Despite recent improvements in molecular techniques, biological knowledge remains incomplete. Any theorizing about living systems is therefore necessarily based on the use of heterogeneous and partial information. Much current research h as focussed successfully on the qualitative behaviors of macromolecular networks. Nonetheless, it is not able to take available quantitative information such as time series protein concentration variations into account. The present work proposes a probabilistic modeling framework that integrates both kinds of information. Average case analysis methods are used in association to Markov chains in order to link qualitative information about transcriptional regulations to quantitative information about protein concentrations. The approach is illustrated by modeling the carbon starvation response in Escherichia coli. Its use accurately predicts the quantitative timeseries evolution of several protein concentrations by using only discrete gene interaction knowledge and very few quantitative observation on a single protein concentration. From that, the modeling technique also derives a ranking of interactions with respect to their importance during the considered experiment. Such a classification is confirmed by literature knowledge. Therefore, as a main novelty, our method permits (i) to integrate few quantitative information into an existing qualitative discrete model and derive new quantitative predictions, (ii) to precisely quantify the robustness and relevance of interactions with respect to phenotypic criteria, and (iii) to extract the key features of the model and design some new experiments.