The Striatum Organizes 3D Behavior via Moment-to-Moment Action Selection

Authors

Jeffrey E. Markowitz^1,2,5, Winthrop Gillis¹, Celion Beron^1,2, Shay Neufeld^1,2, Keiramarie Robertson^1,2, Neha Bhagat¹, Ralph Peterson¹, Emalee Peterson¹, Minsuk Hyun^1,2, Scott Linderman^3,4, Bernardo L. Sabatini^1,2, Sandeep R. Datta^1,#

¹Department of Neurobiology, Harvard Medical School, Boston, Massachusetts, United States
²Howard Hughes Medical Institute, Chevy Chase, Maryland, United States
³Grossman Center for the Statistics of Mind, Columbia University, New York, New York, United States
⁴Departments of Statistics and Computer Science, Columbia University, New York, New York, United States
⁵Present address: Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University. Atlanta, Georgia, United States

#Corresponding Author

Overview

This contains code necessary to regenerate key figures and analysis from this paper. All code is in MATLAB and has been tested against versions up to 2022b. The directories are arranged as follows:

1. fig_scripts MATLAB scripts to directly generate figure panels. It is recommended to clear the MATLAB workspace prior to running each script.
2. lib contains custom MATLAB classes and functions for data processing.
3. preprocessing contains MATLAB scripts to regenerate long-running intermediate computations. It is not necessary to run these prior to running the figure-generation scripts. These are included for completeness.
4. third_party useful third-party functions.

Installation

Installation instructions

1. Download the repository onto your local machine.
2. Add the path to the github repository (including subdirectories) to your MATLAB path. Note that you'll need the Signal Processing, Bioinformatics, Image Processing, and Parallel Computing Toolboxes installed.

Data and Panel Generation

How to obtain data.

1. Download data from Zenodo here . You should see the following directory structure after unzipping the data. Note that it is recommended to decompress the data in ~/Desktop, which will create the directory ~/Desktop/phanalysis_images – this is the default path for loading data in most scripts.

    .
    ├── 1pimaging_dls
    │   └── phanalysis_object.mat
    ├── decoding_results
    │   ├── decoding_results_1pimaging_cell_types.mat
    │   ├── decoding_results_1pimaging_moseq_hierarchy.mat
    │   ├── decoding_results_1pimaging_ncells.mat
    │   ├── decoding_results_1pimaging_ncells_zoom.mat
    │   ├── decoding_results_1pimaging_twocolor_cell_types.mat
    │   ├── decoding_results_1pimaging_twocolor_pseudopop.mat
    │   ├── decoding_results_1pimaging_twocolor_withinanimal.mat
    │   ├── decoding_results_photometry_moseq_hierarchy.mat
    │   └── decoding_results_photometry_regress.mat
    ├── dls_lesion_round1
    │   └── phanalysis_object.mat
    ├── dls_lesion_round2
    │   └── phanalysis_object.mat
    ├── photomephys
    │   ├── ephys_kernel.mat
    │   └── photomephys_analysis.mat
    ├── photometry_a2a
    │   └── phanalysis_object.mat
    ├── photometry_crosstalk
    │   ├── both
    │   │   └── hifiber_object.mat
    │   ├── greenonly
    │   │   └── hifiber_object.mat
    │   └── redonly
    │       └── hifiber_object.mat
    ├── photometry_dls
    │   ├── lasso_distance.mat
    │   ├── modelr_randomizations.mat
    │   ├── modelr_randomizations_warped.mat
    │   └── phanalysis_object.mat
    └── photometry_nac
    	└── phanalysis_object.mat
   

2. This contains everything you need to run the scripts in fig_scripts; this directory contains MATLAB scripts for generating panels from the manuscript. Be sure to clear all variables in between running scripts (clear all).