CARMEN – Calcium Multiplexing Biosensor System

Preprint:
Anna Lischnig, Yusuf Erdoğan, Benjamin Gottschalk, et al.
A Quad-Cistronic Fluorescent Biosensor System for Real-Time Detection of Subcellular Ca²⁺ Signals.
Authorea, March 03, 2025. DOI: 10.22541/au.174098553.38904074/v1

Status: Under revision
Addgene: CARMEN will be available via Addgene pending publication.

---

This repository contains a Fiji macro and a set of Python scripts for processing and analysing four-channel calcium imaging data exported from Fiji (ImageJ).
The workflow is designed for the CARMEN calcium multiplexing approach, including steps for:

• ROI-based signal extraction
• Background subtraction
• Data aggregation
• Optional curve fitting for fluorescence bleaching analysis

---

📁 Overview of Workflow Components

1. Fiji Macro

Fiji_Makro_CARMEN_MeasureFourChannelROI.ijm
This ImageJ macro extracts mean grey values from manually selected ROIs across four imaging channels.
It saves both the ROI set and one CSV per image and channel.

---

2. Python Scripts

CARMEN_combine_and_backgroundsubstract_CSVs.ipynb

• Combines and background-subtracts ROI intensity values.
• Assumes the last ROI in each image represents background.
• Merges four per-image CSVs (e.g. Image01_Intensity FP1.csv to FP4.csv) into one Excel file.
• Includes both raw and background-corrected data.

---

CARMEN_combine_and_backgroundsubstract_CSVs_endpoint.ipynb

Two-part script:

Step 1:
Groups four per-channel CSV files (FB1 to FB4) per image and saves them as one Excel file.

Step 2:
Combines data from multiple Excel files into a single table per channel, extracts the second data row,
automatically determines background (last valid MeanX), and performs background subtraction (BS1–BS50).

---

CARMEN_combine_and_curvefitting_xlsmtemplate.ipynb

(or: CARMEN_combine_and_curvefitting_xlsxtemplate.ipynb)

Includes use of either CARMEN_template.xlsm (macro-enabled) or CARMEN_template.xlsx.

Step 1:
Groups four-channel CSVs by image and merges them into one Excel file per image.

Step 2:
Inserts raw data into the Excel template and maps channels to specific sheets (BFP, GFP, RFP, NIR).
Cleans up template sheets by removing trailing rows.

Step 3:
Fits exponential decay curves to ROI bleaching traces using scipy.optimize.curve_fit.
Background correction and optional outlier removal are included.

---

CARMEN_onset_return_oscillation.ipynb

Fully automated batch analysis of multi-channel calcium imaging data. Processes all .xlsm files in a folder and analyses each 3-channel dataset (e.g. GFP, RFP, NIR).

• Applies Savitzky-Golay smoothing and derivative-based onset detection.
• Calculates return-to-baseline times and response durations.
• Detects oscillations (peaks and troughs) between onset and return.
• Generates three plots per dataset (zoomed view, oscillations, global signal) and saves them as PNGs.
• Saves per-dataset results (timing, durations, oscillation count) into an Excel file.
• Logs all parameter settings to a .txt file for reproducibility.

Step 1:
Test and adjust parameters for peak detection, smoothing, and baseline return.

Step 2:
Run full batch analysis on all .xlsm files in the selected folder.

---

📥 Input Format

• .csv files from Fiji with filenames like:
  Image01_Intensity FP1.csv, ..., Image01_Intensity FP4.csv
• Each file should include one ROI per row and have the last ROI as background.
• Templates for later steps:
  • CARMEN_template.xlsm or CARMEN_template.xlsx must be present in the specified folder.

---

📤 Output

• .xlsx files per image containing:
  • Sheets: FP1–FP4 (raw)
  • Sheets: FP1bs–FP4bs (background-subtracted)
• Combined .xlsx across multiple images, with:
  • Mean values (Mean1–Mean50)
  • Auto-selected background value
  • Background-subtracted values (BS1–BS50)
• Optional: Excel files with fitted bleach curves starting in column DF.

---

📦 Requirements

• Required libraries (can be installed via pip install -r requirements.txt):

text pandas numpy scipy matplotlib openpyxl xlwings

---

📚 Citation

If you use CARMEN in your research, please cite:

Anna Lischnig, Yusuf Erdoğan, Benjamin Gottschalk, et al.
A Quad-Cistronic Fluorescent Biosensor System for Real-Time Detection of Subcellular Ca²⁺ Signals.
Authorea, March 03, 2025. DOI: 10.22541/au.174098553.38904074/v1

Lischnig, A. (2025). CARMEN: Calcium Multiplexing Analysis Workflow (Version v1.1) [Computer software]. https://doi.org/10.5281/zenodo.15705614