Input, Output and the main processes of MIKANA.

MIKANA's input is composed of a time course data and a set of reactions known to be absent in the final solution set. There are three different types of information criteria: (i) Empirical; (ii) Akaike; and (iii) Bayesian. The user can also select between first– or second–order reactions (Reaction Order) and a maximum molecularity (one or two molecules per chemical species per reaction) to work with. The core process of MIKANA is the construction of a model design matrix with all possible elementary chemical reactions. Other processes include the exclusion of specified reactions and a model selection process. The output of MIKANA consists of a set of elementary reactions used in the model selection process and a set of predicted reactions from which we obtain the differential equations.