Simulating microbial community patterning using Biocellion.

Area of Science:

• Computational Biology
• Systems Biology
• Biophysics

Background:

• Mathematical modeling and computer simulations are crucial for understanding cellular interactions and environmental influences.
• Previous simulations, like those by Momeni et al., faced significant time constraints (e.g., MATLAB simulations taking over a week).
• The computational demands of simulating millions of cells in 3D limit the exploration of model parameters and assumptions.

Purpose of the Study:

• To introduce Biocellion, a high-performance software framework designed to accelerate discrete agent-based simulations of biological systems.
• To demonstrate the adaptation of a previously published model (Momeni et al.) to the Biocellion framework as a case study.
• To enable faster hypothesis formation and discovery by improving simulation speed, scale, and accuracy.

Main Methods:

• Biocellion utilizes a discrete agent-based simulation approach.
• The framework is optimized for high-performance computing on multicore workstations and cluster computers.
• Simulations are accelerated by partitioning computational tasks across multiple nodes.

Main Results:

• Biocellion significantly reduces simulation time, completing tasks in hours that previously took over a week using MATLAB.
• The framework supports simulations involving millions to trillions of cells with high accuracy and resolution.
• Adaptation of the Momeni et al. model to Biocellion serves as a practical demonstration of its capabilities.

Conclusions:

• Biocellion offers a substantial improvement in computational efficiency for large-scale biological simulations.
• The framework empowers computational biologists to explore complex biological systems more effectively.
• Faster and more scalable simulations expedite the process of scientific discovery and hypothesis testing.