This repository contains analysis code and data for the paper Neural Networks of the Mouse Primary Visceromotor Cortex.
This software is designed to work with any computer running Python 3 but was developed for Linux, and specifically tested on Ubuntu 22.04.5 and Linux Mint 21.3. As detailed in the installation guide, all package requirements are listed in requirements.txt. No non-standard hardware is required.
Installation can be performed in seconds using a Python 3 virtual environment:
python3 -m venv mpf_venv
. mpf_venv/bin/activate
pip install -r requirements.txt
The directories SW******-**A each contains anterograde and retrograde tracer quantificatification data for a single rodent brain.
The file case_sheet.csv contains the injection site and tracers used for each rodent brain.
The mpfc_analysis.py program performs analysis and visualization for MPFC regions. It reads the raw data listed in case_sheet.csv, computes the fraction and density connectivity matrices, and plots the connectivity matrices as scatter plots and 2D heatmaps. You can run script from within the repo:
python -m mpfc_analysis
resulting in the output of approximately 100 images with corresponding metadata will be written to plots/ over the span of about 5 minutes.
The notebook do_mpfc_analysis.ipynb computes connectivity "reciprocity" between brain regions, using fraction matrices of both anterograde and retrograde tracers. mpfc_anterograde_ctx_fractions_all_merge.csv and mpfc_retrograde_ctx_fractions_all_merge.csv are example inputs. The numbers in these sheets are merged fraction results across multiple rodent brains.