematical Neuroscience. The conference submission and presentation can be found in icmns_submission

Data preprocessing

Run the script combine_datasets.m to pre-process the data. The data should be stored in the data/ directory. Each directory contains a different experiment. For each experiment, there are simultaneous recording from the primary visual cortex (V1, Combo3_V1.mat) and a higher visual cortex (area AL, Combo3_AL.mat).

These files can be opened via Matlab, or using the mat4py library.

Decrease sparsity over trials

Run the script sparsity_over_trial.py to see the decreasing sparsity across repeated trials. You can change the area variable to see the result for V1 or AL. The summary plots will be saved in figures/sparsity/. Below is an example result for area AL.

Presence of high-order cell assemblies

Run the script analyze_data_populaiton.m to see evidence of high-order cell assemblies. The script will generate the following result for each experiment session

• sparsity over trial (figures/sparseness_over_timestim*, Fig. 1C)
• subset index calculation (figures/subset_indexstim* Fig. 1D)
• Number of unique high-order cell assemblies (cellAssembly_* Fig. 1E)
• Accuracy of cell assembly occurance time (cellAssemblyjitter*, Fig. 1F)
• Example raster plots (figures/exampleRasterStim* slide #27)