STEPS

STEPS simulates exact stochastic reaction-diffusion dynamics of biochemical systems within complex three-dimensional geometries to model spatially resolved molecular interactions and transport.

Key Features:

• Complex Geometry Representation: Represents cellular morphologies using detailed 3D tetrahedral meshes to localize chemical species and diffusion within arbitrarily shaped geometries.
• Efficient Simulation Engine: Employs the composition and rejection method, a variation of the Gillespie Stochastic Simulation Algorithm (SSA), with a search-and-update engine to manage diffusion between tetrahedral elements efficiently.
• Systems Biology Markup Language (SBML) support: Supports SBML for importing existing biochemical models.
• Python interface: Provides a Python interface for model construction and simulation control.
• Hybrid Simulation Capabilities: Supports deterministic solvers and well-mixed (non-spatial) conditions in addition to stochastic spatial methods.
• Validation and Performance: Accuracy validated against isolated reaction, diffusion, and reaction-diffusion systems and compared with Smoldyn and MesoRD, often outperforming voxel-based methods in speed for larger systems.
• Solver Constraints: Specifies upper and lower limits on tetrahedron sizes required to maintain solver accuracy.

Scientific Applications:

• Cellular signaling pathways: Models biochemical signaling with spatial heterogeneity and complex cell geometries.
• Systems biology: Enables spatially resolved analysis of biochemical networks and pathway dynamics.
• Cellular biophysics: Simulates diffusion, transport phenomena, and molecular interactions within realistic 3D morphologies.
• Pharmacology: Models spatial aspects of molecular transport and reactions relevant to drug action and distribution.

Methodology:

Simulations run on 3D tetrahedral meshes using the composition and rejection method (a variation of the Gillespie SSA) with a search-and-update engine to handle diffusion between tetrahedra; deterministic and well-mixed solvers are also supported.