CalDyn: an advanced high-throughput automated algorithm for monitoring intracellular calcium dynamics and electrophysiological signals in patient-derived cells
12 affiliations
- 1PhyMedExp, Univ Montpellier, INSERM, CNRS, Montpellier, France.
- 2Cardiovascular Research Program, VA New York Harbor Healthcare System, New York, United States.
- 3International Center in Fundamental and Engineering Thermophysics, Tianjin University of Commerce, Tianjin, China.
- 4Energy and Thermo-Fluid group, Lebanese International University, Bekaa, Lebanon.
- 5Institute of Myology, Center of Research in Myology, Sorbonne University, Inserm, Paris, France.
- 6MicroBrain Biotech S.A.S., Marly Le-Roi, France.
- 7Department of Medicine, Cell Biology and Pharmacology, State University of New York Downstate Health Sciences University, New York, NY, United States.
- 8IGF, Univ Montpellier, INSERM, CNRS, Montpellier, France.
- 9Biofabics, Porto, Portugal.
- 10Department of Medicine, New York University Grossman School of Medicine, New York, United States.
- 11University of Lyon, CNRS, Institut Lumière Matière, Villeurbanne, France.
- 12Biocampus, Univ Montpellier, CNRS, INSERM, Montpellier, France.
Abstract
Background Quantitative analysis of intracellular calcium and electrophysiological signals is challenging when signals exhibit baseline drift, low-amplitude events, irregular waveforms, or multicomponent transients. We developed CalDyn, a MATLAB-based graphical application for automated peak detection, waveform characterization, event classification, batch processing, quality control and data export. Methods CalDyn combines Savitzky–Golay filtering, LOESS-assisted derivative detection, optional baseline correction, waveform-boundary fitting and rule-based peak classification. It was evaluated on calcium recordings (healthy and disease-associated hiPSC-derived cardiomyocytes, hiPSC-derived neurons, primary skeletal myotubes), current-clamp action potentials and cardiac optical-mapping signals. Outputs were compared with PeakInspector, Clampfit and BV Workbench using paired and Bland–Altman analyses. Detection was assessed against 142 expert-annotated events and synthetic signals of known timing and amplitude. Results CalDyn achieved 94.4% sensitivity, 89.9% precision and an F1 score of 0.92. Performance peaked for healthy cardiomyocytes, paced myotubes and action potentials, with lower recall in arrhythmic cardiomyocytes and variable precision in neurons. CalDyn detected candidate low-amplitude diastolic and multicomponent Ca2+ events in arrhythmic cardiomyocytes, although the biological identity of CalDyn-only events was not independently confirmed. Comparisons with software revealed parameter-specific systematic differences, notably for rise rates and classifications. Action-potential measurements showed limited mean bias but non-negligible limits of agreement. In synthetic signals, amplitude error increased with noise and uncorrected drift, and timing error with low sampling rates. Detection remained stable across moderate parameter variation. Conclusions CalDyn provides an accessible, reproducible and versatile workflow for automated analysis of heterogeneous calcium and electrophysiological signals. With appropriate quality control, it offers a robust alternative to modality-specific tools.
Related
- Validation of a Revised V-Slope Algorithm for Automated First Ventilatory Threshold Detection in Healthy Adults
- Development and validation of a higher-throughput, automated FAMA assay for varicella-zoster virus antibody detection, offering a scalable alternative to the classic 12-well slide method
- Development of an artificial intelligence algorithm for automated detection of cardiovascular intervention-related complications.
- A novel morphological assessment approach for abdominal aortic aneurysms: automated segmentation and measurement using a semi-supervised deep learning algorithm
- Continuous glucose monitoring, automated insulin delivery, and remote care in adolescents with type 1 diabetes: advances toward integrated digital management
BibTeX
@article{sleiman2026caldyn,
title = {{CalDyn: an advanced high-throughput automated algorithm for monitoring intracellular calcium dynamics and electrophysiological signals in patient-derived cells}},
author = {Yvonne Sleiman and Georges El Achkar and Sarah Colombani and Eric Estève and Jean-Baptiste Reisqs and Marie Kervella and Lamia Goual and Elisa Bounasri and Maria João Ferreira and Michael Cupelli and Angelo G Torrente and Jérôme Thireau and Pedro F Costa and Mohamed Boutjdir and Alain Lacampagne and Antoine Muchir and Pascal Seyer and Hélène Delanoë-Ayari and Albano C Meli},
journal = {Briefings in Bioinformatics},
year = {2026},
month = nov,
doi = {10.1093/bib/bbag602},
url = {https://hal.science/hal-05857114v1},
note = {CC BY},
}
RIS
TY - JOUR
TI - CalDyn: an advanced high-throughput automated algorithm for monitoring intracellular calcium dynamics and electrophysiological signals in patient-derived cells
AU - Yvonne Sleiman
AU - Georges El Achkar
AU - Sarah Colombani
AU - Eric Estève
AU - Jean-Baptiste Reisqs
AU - Marie Kervella
AU - Lamia Goual
AU - Elisa Bounasri
AU - Maria João Ferreira
AU - Michael Cupelli
AU - Angelo G Torrente
AU - Jérôme Thireau
AU - Pedro F Costa
AU - Mohamed Boutjdir
AU - Alain Lacampagne
AU - Antoine Muchir
AU - Pascal Seyer
AU - Hélène Delanoë-Ayari
AU - Albano C Meli
PY - 2026
DA - 2026/11/01
JO - Briefings in Bioinformatics
DO - 10.1093/bib/bbag602
UR - https://hal.science/hal-05857114v1
AB - Background Quantitative analysis of intracellular calcium and electrophysiological signals is challenging when signals exhibit baseline drift, low-amplitude events, irregular waveforms, or multicomponent transients. We developed CalDyn, a MATLAB-based graphical application for automated peak detection, waveform characterization, event classification, batch processing, quality control and data export. Methods CalDyn combines Savitzky–Golay filtering, LOESS-assisted derivative detection, optional baseline correction, waveform-boundary fitting and rule-based peak classification. It was evaluated on calcium recordings (healthy and disease-associated hiPSC-derived cardiomyocytes, hiPSC-derived neurons, primary skeletal myotubes), current-clamp action potentials and cardiac optical-mapping signals. Outputs were compared with PeakInspector, Clampfit and BV Workbench using paired and Bland–Altman analyses. Detection was assessed against 142 expert-annotated events and synthetic signals of known timing and amplitude. Results CalDyn achieved 94.4% sensitivity, 89.9% precision and an F1 score of 0.92. Performance peaked for healthy cardiomyocytes, paced myotubes and action potentials, with lower recall in arrhythmic cardiomyocytes and variable precision in neurons. CalDyn detected candidate low-amplitude diastolic and multicomponent Ca2+ events in arrhythmic cardiomyocytes, although the biological identity of CalDyn-only events was not independently confirmed. Comparisons with software revealed parameter-specific systematic differences, notably for rise rates and classifications. Action-potential measurements showed limited mean bias but non-negligible limits of agreement. In synthetic signals, amplitude error increased with noise and uncorrected drift, and timing error with low sampling rates. Detection remained stable across moderate parameter variation. Conclusions CalDyn provides an accessible, reproducible and versatile workflow for automated analysis of heterogeneous calcium and electrophysiological signals. With appropriate quality control, it offers a robust alternative to modality-specific tools.
SN - 1467-5463
SN - 1477-4054
ER -