Quantitative metrics for bio-modeling algorithm selection.

In this paper, we report our efforts in developing guidelines that are capable of helping researchers to select algorithms in systems biology modeling. We propose a set of metrics based on discrete observable units in terms of key bio-modeling considerations. We accomplish this by (i) reviewing classical metric definitions, (ii) implementing widely used modeling algorithms on a specific case study, and (iii) testing metrics that are a hybrid of classical metrics and key bio-modeling considerations. The modeling algorithms implemented are Michaelis-Menten kinetics, generalized mass action, flux balance analysis, and metabolic control analysis. This work extends our previous work in developing qualitative guidelines to select bio-modeling algorithms. Our results impact systems biology modeling specifically by increasing the level of confidence for users to select bio-modeling algorithms by using quantitative metrics appropriately.