Demo of CRF code used in:

Identification of Pattern Completion Neurons in Neuronal Ensembles using Probabilistic Graphical Models

Luis Carrillo-Reid^, Neurobiology Institute UNAM

Shuting Han^, Columbia University

Darik O'Neil^, Columbia University

Ekaterina Taralova, Columbia University

Tony Jebara, Columbia University

Rafael Yuste, Columbia University

^denotes equal contribution

Journal of Neuroscience (2021)

https://www.jneurosci.org/content/early/2021/08/19/JNEUROSCI.0051-21.2021

Current Code

This repo contains the code as used in the associated publication.
To see the latest implementation and toolbox, click here

Software Requirements

Linux only (Tested on Ubuntu 18.04.5 LTS)
Recompilation of associated thirdparty mex-files requires GCC/G++ version 6.3.X

MATLAB Requirements

MATLAB 2019b with symbolic links (More recent versions of MATLAB are usually compatible)
Signal Processing Toolbox
Parallel-Computing Toolbox is not necessary, this demo is a single-process implementation

Third Party Dependencies

QPBO 1.32
GLMNet
MexCPP

Installation

Open a terminal in the folder "Pattern_Completor_Modeling"
Enter bash UBUNTU_SETUP.sh in the terminal
Enter "1" for "Yes" to install

Running the Demo

Open Matlab
Run the Run_Demo script
For ease of demoing, a small model option was included that is 1/10th the size of the network used in the publication.

Output

A figure highlighting the identified pattern completors and their respective coordinates is produced for each detected ensemble.