Analysis and remedy of negativity problem in hybrid stochastic simulation algorithm and its application

Area of Science:

• Biochemical Systems Analysis
• Computational Biology
• Stochastic Modeling

Background:

• The Haseltine-Rawlings (HR) hybrid method combines deterministic and stochastic simulation algorithms for multiscale biochemical networks.
• This hybrid approach enhances computational efficiency for complex systems.
• A potential issue is the occurrence of negative reactant populations in hybrid simulations.

Purpose of the Study:

• To investigate the impact of negative populations on the accuracy and stability of the HR hybrid method.
• To analyze the negativity problem across various biochemical models.
• To evaluate potential remedies for the negativity issue.

Main Methods:

• Analysis and testing of the HR hybrid method using linear chain, nonlinear reaction, and cell cycle systems.
• Benchmarking accuracy using the second slow reaction firing time.
• Evaluation of three proposed remedies for negative population issues.

Main Results:

• Negative populations typically introduce negligible errors compared to inherent approximation errors of the HR method.
• In some cases, negative populations may even enhance simulation accuracy.
• System stability can be compromised in nonlinear or sensitive systems due to negative populations, potentially causing simulation failure.

Conclusions:

• The Zero-Reaction rule is identified as an efficient and simple remedy for addressing negativity problems in nonlinear and sensitive biochemical systems.
• The HR hybrid method remains a valuable tool for efficient biochemical simulations, with specific considerations for system sensitivity.