Paper Figure Reproduction

"Chen X.F., Tao X.M., Zhong Z.C., Zhu F.Y., Zhang Y.Q., Wang Y.Z., Wang M., Sun L., Li Y.X., Ouyang Y., Ding Z.Y., An M., Zhao Y.Y., Zhou J.F., Zhang R., Xiong W., Ji N., Li Y., 2026. Sub-second dopaminergic reinforcement orchestrates juvenile social play and is disrupted in Shank3 deficiency. In preparation"

Author: Chen Xinfeng (2025-08-09). Revised: Tao Xianming (2025-08-15), Chen Xinfeng (2026-08-07).

Run the code in this repository to reproduce all the figures in the paper. The images below show example outputs. Windows, Linux, and macOS are supported.

Fig1_S1 Fig2_S2 Fig3_S3 Fig4_S4 Fig5_S5 Fig6_S6 Fig7_S7

1. Code and Data Download

The data are openly available for download (file: Social_Seq_DATA.rar, ~4.1 GB) via Mendeley Data🔗 or Baidu Cloud🔗. After downloading, extract the archive and rename the folder to Figshare.

The source code is openly available at https://github.com/LiLab-CIBR/Social_Seq-Paper-Figure-Replication.

The directory structures of the data and code are one-to-one matched. Please manually copy the contents of Fig*/data/* from the data folder into the corresponding code folder.

You can use AI coding assistants such as OpenCode, Codex, or Claude. Just provide one of the prompts below, and they will automatically install the environment and generate all the figures.

OpenCode

Provide the full prompt:

Example 1 — install everything and generate all figures:

"Please follow the guide at https://lilab-cibr.github.io/Social_Seq/assets/figure_reproduce/AGENTS.md to install and run the project and generate all figures. My data is at C:\<<<<path>>>>\Social_Seq_DATA-v20260809.zip; please use the domestic mirror, otherwise installation may be very slow."

Example 2 — clean all results / restore initial state:

"Clean all results and restore the initial state."

Example 3 — generate a single figure:

"Please generate Figure 1D for me."

For the detailed manual installation steps, please refer to "https://github.com/LiLab-CIBR/Social_Seq-Paper-Figure-Replication". The general process is as follows.

First, install the dependency environment using uv.

git clone https://github.com/LiLab-CIBR/Social_Seq-Paper-Figure-Replication .  # Download the code to the current folder

python3 -m pip install uv          # uv is a lightweight Python package manager, similar to conda

uv sync --index-url https://pypi.tuna.tsinghua.edu.cn/simple  # Users in mainland China may use the Tsinghua mirror; otherwise, the installation will be very slow
uv run python --version  # Verify the Python version (should be 3.12)

Note: Do not use uv and conda simultaneously, as this will cause package installation conflicts. Please run conda deactivate first before running the commands below.

Install Custom Packages 

uv pip install git+https://github.com/chenxinfeng4/multiview_calib.git
uv pip install git+https://github.com/chenxinfeng4/LILAB-py.git ../LILAB-py

Generate a Single Figure: Fig1D.pdf

You can also run the scripts Fig*/Fig*.py one by one to generate a single figure. 

uv run python Fig1_S1/Fig1D.py   # The result is saved to Fig1_S1/result/Fig1D.pdf

Note: If any error occurs, please make sure all the data files have been downloaded correctly.

Generate All Figures with One Command

uv run python main.py test
Figures such as Fig1.pdf, FigS1.pdf, Fig2*.pdf, ... will be generated and saved to the corresponding folders: Fig*/result/*.pdf, e.g., Fig1_S1/result/Fig1G.pdf.

Clean All Results and Reset

uv run python main.py clean

Or manually delete all the result files: 

rm Fig*/result/*pdf Fig*/result/*pkl Fig*/result/*png