R-leaping:: Accelerating the stochastic simulation algorithm by reaction leaps

Anne Auger

doi:10.1063/1.2218339

Return

R-leaping:: Accelerating the stochastic simulation algorithm by reaction leaps

2006-08-23

PRE

Anne Auger *

Philippe Chatelain

Petros Koumoutsakos

DOI：10.1063/1.2218339

Original

Original request for help

Save

Abstract

En 中文

A novel algorithm is proposed for the acceleration of the exact stochastic simulation algorithm by a predefined number of reaction firings (R-leaping) that may occur across several reaction channels. In the present approach, the numbers of reaction firings are correlated binomial distributions and the sampling procedure is independent of any permutation of the reaction channels. This enables the algorithm to efficiently handle large systems with disparate rates, providing substantial computational savings in certain cases. Several mechanisms for controlling the accuracy and the appearance of negative species are described. The advantages and drawbacks of R-leaping are assessed by simulations on a number of benchmark problems and the results are discussed in comparison with established methods. (c) 2006 American Institute of Physics.

Keywords:

SYSTEMS

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.