Stochastic Simulation
Also known as: Gillespie algorithm, SSA
Simulation method that models biochemical reactions as discrete, probabilistic events to capture intrinsic noise in gene expression.
Stochastic Simulation is a computational method that models each biochemical reaction as a discrete random event, capturing the inherent probabilistic nature of molecular interactions in living cells 1.
How It Works
The Stochastic Simulation Algorithm (SSA), introduced by Gillespie in 1977, tracks integer molecule counts and simulates one reaction at a time. At each step, the algorithm calculates propensity functions for all possible reactions based on current molecule numbers and rate constants, then randomly selects which reaction fires and when.
This approach produces individual trajectories that differ from run to run, reflecting the noise observed in real biological systems. Ensemble averages over many trajectories yield probability distributions of molecular species, revealing phenomena like bimodal gene expression that deterministic ODE models cannot capture.
Because the exact SSA simulates every single reaction event, it becomes computationally expensive for systems with abundant, fast-reacting species. Approximate methods such as tau-leaping bundle multiple reactions into time intervals, trading some accuracy for significant speedups 2.
Computational Considerations
Modern implementations parallelize thousands of SSA trajectories on GPUs using frameworks like CUDA or JAX. Machine learning surrogate models trained on SSA output can predict distribution statistics without running full simulations, enabling rapid design-space exploration for synthetic circuit engineering 1.
Woolf Software specializes in computational modeling and simulation for biological systems. Get in touch.
GPU-accelerated and tau-leaping variants enable high-throughput stochastic simulations, while ML surrogate models can approximate ensemble statistics orders of magnitude faster.