Skip to content. | Skip to navigation

Personal tools

Stefan Leye, Corrado Priami, and Adelinde M Uhrmacher (2007)

POp-BetaSim - A Parallel, Bounded Optimistic Simulator for Beta-binders

Poster, , Computational Methods in Systems Biology.

Beta-binders is a comparatively new modeling formalism introduced for systems biology. A stochastic extension (analogous to the stochastic π-Calculus) makes a stochastic discrete event simulation of Beta-binders models possible. Since this type of simulation can be very expensive (especially for simulating cell biological systems), executing the simulation in parallel can be very beneficial. To ensure that the events in a parallel simulation are processed in time stamp order, two main concepts exist. Conservative approaches execute only those events, which are safe to process, by using look aheads. Optimistic attempts execute all events which are able to process. If in a simulator an event is received with a lower time stamp than the last executed event a roll back must be performed. Since look aheads are hard to define in a  stochastic environment an optimistic simulator appears more promising for realizing a parallel execution. The parallel simulator POp-BetaSim is introduced. It is devided into simulator components, responsible for the execution of the internal π-processes and a coordinator component managing the interaction between the boxes. This concept benefits from a distributed execution by running the components on several physical processors. The parallel execution is put into effect by a bounded optimistic concept using the next interaction event as a dynamic barrier until which internal events can process freely. This reduces unnecessary rollbacks by preventing the components from processing too far into the future.

Document Actions