Year: 2008
Type: report
Abstract: We define the syntax and semantics of a new temporal logic called probabilistic LTL with numerical constraints (PLTLc). \nWe introduce an efficient model checker for PLTLc properties. The efficiency of the model checker is through approximation \nusing Monte Carlo sampling of finite paths through the model’s state space (simulation outputs) and parallel model checking \nof the paths. Our model checking method can be applied to any model producing quantitative output – continuous or \nstochastic, including those with complex dynamics and those with an infinite state space. Furthermore, our offline approach \nallows the analysis of observed (real-life) behaviour traces. We find in this paper that PLTLc properties with constraints \nover free variables can replace full model checking experiments, resulting in a significant gain in efficiency. This overcomes \none disadvantage of model checking experiments which is that the complexity depends on system granularity and number of \nvariables, and quickly becomes infeasible. We focus on models of biochemical networks, and specifically in this paper on \nintracellular signalling pathways; however our method can be applied to a wide range of biological as well as technical \nsystems and their models. Our work contributes to the emerging field of synthetic biology by proposing a rigourous approach \nfor the structured formal engineering of biological systems.
Authors: David Gilbert, Robin Donaldson
Institution: University of Glasgow
Language: English
Subfield: Computational Theory and Mathematics
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W138443365