A community-driven benchmark initiative for evaluating and comparing Center-based agent-Based Models (CBM) in biomedical applications.
This repository contains the benchmark framework and results for comparing various cell-based multiscale models, specifically focusing on off-lattice center-based agent-based models (CBM). The initiative aims to establish standardized benchmarks for evaluating different CBM tools, similar to successful benchmarking efforts like CASP or the MICCAI's medical segmentation one.
The benchmark includes the following CBM tools:
The repository includes several unit tests designed to evaluate different aspects of CBM implementations:
- Single-cell diffusion scenarios (
diffusion_single/
) - Multiple-cell diffusion interactions (work in progress)
- Friction-based mechanics (
mechanics_friction/
) - Pushing mechanics (
mechanics_pushing/
)
- Fixed cell cycle simulations (
fixed_cell_cycle/
) - Stochastic cell cycle simulations (work in progress)
- Monolayer growth simulations (
monolayer/
) - Complex multi-cellular organization (work in progress)
├── Biodynamo/ # BioDynaMo implementation and results
├── Chaste/ # Chaste implementation and results
├── CompuTiX/ # CompuTiX implementation and results
├── PhysiCell/ # PhysiCell implementation and results
├── Tisim/ # TiSim implementation and results
├── experimental_data/ # Reference experimental datasets
└── ResultAnalysis/ # Analysis scripts and comparative results
Each tool's directory contains:
- Source code for benchmark implementations (when available)
- Results from benchmark runs
- Tool-specific configuration files
- Analysis and visualization scripts
The experimental_data/
directory contains reference datasets used for validation:
- 2D monolayer growth data from Bru et al. 1998
- Cell cycle volume dynamics measurements
The ResultAnalysis/
directory provides:
- Scripts for analyzing benchmark results
- Comparative analysis across different tools
- Ground truth data for validation
- Visualization and plotting tools
Each tool's directory contains specific instructions for running the benchmarks. Please refer to the individual README files in each directory for detailed setup and execution instructions.
This is a community-driven initiative, and contributions are welcome. Please contact the repository maintainers or create an issue to discuss potential contributions.
This work is associated with the following publication:
- Ntiniakou et al., "Open Benchmarking for Cell-Based Multiscale Models: Lessons from a Community Initiative", 2025, DOI:TBD
Please refer to the LICENSE file of each tool for terms of use and distribution.
This work is part of a collaborative effort to strengthen the computational biology simulation community and advance the development of Virtual Human Twins (VHT) in biomedical research.
For questions or feedback about this benchmark initiative, please open an issue in this repository or send a mail to Arnau Montagud.
Initiative created and maintained as part of the PerMedCoE project.