Fig. 2

Regulatory network steady-state analysis of FlexFlux. 1) The initial state is constructed from qualitative values present in the SBML-qual file (for R1) and quantitative values in the constraint file that are translated into qualitative values by the equivalences written in the SBML-qual network (for M1). 2) From this initial state, the network is iteratively updated using a synchronous update of all components. Integer values represent qualitative states of the components of the network. The update of the network state stops when a state which was already calculated is found (the states shown in red here). The attractor is composed of this state plus the states between the two identical states. 3) The link between qualitative states and continuous intervals must be specified in the SBML-qual file describing the regulatory network in the notes of the QualitativeSpecies tag. For each component where they are provided, all the states of the attractor are translated into intervals. The example is shown for R1 here. 4) When the attractor contains more than one state (cyclic attractor), the average of the upper bounds and lower bounds of all the intervals is used to form a steady-state constraint for components containing equivalences