Modeling reaction–diffusion processes in simplified to complex astroglial branchlet geometries using different simulation tools
- 1Faculty of Medicine and Health Technology, Tampere University, Tampere, Finland.
Abstract
Computational modeling of spatial biochemical processes in neuroscience is increasingly important as advances in experimental and imaging techniques provide detailed insights into cellular organization. Reaction–diffusion models are commonly used to study the diffusion and interactions of ions and molecules within neuronal cells, and multiple open-access simulation platforms have been developed to support such analysis. However, a comprehensive analysis of commonly used simulation tools, including MCell, STEPS, NeuroRD, and NEURON, remains lacking, particularly for realistic glial cell geometries, as these tools were originally developed for neurons. Here, we present a systematic analysis of these four simulation environments by implementing equivalent calcium diffusion and calcium–calmodulin reaction–diffusion models across a range of spatial domains. These include simplified geometries (cylinders and rectangular cuboids) as well as biologically realistic astroglial morphologies derived from electron microscopy data, including mouse Bergmann glia and mouse and human cortical astrocytes. Using matched kinetic parameters and diffusion constants, we examine simulation outputs and morphological variations across platforms under comparable conditions. We find that all tools produce overall consistent results for the systems considered, indicating that each can reliably capture diffusion and reaction–diffusion dynamics in simplified and complex geometries. At the same time, we show that the structural heterogeneity of branchlets from human astrocytes and mouse Bergmann glia gives rise to distinct calcium signaling, demonstrating that fine-scale morphology critically shapes astroglial dynamics and must be accurately captured by simulation frameworks. In addition, notable differences emerge in geometry representation, implementation strategies, computational aspects, and usability of the tools. By highlighting key trade-offs between geometric fidelity, flexibility, and computational cost in modeling intracellular biochemical processes, these findings provide practical guidance for selecting appropriate simulation tools. More broadly, our results support the use of reaction–diffusion modeling in morphologically realistic astrocytic environments and contribute to enabling mechanistic studies of localized signaling events at synaptic and perisynaptic sites. Author summary: Simulation of cellular reaction–diffusion systems is receiving considerable attention due to increasing computing power and biological knowledge. Several open-access simulation tools have been developed for this purpose and, in this study, we are interested in utilizing and examining four of them, MCell, STEPS, NeuroRD, and NEURON. With these four selected tools, we implement the same diffusion and reaction–diffusion models in cylinder, rectangular cuboid, and branchlets of real astroglial morphologies to perform the simulations. We conclude that the simulation results are consistent between the simulation tools. However, the tools differ in their usability, approaches to morphological representation and implementation, and the support for simulations in different dimensionalities. Understanding the properties and capabilities of these simulation tools will help develop better data-based models.
Related
- A Simplified Density Gradient–Free Method for Isolating Functional Symbiosomes From Legume Nodules
- Development of a simplified intensified fed-batch CHO cell culture processbased on a commercial single-feed strategy
- Simplified in vitro generation of human gastruloids for modelling early development
- A Simplified and Solvent-Reduced HPLC-UV/Vis Method for the Simultaneous Determination of Major Marker Compounds in Rhodiola rosea L.
- TTF2 processes sites of incomplete DNA replication during mitosis via sister-chromatid exchanges
BibTeX
@article{keto2026modeling,
title = {{Modeling reaction–diffusion processes in simplified to complex astroglial branchlet geometries using different simulation tools}},
author = {Laura Keto and Saana Seppälä and Derek Ndubuaku and Annika Mäki and Tuomo Mäki-Marttunen and Marja-Leena Linne and Tiina Manninen},
journal = {PLOS Computational Biology},
year = {2026},
month = oct,
doi = {10.1371/journal.pcbi.1014878},
url = {https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1014878},
note = {CC BY},
}
RIS
TY - JOUR
TI - Modeling reaction–diffusion processes in simplified to complex astroglial branchlet geometries using different simulation tools
AU - Laura Keto
AU - Saana Seppälä
AU - Derek Ndubuaku
AU - Annika Mäki
AU - Tuomo Mäki-Marttunen
AU - Marja-Leena Linne
AU - Tiina Manninen
PY - 2026
DA - 2026/10/28
JO - PLOS Computational Biology
DO - 10.1371/journal.pcbi.1014878
UR - https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1014878
AB - Computational modeling of spatial biochemical processes in neuroscience is increasingly important as advances in experimental and imaging techniques provide detailed insights into cellular organization. Reaction–diffusion models are commonly used to study the diffusion and interactions of ions and molecules within neuronal cells, and multiple open-access simulation platforms have been developed to support such analysis. However, a comprehensive analysis of commonly used simulation tools, including MCell, STEPS, NeuroRD, and NEURON, remains lacking, particularly for realistic glial cell geometries, as these tools were originally developed for neurons. Here, we present a systematic analysis of these four simulation environments by implementing equivalent calcium diffusion and calcium–calmodulin reaction–diffusion models across a range of spatial domains. These include simplified geometries (cylinders and rectangular cuboids) as well as biologically realistic astroglial morphologies derived from electron microscopy data, including mouse Bergmann glia and mouse and human cortical astrocytes. Using matched kinetic parameters and diffusion constants, we examine simulation outputs and morphological variations across platforms under comparable conditions. We find that all tools produce overall consistent results for the systems considered, indicating that each can reliably capture diffusion and reaction–diffusion dynamics in simplified and complex geometries. At the same time, we show that the structural heterogeneity of branchlets from human astrocytes and mouse Bergmann glia gives rise to distinct calcium signaling, demonstrating that fine-scale morphology critically shapes astroglial dynamics and must be accurately captured by simulation frameworks. In addition, notable differences emerge in geometry representation, implementation strategies, computational aspects, and usability of the tools. By highlighting key trade-offs between geometric fidelity, flexibility, and computational cost in modeling intracellular biochemical processes, these findings provide practical guidance for selecting appropriate simulation tools. More broadly, our results support the use of reaction–diffusion modeling in morphologically realistic astrocytic environments and contribute to enabling mechanistic studies of localized signaling events at synaptic and perisynaptic sites. Author summary: Simulation of cellular reaction–diffusion systems is receiving considerable attention due to increasing computing power and biological knowledge. Several open-access simulation tools have been developed for this purpose and, in this study, we are interested in utilizing and examining four of them, MCell, STEPS, NeuroRD, and NEURON. With these four selected tools, we implement the same diffusion and reaction–diffusion models in cylinder, rectangular cuboid, and branchlets of real astroglial morphologies to perform the simulations. We conclude that the simulation results are consistent between the simulation tools. However, the tools differ in their usability, approaches to morphological representation and implementation, and the support for simulations in different dimensionalities. Understanding the properties and capabilities of these simulation tools will help develop better data-based models.
SN - 1553-7358
SN - 1553-734X
ER -