Sociability Learning

Code to analyze the data from the paper "Conspecific sociability is regulated by associative learning circuits" Victor Lobato-Rios, Thomas Ka Chung Lam, Pavan Ramdya 2024

Installation

• Create a conda environment with python 3.7 and SLEAP version 1.2.8:

conda create -y -n sociability -c sleap -c nvidia -c conda-forge sleap=1.2.8 python=3.7

• Clone this repository:

$git clone https://github.com/NeLy-EPFL/Sociability_Learning.git

• Go into the repository folder and activate the conda environment:

$cd Sociability_Learning

$conda activate sociability

• Install this repository as a package:

$pip install –e .

Usage

The folder scripts contains examples of using the Sociability_Learning package. These scripts replicate the analysis and generate the plots from the paper.

• compare_model_with_random_networks.py: compares the hits from our network (defined in the script Sociability_Learning/utils_connectomics.py) with randomly generated networks that conserve or not the proportion of neuronal `hits' for each brain region (Fig. 3e).
• embedding.ipynb: generates the UMAP embedding based on proximity events from female-female control experiments (Fig. 1e-f; EDFig. 2).
• folders_to_process_*.yaml: list of data to analyze. Data can be downloaded from our Dataverse.
• generate_figures_videos.sh: runs several scripts to generate Fig. 1e-f, EDFig. 2, and Videos 3-7.
• get_sociability_index.py: computes the sociability index from the data specified in folders_to_process_behavior.yaml. Several parameters, including the choice of control experiments for computing the metrics' thresholds, are located at the top of the script (Fig. 1h-k; Fig. 2, Fig. 3b-c, EDFig. 5, 7, 8).
• get_sociability_model.py: obtains the network formed in the connectome by the cell types specified in the script Sociability_Learning/utils_connectomics.py.
• preprocess_data_from_2p_setup.py: obtains Delta F/F time-series from neural activity recordings, treadmill rotations from Fictrac, and the location of the freely-moving fly from SLEAP.
• twop_analysis.py: analyzes neural and behavioral data from two-photon recordings specified in folders_to_process_.yaml.
• video*.py: compiles videos specified in the name of the script.