Gillespie-based simulation and inference for non-Markovian stochastic reaction networks.

Non-Markovian Gillespie algorithms reveal that history-dependent reaction times can significantly alter biological population predictions, challenging traditional memoryless models.

• Why it matters: Most current simulators assume Markovian kinetics, which overlook memory effects observed in gene regulation, RNA transcription, and infection dynamics, limiting realistic modeling and analysis.
• What they did: The authors developed and benchmarked a unified framework with multiple Gillespie-based algorithms, including exact, rejection-based, delay-based, and hybrid schemes, tested across biological models.
• The result: Non-Markovian waiting times can change population outcomes, and the new methods enable inference of hidden waiting-time distributions, supporting sensitivity analysis; the open-source NoMaSS library facilitates broad application.