A systems biology approach to qualitative models validation in signaling pathways

Recently, different computational toolkits have been proposed that allow both to formally describe a biochemical system and to perform in silico experimentation mimicking perturbative wet-lab experiments. They should provide those features needed for performing a what-if analysis of the considered biological system, where the operator gives some perturbations and observes the responses of the system. This kind of experiments is particularly useful when the studied system is represented by externally-driven metabolic pathways, e.g. signaling pathways triggered by exogenous signals. We developed a simulator called Quick Direct-method Controlled (QDC for short) that provide a quick and biologists-friendly representation of metabolic experiments with a comprehensive description of the most frequently used experimental controls in the wet-lab. QDC uses the Gillespies' direct method to simulate a biochemical system, where the user can also specify control statements like: the addition at a given time of a given number of molecules of a given biochemical species; the change, at any time, of the rate of any reaction; the specification of stoichiometric based conditional events, etc. We exploited the possibility to describe, by using QDC, the dynamics of a metabolic system to test the two most reliable models proposed in the literature to explain the photoresponses in the Archaeon Halobacterium salinarium. Both these models are, in fact, qualitative and based on the Sensory Rodhopsin I photocycle, even they differ in counting the different spectroscopic states in the photocycle and in assigning the biochemical role of them. We described these models in QDC, by optimizing the collection from the literature (when available) or the inference of the dynamic coefficients of the various biochemical reactions involved in the signaling pathway. We then derive from each original qualitative model, a dynamic model capable to respond to simulated stimuli. We run several simulations by screening any experimental condition in which the photobehavior of H. salinarium has been reported in the literature. We compared the responses obtained by the two simulated models with the experimentally collected data and discuss the performances of both the models for this broad spectrum of trials.