Swiss-knife for Single-Cell Electrophysiology

This repository contains the analysis code, Jupyter notebooks, and extracted feature files accompanying the following publications:

Joshi N, van Der Burg S, Celikel T, Zeldenrust F (2025) Neuronal identity is not static: An input-driven perspective. PLoS Comput Biol 21(12): e1013821. https://doi.org/10.1371/journal.pcbi.1013821

Neuromodulatory Control of Cortical Function: Cell-Type Specific Regulation of Neuronal Information Transfer Nishant Joshi, Xuan Yan, Niccolo Calcini, Payam Safavi, Asli Ak, Koen Kole, Sven van der Burg, Tansu Celikel, Fleur Zeldenrust bioRxiv 2026.03.13.711516; doi: https://doi.org/10.64898/2026.03.13.711516

Extracted Features

Joshi, N., Zeldenrust, F.& Celikel, T. (2025). Data files for: Neuronal Identity is not static: an input-driven perspective [Dataset]. Zenodo. https://doi.org/10.5281/zenodo.17625155

---

Overview

This project provides the computational workflows for analyzing single-cell electrophysiological recordings and reproduces the results presented in the associated paper.

Key methods and analyses:

• Feature extraction from electrophysiological recordings
• Model fitting using GLIF models (see https://github.com/Nishant-codex/GIFFittingToolbox/tree/master)
• Clustering of single-cell responses across experimental conditions (ACSF vs. drug)
• Dimensionality reduction: PCA, UMAP, and probabilistic CCA (pCCA)
• Manifold alignment for comparing cell populations
• MCFA for comparing extracted features for heterogeneity
• Exploratory analyses of cluster stability, spike train metrics, and impedance profiles

---

Repository Structure

notebooks_acsf/        # Analysis under ACSF conditions
notebooks_drug/        # Analysis under drug conditions
residual_code/         # Exploratory / legacy notebooks and scripts
│   └── Infomation_transfer/   # Cluster analysis, PCA, UMAP, stability checks
scripts/         # Containing scripts for visualiztion and cluster stability
src/         # Source code for extracting features and plotting utilities
│   └── single_cell_analysis/   # Cluster analysis, PCA, UMAP, stability checks
	│   └── CC_analysis_utils.py                                                          
			CC_ephys_features.py                                                          
			FN_ephys_features.py                                                          
			neuromod_utils.py                                                             
			plot_utils.py                                                                 
			UMAP_utils.py                                                                 
			utils.py                                                                      
/
LICENSE
README.md              # Project description and usage

---

Installation

Clone the repository and install dependencies:

git clone https://github.com/your-username/single_cell_analysis-clustering.git
cd single_cell_analysis-clustering
python -m venv .venv
# activate the environment
python -m pip install --upgrade pip
python -m pip install -r requirements.txt
python -m pip install -e .

Usage

To reproduce the analyses, open the Jupyter notebooks:

jupyter notebook notebooks_acsf/Clustering_attribute_sets.ipynb

• notebooks_acsf/ → Main analysis for ACSF
• notebooks_drug/ → Drug condition analysis
• residual_code/ → Supplemental and exploratory analyses

Figures and results generated from these notebooks correspond directly to the analyses presented in the paper.

Feature extraction examples

You can run the feature extraction scripts directly from the repository root with your own input and output paths.

python -m src.single_cell_analysis.feature_extraction.Action_potential --input /path/to/data --output /path/to/output

python -m src.single_cell_analysis.feature_extraction.impedance --input /path/to/data --output /path/to/output

python -m src.single_cell_analysis.feature_extraction.STA --input /path/to/data --output /path/to/output

python -m src.single_cell_analysis.feature_extraction.cc_features --input /path/to/data --output /path/to/output --mode both

Use --mode ephys, --mode waveform, or --mode both with the CC extractor depending on which feature set you want to produce.

---

Workflow

Below is a simplified workflow of the analysis pipeline:

flowchart TD
    A[Raw Electrophysiology Data] --> B[Feature Extraction]
    B --> D[Clustering & Manifold Alignment]
	D --> E[Multi-set correlation and Factor Analysis]
    E --> F[Comparative Analysis: ACSF vs Drug]
    F --> G[Figures & Results]

Loading

---

Citation

If you use this code or workflows in your research, please cite:

@article{joshi2025neuronal,
  title={Neuronal identity is not static: An input-driven perspective},
  author={Joshi, Nishant and van Der Burg, Sven and Celikel, Tansu and Zeldenrust, Fleur},
  journal={PLOS Computational Biology},
  volume={21},
  number={12},
  pages={e1013821},
  year={2025},
  publisher={Public Library of Science San Francisco, CA USA}
}

---

License

This project is licensed under the MIT License.